diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,109398 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 15624,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 74.75566446547442,
+      "learning_rate": 2.1321961620469085e-08,
+      "loss": 1.9973,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 91.65379509395875,
+      "learning_rate": 4.264392324093817e-08,
+      "loss": 1.9259,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 38.6705808580265,
+      "learning_rate": 6.396588486140725e-08,
+      "loss": 1.6888,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 84.20240444687236,
+      "learning_rate": 8.528784648187634e-08,
+      "loss": 1.9685,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 55.46876633349456,
+      "learning_rate": 1.0660980810234542e-07,
+      "loss": 1.9508,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.21171046284991,
+      "learning_rate": 1.279317697228145e-07,
+      "loss": 2.1308,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 75.24800824628849,
+      "learning_rate": 1.4925373134328358e-07,
+      "loss": 1.9951,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 83.31090575969453,
+      "learning_rate": 1.7057569296375268e-07,
+      "loss": 1.8936,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 27.50669913532547,
+      "learning_rate": 1.918976545842218e-07,
+      "loss": 1.9997,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 59.95940536985684,
+      "learning_rate": 2.1321961620469084e-07,
+      "loss": 1.8446,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 54.32684733645877,
+      "learning_rate": 2.3454157782515995e-07,
+      "loss": 1.7228,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 390.7101171781313,
+      "learning_rate": 2.55863539445629e-07,
+      "loss": 1.8747,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.227197760929588,
+      "learning_rate": 2.771855010660981e-07,
+      "loss": 1.7908,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 109.39168455483434,
+      "learning_rate": 2.9850746268656716e-07,
+      "loss": 1.9734,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.213610757289118,
+      "learning_rate": 3.1982942430703626e-07,
+      "loss": 1.835,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 45.91261211926917,
+      "learning_rate": 3.4115138592750537e-07,
+      "loss": 1.921,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.18393428039165,
+      "learning_rate": 3.624733475479744e-07,
+      "loss": 1.834,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.57348286581749,
+      "learning_rate": 3.837953091684436e-07,
+      "loss": 1.9299,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.585655633750115,
+      "learning_rate": 4.0511727078891263e-07,
+      "loss": 1.9469,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 60.94223848739873,
+      "learning_rate": 4.264392324093817e-07,
+      "loss": 1.869,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.57714840363393,
+      "learning_rate": 4.4776119402985074e-07,
+      "loss": 2.0122,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.678568402502876,
+      "learning_rate": 4.690831556503199e-07,
+      "loss": 1.9627,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.777067481596594,
+      "learning_rate": 4.904051172707889e-07,
+      "loss": 0.8568,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.59242729609809,
+      "learning_rate": 5.11727078891258e-07,
+      "loss": 1.9015,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 202.73017865686384,
+      "learning_rate": 5.33049040511727e-07,
+      "loss": 1.7323,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.964599511174995,
+      "learning_rate": 5.543710021321962e-07,
+      "loss": 0.9026,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.737608458054458,
+      "learning_rate": 5.756929637526653e-07,
+      "loss": 0.9136,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.12108998362152,
+      "learning_rate": 5.970149253731343e-07,
+      "loss": 1.8423,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.266371818864066,
+      "learning_rate": 6.183368869936035e-07,
+      "loss": 1.8633,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.123057032072364,
+      "learning_rate": 6.396588486140725e-07,
+      "loss": 1.6249,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 46.065811437693874,
+      "learning_rate": 6.609808102345417e-07,
+      "loss": 1.8451,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 59.4734016081285,
+      "learning_rate": 6.823027718550107e-07,
+      "loss": 1.7741,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 42.77093313117688,
+      "learning_rate": 7.036247334754798e-07,
+      "loss": 1.4831,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.97641684097191,
+      "learning_rate": 7.249466950959488e-07,
+      "loss": 1.7792,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.935501124028264,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 1.6106,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 99.78303234465945,
+      "learning_rate": 7.675906183368872e-07,
+      "loss": 1.6622,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.677559682342803,
+      "learning_rate": 7.889125799573562e-07,
+      "loss": 1.6344,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.15448752703346,
+      "learning_rate": 8.102345415778253e-07,
+      "loss": 1.5766,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.13797235316429,
+      "learning_rate": 8.315565031982943e-07,
+      "loss": 1.6363,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.727516765140575,
+      "learning_rate": 8.528784648187634e-07,
+      "loss": 1.4976,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.603623102670223,
+      "learning_rate": 8.742004264392324e-07,
+      "loss": 1.4115,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.495398936006681,
+      "learning_rate": 8.955223880597015e-07,
+      "loss": 0.7511,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.142525542449864,
+      "learning_rate": 9.168443496801707e-07,
+      "loss": 1.3118,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.85210903182349,
+      "learning_rate": 9.381663113006398e-07,
+      "loss": 1.3083,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.59770840015882,
+      "learning_rate": 9.594882729211088e-07,
+      "loss": 1.5182,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.446754004962443,
+      "learning_rate": 9.808102345415779e-07,
+      "loss": 1.2753,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.993845548365407,
+      "learning_rate": 1.002132196162047e-06,
+      "loss": 1.4776,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.014436115439098,
+      "learning_rate": 1.023454157782516e-06,
+      "loss": 1.2679,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.186550401356996,
+      "learning_rate": 1.044776119402985e-06,
+      "loss": 1.1392,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.666026037825716,
+      "learning_rate": 1.066098081023454e-06,
+      "loss": 1.1814,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.94640533330083,
+      "learning_rate": 1.0874200426439234e-06,
+      "loss": 1.4649,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.880979352545676,
+      "learning_rate": 1.1087420042643924e-06,
+      "loss": 1.4239,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.02563582325072,
+      "learning_rate": 1.1300639658848615e-06,
+      "loss": 0.6884,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.52058931086476,
+      "learning_rate": 1.1513859275053305e-06,
+      "loss": 1.3493,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.054409686684624,
+      "learning_rate": 1.1727078891257996e-06,
+      "loss": 1.1768,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.73362326265402,
+      "learning_rate": 1.1940298507462686e-06,
+      "loss": 1.2696,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.397851676032585,
+      "learning_rate": 1.2153518123667379e-06,
+      "loss": 0.7726,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.676044200396037,
+      "learning_rate": 1.236673773987207e-06,
+      "loss": 1.3939,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.161261659978955,
+      "learning_rate": 1.257995735607676e-06,
+      "loss": 1.1889,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5926913317703881,
+      "learning_rate": 1.279317697228145e-06,
+      "loss": 0.6688,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.93312244566375,
+      "learning_rate": 1.300639658848614e-06,
+      "loss": 1.4416,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.933032483875312,
+      "learning_rate": 1.3219616204690834e-06,
+      "loss": 1.2598,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.967848744275823,
+      "learning_rate": 1.3432835820895524e-06,
+      "loss": 1.2783,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.340276980853929,
+      "learning_rate": 1.3646055437100215e-06,
+      "loss": 0.994,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.262576178875534,
+      "learning_rate": 1.3859275053304905e-06,
+      "loss": 1.1543,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.528220629297374,
+      "learning_rate": 1.4072494669509596e-06,
+      "loss": 1.0792,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.795280228304279,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 1.0893,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.437849568024497,
+      "learning_rate": 1.4498933901918977e-06,
+      "loss": 1.195,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 38.82317123725269,
+      "learning_rate": 1.4712153518123667e-06,
+      "loss": 1.1035,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.145922492006004,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 1.3928,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.3286771129107935,
+      "learning_rate": 1.5138592750533053e-06,
+      "loss": 1.1409,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.410535156800897,
+      "learning_rate": 1.5351812366737743e-06,
+      "loss": 1.1265,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.332049785423163,
+      "learning_rate": 1.5565031982942434e-06,
+      "loss": 1.2574,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.638260425266138,
+      "learning_rate": 1.5778251599147124e-06,
+      "loss": 1.2196,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.938838580240725,
+      "learning_rate": 1.5991471215351815e-06,
+      "loss": 1.0632,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.055389497601755,
+      "learning_rate": 1.6204690831556505e-06,
+      "loss": 1.2252,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.61373064692541,
+      "learning_rate": 1.6417910447761196e-06,
+      "loss": 0.7141,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.563077448934267,
+      "learning_rate": 1.6631130063965886e-06,
+      "loss": 1.131,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5886656692796963,
+      "learning_rate": 1.6844349680170577e-06,
+      "loss": 1.1594,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.412013711668578,
+      "learning_rate": 1.7057569296375267e-06,
+      "loss": 1.0568,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4775552168645025,
+      "learning_rate": 1.7270788912579958e-06,
+      "loss": 0.7265,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 33.74671378499939,
+      "learning_rate": 1.7484008528784648e-06,
+      "loss": 1.1688,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8714828765782263,
+      "learning_rate": 1.7697228144989339e-06,
+      "loss": 1.0611,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.800700311355135,
+      "learning_rate": 1.791044776119403e-06,
+      "loss": 1.0512,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.178209639833226,
+      "learning_rate": 1.812366737739872e-06,
+      "loss": 1.0955,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.850338088395606,
+      "learning_rate": 1.8336886993603415e-06,
+      "loss": 1.1512,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.108077873613537,
+      "learning_rate": 1.8550106609808105e-06,
+      "loss": 1.0072,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.42477604880931,
+      "learning_rate": 1.8763326226012796e-06,
+      "loss": 1.1581,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.291497760145232,
+      "learning_rate": 1.8976545842217486e-06,
+      "loss": 1.1741,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.795896011414218,
+      "learning_rate": 1.9189765458422177e-06,
+      "loss": 1.1019,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.980961330025634,
+      "learning_rate": 1.9402985074626867e-06,
+      "loss": 1.1584,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0658502746639766,
+      "learning_rate": 1.9616204690831558e-06,
+      "loss": 1.1603,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.854658885282486,
+      "learning_rate": 1.982942430703625e-06,
+      "loss": 1.0679,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9759137022141797,
+      "learning_rate": 2.004264392324094e-06,
+      "loss": 0.6577,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.458671886067648,
+      "learning_rate": 2.025586353944563e-06,
+      "loss": 0.9854,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4087091852764033,
+      "learning_rate": 2.046908315565032e-06,
+      "loss": 0.7764,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3473219201349185,
+      "learning_rate": 2.068230277185501e-06,
+      "loss": 0.8168,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.24017237992957,
+      "learning_rate": 2.08955223880597e-06,
+      "loss": 0.9986,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.271280013563478,
+      "learning_rate": 2.110874200426439e-06,
+      "loss": 1.187,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6141084380888429,
+      "learning_rate": 2.132196162046908e-06,
+      "loss": 0.8403,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0873842478238975,
+      "learning_rate": 2.1535181236673773e-06,
+      "loss": 1.0857,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.175568929003719,
+      "learning_rate": 2.1748400852878467e-06,
+      "loss": 1.1336,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.81300275177583,
+      "learning_rate": 2.1961620469083158e-06,
+      "loss": 1.0611,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9660203555722933,
+      "learning_rate": 2.217484008528785e-06,
+      "loss": 0.9257,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.325048179950666,
+      "learning_rate": 2.238805970149254e-06,
+      "loss": 1.0623,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8888509944583909,
+      "learning_rate": 2.260127931769723e-06,
+      "loss": 0.7216,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4503986415603831,
+      "learning_rate": 2.281449893390192e-06,
+      "loss": 0.6774,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.628600619860332,
+      "learning_rate": 2.302771855010661e-06,
+      "loss": 1.1485,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.912538384503131,
+      "learning_rate": 2.32409381663113e-06,
+      "loss": 1.088,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.126607546904964,
+      "learning_rate": 2.345415778251599e-06,
+      "loss": 0.9893,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 109.89897590220892,
+      "learning_rate": 2.366737739872068e-06,
+      "loss": 1.1651,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.125773330417651,
+      "learning_rate": 2.3880597014925373e-06,
+      "loss": 0.9954,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.215364970579222,
+      "learning_rate": 2.4093816631130067e-06,
+      "loss": 1.0986,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.286275598754983,
+      "learning_rate": 2.4307036247334758e-06,
+      "loss": 1.1521,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 52.980919444311596,
+      "learning_rate": 2.452025586353945e-06,
+      "loss": 1.1604,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.9085310920943,
+      "learning_rate": 2.473347547974414e-06,
+      "loss": 1.1749,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.828364228614898,
+      "learning_rate": 2.494669509594883e-06,
+      "loss": 0.9211,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.831412107592298,
+      "learning_rate": 2.515991471215352e-06,
+      "loss": 1.1119,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.725847100569869,
+      "learning_rate": 2.537313432835821e-06,
+      "loss": 1.1391,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.848001479866944,
+      "learning_rate": 2.55863539445629e-06,
+      "loss": 1.0992,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2268893217593397,
+      "learning_rate": 2.579957356076759e-06,
+      "loss": 0.5676,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.560531722961582,
+      "learning_rate": 2.601279317697228e-06,
+      "loss": 1.0214,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.652712369483309,
+      "learning_rate": 2.6226012793176977e-06,
+      "loss": 1.0499,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.772981802804721,
+      "learning_rate": 2.6439232409381667e-06,
+      "loss": 1.0375,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3825513989693334,
+      "learning_rate": 2.6652452025586358e-06,
+      "loss": 1.0373,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2782635702044827,
+      "learning_rate": 2.686567164179105e-06,
+      "loss": 0.6602,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.286853779307659,
+      "learning_rate": 2.707889125799574e-06,
+      "loss": 0.9742,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.904758010191905,
+      "learning_rate": 2.729211087420043e-06,
+      "loss": 1.0042,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6786297665547942,
+      "learning_rate": 2.750533049040512e-06,
+      "loss": 1.0645,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.4160743162316,
+      "learning_rate": 2.771855010660981e-06,
+      "loss": 1.101,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0847082711024725,
+      "learning_rate": 2.79317697228145e-06,
+      "loss": 0.9976,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.14698090074991,
+      "learning_rate": 2.814498933901919e-06,
+      "loss": 0.9802,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6886095392691223,
+      "learning_rate": 2.835820895522388e-06,
+      "loss": 1.0987,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.180134186371319,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.9921,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.027770681042015,
+      "learning_rate": 2.8784648187633263e-06,
+      "loss": 1.0334,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 37.95427091921639,
+      "learning_rate": 2.8997867803837954e-06,
+      "loss": 1.0212,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8935688975941294,
+      "learning_rate": 2.9211087420042644e-06,
+      "loss": 0.9739,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9723945162723224,
+      "learning_rate": 2.9424307036247335e-06,
+      "loss": 0.7736,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.569738738640676,
+      "learning_rate": 2.9637526652452025e-06,
+      "loss": 0.9971,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.754210054657566,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 1.0319,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.5446006812384985,
+      "learning_rate": 3.006396588486141e-06,
+      "loss": 0.8556,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.095394843274608,
+      "learning_rate": 3.0277185501066105e-06,
+      "loss": 0.9727,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.86376586235683,
+      "learning_rate": 3.0490405117270796e-06,
+      "loss": 0.9073,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.662406336926879,
+      "learning_rate": 3.0703624733475486e-06,
+      "loss": 0.9842,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6067959395432805,
+      "learning_rate": 3.0916844349680177e-06,
+      "loss": 0.9029,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2950058386403724,
+      "learning_rate": 3.1130063965884867e-06,
+      "loss": 0.6515,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5687185671629458,
+      "learning_rate": 3.1343283582089558e-06,
+      "loss": 0.7011,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.07848243396814,
+      "learning_rate": 3.155650319829425e-06,
+      "loss": 1.0291,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4346353590263594,
+      "learning_rate": 3.176972281449894e-06,
+      "loss": 1.093,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 63.6742879483503,
+      "learning_rate": 3.198294243070363e-06,
+      "loss": 0.9758,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.117016185355896,
+      "learning_rate": 3.219616204690832e-06,
+      "loss": 1.0009,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.548381204978389,
+      "learning_rate": 3.240938166311301e-06,
+      "loss": 1.0291,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5165285687818515,
+      "learning_rate": 3.26226012793177e-06,
+      "loss": 0.9833,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.523274552488534,
+      "learning_rate": 3.283582089552239e-06,
+      "loss": 1.065,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.742060814076015,
+      "learning_rate": 3.304904051172708e-06,
+      "loss": 1.0866,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.81976732605988,
+      "learning_rate": 3.3262260127931773e-06,
+      "loss": 1.0181,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.352924616120804,
+      "learning_rate": 3.3475479744136463e-06,
+      "loss": 0.8777,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4198200341284422,
+      "learning_rate": 3.3688699360341154e-06,
+      "loss": 0.6617,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.458180736858485,
+      "learning_rate": 3.3901918976545844e-06,
+      "loss": 1.0517,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6806097568496776,
+      "learning_rate": 3.4115138592750535e-06,
+      "loss": 0.9584,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4274415621847685,
+      "learning_rate": 3.4328358208955225e-06,
+      "loss": 1.0007,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.118691966498492,
+      "learning_rate": 3.4541577825159916e-06,
+      "loss": 1.0524,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.280074372204677,
+      "learning_rate": 3.4754797441364606e-06,
+      "loss": 0.8102,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.932381666637639,
+      "learning_rate": 3.4968017057569297e-06,
+      "loss": 0.6876,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.6248824217305655,
+      "learning_rate": 3.5181236673773987e-06,
+      "loss": 0.8543,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3299349495495754,
+      "learning_rate": 3.5394456289978678e-06,
+      "loss": 0.6404,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.917225849369174,
+      "learning_rate": 3.560767590618337e-06,
+      "loss": 0.9923,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3984824273664318,
+      "learning_rate": 3.582089552238806e-06,
+      "loss": 0.6342,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.6082142292597,
+      "learning_rate": 3.603411513859275e-06,
+      "loss": 0.9797,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.428114837510602,
+      "learning_rate": 3.624733475479744e-06,
+      "loss": 1.0952,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.22367768308715,
+      "learning_rate": 3.6460554371002135e-06,
+      "loss": 0.6836,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9579224158971686,
+      "learning_rate": 3.667377398720683e-06,
+      "loss": 1.0057,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.488184026920131,
+      "learning_rate": 3.688699360341152e-06,
+      "loss": 0.9536,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.39886023798639,
+      "learning_rate": 3.710021321961621e-06,
+      "loss": 0.6287,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4374334258521946,
+      "learning_rate": 3.73134328358209e-06,
+      "loss": 0.7423,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.380753951776869,
+      "learning_rate": 3.752665245202559e-06,
+      "loss": 1.0811,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.163003531612366,
+      "learning_rate": 3.773987206823028e-06,
+      "loss": 1.0526,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.540867081107244,
+      "learning_rate": 3.7953091684434973e-06,
+      "loss": 0.9093,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.962806580862798,
+      "learning_rate": 3.816631130063966e-06,
+      "loss": 0.9642,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.584537008347239,
+      "learning_rate": 3.837953091684435e-06,
+      "loss": 0.9619,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.049378933550033,
+      "learning_rate": 3.859275053304904e-06,
+      "loss": 0.908,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.459828666866898,
+      "learning_rate": 3.8805970149253735e-06,
+      "loss": 1.1071,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.660022726918198,
+      "learning_rate": 3.9019189765458425e-06,
+      "loss": 0.9436,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.052709169503742,
+      "learning_rate": 3.9232409381663116e-06,
+      "loss": 1.1742,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2387889561823457,
+      "learning_rate": 3.944562899786781e-06,
+      "loss": 0.7177,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.240689670820267,
+      "learning_rate": 3.96588486140725e-06,
+      "loss": 1.0967,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.883786280040918,
+      "learning_rate": 3.987206823027719e-06,
+      "loss": 1.0158,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.459345200403704,
+      "learning_rate": 4.008528784648188e-06,
+      "loss": 0.9888,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.275535664970997,
+      "learning_rate": 4.029850746268657e-06,
+      "loss": 0.9766,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.192770320236586,
+      "learning_rate": 4.051172707889126e-06,
+      "loss": 0.9858,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.243950366816717,
+      "learning_rate": 4.072494669509595e-06,
+      "loss": 0.9152,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.238053923272308,
+      "learning_rate": 4.093816631130064e-06,
+      "loss": 0.9968,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.112336959976924,
+      "learning_rate": 4.115138592750533e-06,
+      "loss": 1.0295,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.642886849574858,
+      "learning_rate": 4.136460554371002e-06,
+      "loss": 0.8412,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6953443054479873,
+      "learning_rate": 4.157782515991471e-06,
+      "loss": 0.9182,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.501456811928708,
+      "learning_rate": 4.17910447761194e-06,
+      "loss": 0.8952,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.045754964222702,
+      "learning_rate": 4.200426439232409e-06,
+      "loss": 1.0612,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.1190071428130555,
+      "learning_rate": 4.221748400852878e-06,
+      "loss": 0.9418,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9699399753481166,
+      "learning_rate": 4.243070362473347e-06,
+      "loss": 1.0669,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.241327723793498,
+      "learning_rate": 4.264392324093816e-06,
+      "loss": 0.9357,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.422829328550876,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 1.0191,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2950473387284753,
+      "learning_rate": 4.3070362473347545e-06,
+      "loss": 0.9828,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5968400193734,
+      "learning_rate": 4.3283582089552236e-06,
+      "loss": 0.9215,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5172669576126547,
+      "learning_rate": 4.3496801705756935e-06,
+      "loss": 0.9644,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4136208741481644,
+      "learning_rate": 4.3710021321961625e-06,
+      "loss": 0.9927,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.877201931803291,
+      "learning_rate": 4.3923240938166316e-06,
+      "loss": 0.9804,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.34443333669772,
+      "learning_rate": 4.413646055437101e-06,
+      "loss": 0.9667,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.056128239653811,
+      "learning_rate": 4.43496801705757e-06,
+      "loss": 0.9865,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.053886406356322,
+      "learning_rate": 4.456289978678039e-06,
+      "loss": 0.9942,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.835968784392479,
+      "learning_rate": 4.477611940298508e-06,
+      "loss": 1.1026,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.31803150655678,
+      "learning_rate": 4.498933901918977e-06,
+      "loss": 0.945,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.995070659866954,
+      "learning_rate": 4.520255863539446e-06,
+      "loss": 0.9873,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.49284582674016,
+      "learning_rate": 4.541577825159915e-06,
+      "loss": 0.9378,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.494576635409437,
+      "learning_rate": 4.562899786780384e-06,
+      "loss": 1.0385,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7976696894804984,
+      "learning_rate": 4.584221748400853e-06,
+      "loss": 1.0365,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6019133454365053,
+      "learning_rate": 4.605543710021322e-06,
+      "loss": 0.6208,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.6298942037150725,
+      "learning_rate": 4.626865671641791e-06,
+      "loss": 1.107,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.531488577551905,
+      "learning_rate": 4.64818763326226e-06,
+      "loss": 0.9483,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2007138965425934,
+      "learning_rate": 4.669509594882729e-06,
+      "loss": 0.9565,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.321270596913737,
+      "learning_rate": 4.690831556503198e-06,
+      "loss": 0.7395,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.233091764763394,
+      "learning_rate": 4.712153518123667e-06,
+      "loss": 1.0299,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1026858527227166,
+      "learning_rate": 4.733475479744136e-06,
+      "loss": 0.8832,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.873417040779917,
+      "learning_rate": 4.7547974413646055e-06,
+      "loss": 0.979,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.101027726199242,
+      "learning_rate": 4.7761194029850745e-06,
+      "loss": 0.9722,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.091370140592939,
+      "learning_rate": 4.797441364605544e-06,
+      "loss": 0.9571,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3928716630747884,
+      "learning_rate": 4.8187633262260135e-06,
+      "loss": 0.6787,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.610957329904152,
+      "learning_rate": 4.8400852878464825e-06,
+      "loss": 0.9409,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.11885510083222,
+      "learning_rate": 4.8614072494669516e-06,
+      "loss": 0.915,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.90725840141939,
+      "learning_rate": 4.882729211087421e-06,
+      "loss": 0.9618,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.029481063688279,
+      "learning_rate": 4.90405117270789e-06,
+      "loss": 1.0401,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.022097079445353,
+      "learning_rate": 4.925373134328359e-06,
+      "loss": 0.9164,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1600669186754327,
+      "learning_rate": 4.946695095948828e-06,
+      "loss": 0.9792,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.00774474721482,
+      "learning_rate": 4.968017057569297e-06,
+      "loss": 0.9897,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0019753178035335,
+      "learning_rate": 4.989339019189766e-06,
+      "loss": 0.7495,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7096519887624186,
+      "learning_rate": 5.010660980810235e-06,
+      "loss": 0.9497,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1030566970843356,
+      "learning_rate": 5.031982942430704e-06,
+      "loss": 0.991,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6466873675422202,
+      "learning_rate": 5.053304904051173e-06,
+      "loss": 0.995,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.994464486798373,
+      "learning_rate": 5.074626865671642e-06,
+      "loss": 0.9024,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3988523732082916,
+      "learning_rate": 5.095948827292111e-06,
+      "loss": 1.0067,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.449326567889687,
+      "learning_rate": 5.11727078891258e-06,
+      "loss": 0.963,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8555481410074717,
+      "learning_rate": 5.138592750533049e-06,
+      "loss": 0.948,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2369839919355243,
+      "learning_rate": 5.159914712153518e-06,
+      "loss": 1.0558,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.427148284882106,
+      "learning_rate": 5.181236673773987e-06,
+      "loss": 0.9378,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3403852772485119,
+      "learning_rate": 5.202558635394456e-06,
+      "loss": 0.6677,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3519928518869537,
+      "learning_rate": 5.2238805970149255e-06,
+      "loss": 0.6592,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.707623668391948,
+      "learning_rate": 5.245202558635395e-06,
+      "loss": 0.8986,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7424104007841765,
+      "learning_rate": 5.2665245202558636e-06,
+      "loss": 0.9589,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.925963922572754,
+      "learning_rate": 5.2878464818763335e-06,
+      "loss": 0.98,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0129289092743905,
+      "learning_rate": 5.309168443496802e-06,
+      "loss": 1.008,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.148041997266006,
+      "learning_rate": 5.3304904051172716e-06,
+      "loss": 0.9842,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.293940538365192,
+      "learning_rate": 5.351812366737741e-06,
+      "loss": 1.0526,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5029864461238676,
+      "learning_rate": 5.37313432835821e-06,
+      "loss": 0.9388,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.914049556756784,
+      "learning_rate": 5.394456289978679e-06,
+      "loss": 0.9527,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.40247927074392,
+      "learning_rate": 5.415778251599148e-06,
+      "loss": 0.9251,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8761023142703415,
+      "learning_rate": 5.437100213219617e-06,
+      "loss": 0.8993,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.369440357290614,
+      "learning_rate": 5.458422174840086e-06,
+      "loss": 0.6138,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3374202692491117,
+      "learning_rate": 5.479744136460555e-06,
+      "loss": 0.7322,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5258659673919803,
+      "learning_rate": 5.501066098081024e-06,
+      "loss": 0.9697,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2799375290557586,
+      "learning_rate": 5.522388059701493e-06,
+      "loss": 0.9243,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0359080430081056,
+      "learning_rate": 5.543710021321962e-06,
+      "loss": 0.9388,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.209734905820337,
+      "learning_rate": 5.565031982942431e-06,
+      "loss": 0.8325,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1782753870376936,
+      "learning_rate": 5.5863539445629e-06,
+      "loss": 1.0406,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.240726913109007,
+      "learning_rate": 5.607675906183369e-06,
+      "loss": 0.9081,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4511322540965335,
+      "learning_rate": 5.628997867803838e-06,
+      "loss": 0.967,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2882391868681238,
+      "learning_rate": 5.650319829424308e-06,
+      "loss": 0.5714,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1512794053936166,
+      "learning_rate": 5.671641791044776e-06,
+      "loss": 0.9801,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0906672040818721,
+      "learning_rate": 5.692963752665246e-06,
+      "loss": 0.665,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1717546746394056,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.0267,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.163513868447426,
+      "learning_rate": 5.735607675906184e-06,
+      "loss": 0.6262,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1178793684682087,
+      "learning_rate": 5.756929637526653e-06,
+      "loss": 0.9093,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.642168249756498,
+      "learning_rate": 5.7782515991471225e-06,
+      "loss": 0.8477,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.334759862627989,
+      "learning_rate": 5.799573560767591e-06,
+      "loss": 1.0337,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.705081480843519,
+      "learning_rate": 5.820895522388061e-06,
+      "loss": 0.7556,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.000365245410886,
+      "learning_rate": 5.842217484008529e-06,
+      "loss": 0.9212,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.643633935970472,
+      "learning_rate": 5.863539445628999e-06,
+      "loss": 0.6677,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0249594713270835,
+      "learning_rate": 5.884861407249467e-06,
+      "loss": 0.9152,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.515549799443228,
+      "learning_rate": 5.906183368869937e-06,
+      "loss": 0.9882,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.421590343087106,
+      "learning_rate": 5.927505330490405e-06,
+      "loss": 1.1193,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.246158357551085,
+      "learning_rate": 5.948827292110875e-06,
+      "loss": 0.6427,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8832143511959765,
+      "learning_rate": 5.970149253731343e-06,
+      "loss": 0.9948,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.181413713586738,
+      "learning_rate": 5.991471215351813e-06,
+      "loss": 0.9186,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2342513178005263,
+      "learning_rate": 6.012793176972282e-06,
+      "loss": 0.6854,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9923251658375976,
+      "learning_rate": 6.034115138592751e-06,
+      "loss": 0.8874,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0383173533526286,
+      "learning_rate": 6.055437100213221e-06,
+      "loss": 0.933,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.283544336358112,
+      "learning_rate": 6.076759061833689e-06,
+      "loss": 1.0598,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6608220494829387,
+      "learning_rate": 6.098081023454159e-06,
+      "loss": 1.0054,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.778793941669744,
+      "learning_rate": 6.119402985074627e-06,
+      "loss": 0.9767,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6238767522669235,
+      "learning_rate": 6.140724946695097e-06,
+      "loss": 0.8739,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9755207194651476,
+      "learning_rate": 6.1620469083155655e-06,
+      "loss": 0.9938,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.444592845744324,
+      "learning_rate": 6.183368869936035e-06,
+      "loss": 0.9419,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1656303570189115,
+      "learning_rate": 6.2046908315565036e-06,
+      "loss": 0.9263,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1413366167972225,
+      "learning_rate": 6.2260127931769735e-06,
+      "loss": 0.8985,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.457238911551268,
+      "learning_rate": 6.247334754797442e-06,
+      "loss": 1.0255,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1023340023667414,
+      "learning_rate": 6.2686567164179116e-06,
+      "loss": 0.9161,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9498021308696403,
+      "learning_rate": 6.28997867803838e-06,
+      "loss": 0.8277,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8937841848687373,
+      "learning_rate": 6.31130063965885e-06,
+      "loss": 0.9618,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.1301494926394255,
+      "learning_rate": 6.332622601279318e-06,
+      "loss": 1.0586,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.131360108536233,
+      "learning_rate": 6.353944562899788e-06,
+      "loss": 0.9674,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1590534066482174,
+      "learning_rate": 6.375266524520256e-06,
+      "loss": 0.9393,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.658927936092919,
+      "learning_rate": 6.396588486140726e-06,
+      "loss": 0.9547,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4457469301509662,
+      "learning_rate": 6.417910447761194e-06,
+      "loss": 0.9788,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.373846348825217,
+      "learning_rate": 6.439232409381664e-06,
+      "loss": 0.8717,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8430566668485107,
+      "learning_rate": 6.460554371002132e-06,
+      "loss": 0.963,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.224909590337068,
+      "learning_rate": 6.481876332622602e-06,
+      "loss": 0.8784,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4392440135734534,
+      "learning_rate": 6.50319829424307e-06,
+      "loss": 0.9985,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.754207258911546,
+      "learning_rate": 6.52452025586354e-06,
+      "loss": 0.9444,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.714374069313099,
+      "learning_rate": 6.545842217484008e-06,
+      "loss": 1.0267,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7436597689993816,
+      "learning_rate": 6.567164179104478e-06,
+      "loss": 0.8546,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9854427359526927,
+      "learning_rate": 6.5884861407249465e-06,
+      "loss": 0.9612,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0293461006595965,
+      "learning_rate": 6.609808102345416e-06,
+      "loss": 0.967,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8414432704902806,
+      "learning_rate": 6.631130063965885e-06,
+      "loss": 0.8078,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.491791324900091,
+      "learning_rate": 6.6524520255863545e-06,
+      "loss": 0.7802,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.898073597591938,
+      "learning_rate": 6.673773987206824e-06,
+      "loss": 0.8716,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1097930594378065,
+      "learning_rate": 6.695095948827293e-06,
+      "loss": 1.0754,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2565417770777256,
+      "learning_rate": 6.7164179104477625e-06,
+      "loss": 0.8565,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0730410116636167,
+      "learning_rate": 6.737739872068231e-06,
+      "loss": 0.8344,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2537649066467376,
+      "learning_rate": 6.759061833688701e-06,
+      "loss": 0.8573,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1468613345945626,
+      "learning_rate": 6.780383795309169e-06,
+      "loss": 0.9858,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8719006587480826,
+      "learning_rate": 6.801705756929639e-06,
+      "loss": 0.9457,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5203292679436713,
+      "learning_rate": 6.823027718550107e-06,
+      "loss": 0.7089,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.306919313484959,
+      "learning_rate": 6.844349680170577e-06,
+      "loss": 0.9751,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6998464872942156,
+      "learning_rate": 6.865671641791045e-06,
+      "loss": 0.7729,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.713264921851124,
+      "learning_rate": 6.886993603411515e-06,
+      "loss": 0.9249,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.246716168901037,
+      "learning_rate": 6.908315565031983e-06,
+      "loss": 0.9277,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.694448657776868,
+      "learning_rate": 6.929637526652453e-06,
+      "loss": 0.8979,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1475989737181354,
+      "learning_rate": 6.950959488272921e-06,
+      "loss": 0.9808,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4638633084012853,
+      "learning_rate": 6.972281449893391e-06,
+      "loss": 0.9269,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9527020475969081,
+      "learning_rate": 6.993603411513859e-06,
+      "loss": 0.8338,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.177329559910655,
+      "learning_rate": 7.014925373134329e-06,
+      "loss": 0.9015,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2012697725637635,
+      "learning_rate": 7.0362473347547975e-06,
+      "loss": 0.9527,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2158066303374846,
+      "learning_rate": 7.057569296375267e-06,
+      "loss": 1.0,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7489716117713345,
+      "learning_rate": 7.0788912579957356e-06,
+      "loss": 1.0718,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.529242445502372,
+      "learning_rate": 7.1002132196162055e-06,
+      "loss": 0.5992,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.416350583541147,
+      "learning_rate": 7.121535181236674e-06,
+      "loss": 0.8991,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3632284610888887,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 1.0134,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5166505228863714,
+      "learning_rate": 7.164179104477612e-06,
+      "loss": 0.9807,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.88242789412352,
+      "learning_rate": 7.185501066098082e-06,
+      "loss": 0.9394,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6630030800494082,
+      "learning_rate": 7.20682302771855e-06,
+      "loss": 1.032,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.055905510442367,
+      "learning_rate": 7.22814498933902e-06,
+      "loss": 0.8744,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4294066431625405,
+      "learning_rate": 7.249466950959488e-06,
+      "loss": 0.8044,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.484050877756941,
+      "learning_rate": 7.270788912579958e-06,
+      "loss": 0.8299,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0965733740756693,
+      "learning_rate": 7.292110874200427e-06,
+      "loss": 0.8786,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9439038375699649,
+      "learning_rate": 7.313432835820896e-06,
+      "loss": 0.9852,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8740406491203134,
+      "learning_rate": 7.334754797441366e-06,
+      "loss": 0.9619,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2423658187716777,
+      "learning_rate": 7.356076759061834e-06,
+      "loss": 0.6575,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8947750021508654,
+      "learning_rate": 7.377398720682304e-06,
+      "loss": 0.9473,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.114614143247767,
+      "learning_rate": 7.398720682302772e-06,
+      "loss": 0.9232,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7273713945277596,
+      "learning_rate": 7.420042643923242e-06,
+      "loss": 0.9335,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9824337036732451,
+      "learning_rate": 7.44136460554371e-06,
+      "loss": 1.0683,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.385824071399859,
+      "learning_rate": 7.46268656716418e-06,
+      "loss": 0.9363,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9384123103532898,
+      "learning_rate": 7.484008528784648e-06,
+      "loss": 0.9134,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.080441862472887,
+      "learning_rate": 7.505330490405118e-06,
+      "loss": 0.8901,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8809464369546955,
+      "learning_rate": 7.5266524520255865e-06,
+      "loss": 0.9653,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9412937865695985,
+      "learning_rate": 7.547974413646056e-06,
+      "loss": 0.8929,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3806806680495556,
+      "learning_rate": 7.569296375266525e-06,
+      "loss": 0.9809,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.963422726216561,
+      "learning_rate": 7.5906183368869945e-06,
+      "loss": 1.0081,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4670321819183574,
+      "learning_rate": 7.611940298507463e-06,
+      "loss": 0.9087,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8030189456000802,
+      "learning_rate": 7.633262260127933e-06,
+      "loss": 0.8917,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6361289684145204,
+      "learning_rate": 7.654584221748402e-06,
+      "loss": 0.8917,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4529334469524673,
+      "learning_rate": 7.67590618336887e-06,
+      "loss": 0.943,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3712916900962648,
+      "learning_rate": 7.69722814498934e-06,
+      "loss": 0.6921,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3924077787798104,
+      "learning_rate": 7.718550106609809e-06,
+      "loss": 0.9229,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3430856613304745,
+      "learning_rate": 7.739872068230278e-06,
+      "loss": 0.8115,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.203238907587797,
+      "learning_rate": 7.761194029850747e-06,
+      "loss": 0.7322,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3155663458865114,
+      "learning_rate": 7.782515991471216e-06,
+      "loss": 0.9082,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.076329001763282,
+      "learning_rate": 7.803837953091685e-06,
+      "loss": 0.9792,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8928206065485618,
+      "learning_rate": 7.825159914712154e-06,
+      "loss": 0.8807,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9586907541314686,
+      "learning_rate": 7.846481876332623e-06,
+      "loss": 0.935,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3956358079564684,
+      "learning_rate": 7.867803837953092e-06,
+      "loss": 0.8532,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.932101233692548,
+      "learning_rate": 7.889125799573561e-06,
+      "loss": 0.9287,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3772001755843513,
+      "learning_rate": 7.91044776119403e-06,
+      "loss": 1.0711,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.908265327910005,
+      "learning_rate": 7.9317697228145e-06,
+      "loss": 0.9474,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2344908128373437,
+      "learning_rate": 7.953091684434968e-06,
+      "loss": 0.9611,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.081609764297957,
+      "learning_rate": 7.974413646055437e-06,
+      "loss": 0.7924,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9411706305001288,
+      "learning_rate": 7.995735607675907e-06,
+      "loss": 0.9872,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8231733060229343,
+      "learning_rate": 8.017057569296376e-06,
+      "loss": 0.9667,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3580744995261118,
+      "learning_rate": 8.038379530916846e-06,
+      "loss": 0.659,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2963418446141253,
+      "learning_rate": 8.059701492537314e-06,
+      "loss": 0.8866,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1356109488096129,
+      "learning_rate": 8.081023454157784e-06,
+      "loss": 0.5701,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1548228647924645,
+      "learning_rate": 8.102345415778252e-06,
+      "loss": 1.0267,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.913340727073849,
+      "learning_rate": 8.123667377398723e-06,
+      "loss": 0.9174,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.03885404514754,
+      "learning_rate": 8.14498933901919e-06,
+      "loss": 1.0168,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.290386642590129,
+      "learning_rate": 8.16631130063966e-06,
+      "loss": 0.9362,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.016375296552861,
+      "learning_rate": 8.187633262260128e-06,
+      "loss": 0.8422,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.956203670052285,
+      "learning_rate": 8.208955223880599e-06,
+      "loss": 0.822,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.823940918771343,
+      "learning_rate": 8.230277185501066e-06,
+      "loss": 0.964,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3781710291984135,
+      "learning_rate": 8.251599147121537e-06,
+      "loss": 0.8841,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.426107667094325,
+      "learning_rate": 8.272921108742004e-06,
+      "loss": 0.8491,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2597321023593822,
+      "learning_rate": 8.294243070362475e-06,
+      "loss": 0.686,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3572461118480486,
+      "learning_rate": 8.315565031982942e-06,
+      "loss": 0.8971,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2851656611708067,
+      "learning_rate": 8.336886993603413e-06,
+      "loss": 0.9515,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.884629093102199,
+      "learning_rate": 8.35820895522388e-06,
+      "loss": 0.8316,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8724613259287082,
+      "learning_rate": 8.379530916844351e-06,
+      "loss": 0.9499,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.130790103317508,
+      "learning_rate": 8.400852878464819e-06,
+      "loss": 0.6321,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.096202158558814,
+      "learning_rate": 8.42217484008529e-06,
+      "loss": 1.0954,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7665257623617618,
+      "learning_rate": 8.443496801705757e-06,
+      "loss": 0.9369,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6169940901787996,
+      "learning_rate": 8.464818763326227e-06,
+      "loss": 0.8419,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9973899985668788,
+      "learning_rate": 8.486140724946695e-06,
+      "loss": 0.9533,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5398054807275865,
+      "learning_rate": 8.507462686567165e-06,
+      "loss": 0.8701,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5827984302662097,
+      "learning_rate": 8.528784648187633e-06,
+      "loss": 1.0647,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3045615216595654,
+      "learning_rate": 8.550106609808104e-06,
+      "loss": 0.8749,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8196703260750848,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.9149,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7460497129556605,
+      "learning_rate": 8.592750533049042e-06,
+      "loss": 1.0076,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.328495184886864,
+      "learning_rate": 8.614072494669509e-06,
+      "loss": 0.9003,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7261407222271106,
+      "learning_rate": 8.63539445628998e-06,
+      "loss": 0.6688,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.308840449332254,
+      "learning_rate": 8.656716417910447e-06,
+      "loss": 0.943,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.058208045865655,
+      "learning_rate": 8.678038379530918e-06,
+      "loss": 0.8305,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8571959262757445,
+      "learning_rate": 8.699360341151387e-06,
+      "loss": 0.7764,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2816899260122598,
+      "learning_rate": 8.720682302771856e-06,
+      "loss": 0.7547,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3918115672523586,
+      "learning_rate": 8.742004264392325e-06,
+      "loss": 0.9395,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1153984958314447,
+      "learning_rate": 8.763326226012794e-06,
+      "loss": 0.8794,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2367720947801217,
+      "learning_rate": 8.784648187633263e-06,
+      "loss": 1.0252,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2536979785000748,
+      "learning_rate": 8.805970149253732e-06,
+      "loss": 0.7189,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1224312048763503,
+      "learning_rate": 8.827292110874201e-06,
+      "loss": 0.9029,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7499133411034,
+      "learning_rate": 8.84861407249467e-06,
+      "loss": 0.8521,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6037588495396562,
+      "learning_rate": 8.86993603411514e-06,
+      "loss": 0.7675,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5842159357719896,
+      "learning_rate": 8.891257995735608e-06,
+      "loss": 0.8717,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4833669406174477,
+      "learning_rate": 8.912579957356077e-06,
+      "loss": 1.0202,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.585251957455893,
+      "learning_rate": 8.933901918976547e-06,
+      "loss": 0.9375,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.020655326727032,
+      "learning_rate": 8.955223880597016e-06,
+      "loss": 0.8163,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.186087970002137,
+      "learning_rate": 8.976545842217485e-06,
+      "loss": 0.8393,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1022116852539003,
+      "learning_rate": 8.997867803837954e-06,
+      "loss": 0.9957,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8212001867124488,
+      "learning_rate": 9.019189765458423e-06,
+      "loss": 0.9432,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9329615591035667,
+      "learning_rate": 9.040511727078892e-06,
+      "loss": 0.8439,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8304628392512616,
+      "learning_rate": 9.06183368869936e-06,
+      "loss": 0.8758,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2889996527970644,
+      "learning_rate": 9.08315565031983e-06,
+      "loss": 0.8186,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8278176967033548,
+      "learning_rate": 9.104477611940299e-06,
+      "loss": 0.7865,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4264292370834912,
+      "learning_rate": 9.125799573560768e-06,
+      "loss": 0.8444,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5873679925598823,
+      "learning_rate": 9.147121535181237e-06,
+      "loss": 0.8493,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.323283848219806,
+      "learning_rate": 9.168443496801706e-06,
+      "loss": 0.9592,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1823887024046074,
+      "learning_rate": 9.189765458422175e-06,
+      "loss": 0.8947,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.105874441660986,
+      "learning_rate": 9.211087420042644e-06,
+      "loss": 0.8909,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0039451544647595,
+      "learning_rate": 9.232409381663113e-06,
+      "loss": 1.0047,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1680006777947503,
+      "learning_rate": 9.253731343283582e-06,
+      "loss": 0.9472,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3038053627168047,
+      "learning_rate": 9.275053304904051e-06,
+      "loss": 0.9701,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9314646659715315,
+      "learning_rate": 9.29637526652452e-06,
+      "loss": 0.9723,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.868488197697785,
+      "learning_rate": 9.31769722814499e-06,
+      "loss": 0.9596,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0496194712816695,
+      "learning_rate": 9.339019189765458e-06,
+      "loss": 0.8799,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9545911440066979,
+      "learning_rate": 9.36034115138593e-06,
+      "loss": 0.9433,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1064664029024605,
+      "learning_rate": 9.381663113006397e-06,
+      "loss": 0.9758,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0883985689407492,
+      "learning_rate": 9.402985074626867e-06,
+      "loss": 0.6514,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.135616775468958,
+      "learning_rate": 9.424307036247335e-06,
+      "loss": 0.9222,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8955704087606677,
+      "learning_rate": 9.445628997867805e-06,
+      "loss": 0.9517,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9521089438391648,
+      "learning_rate": 9.466950959488273e-06,
+      "loss": 0.9691,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7366744966370586,
+      "learning_rate": 9.488272921108744e-06,
+      "loss": 0.7737,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6596408803791167,
+      "learning_rate": 9.509594882729211e-06,
+      "loss": 1.0342,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8400224554694296,
+      "learning_rate": 9.530916844349682e-06,
+      "loss": 0.9194,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9930960797198964,
+      "learning_rate": 9.552238805970149e-06,
+      "loss": 0.8442,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1393097342317517,
+      "learning_rate": 9.57356076759062e-06,
+      "loss": 0.9537,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.345037058324702,
+      "learning_rate": 9.594882729211089e-06,
+      "loss": 0.9195,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.789620211403352,
+      "learning_rate": 9.616204690831558e-06,
+      "loss": 0.8295,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1812704671705614,
+      "learning_rate": 9.637526652452027e-06,
+      "loss": 0.8213,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1000685395984244,
+      "learning_rate": 9.658848614072496e-06,
+      "loss": 0.915,
+      "step": 453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.122924205500287,
+      "learning_rate": 9.680170575692965e-06,
+      "loss": 0.8357,
+      "step": 454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8491909244193483,
+      "learning_rate": 9.701492537313434e-06,
+      "loss": 0.8141,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4503108927333184,
+      "learning_rate": 9.722814498933903e-06,
+      "loss": 0.8418,
+      "step": 456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.578364475815341,
+      "learning_rate": 9.744136460554372e-06,
+      "loss": 0.8379,
+      "step": 457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.876222959491142,
+      "learning_rate": 9.765458422174841e-06,
+      "loss": 0.8293,
+      "step": 458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9854168941340664,
+      "learning_rate": 9.78678038379531e-06,
+      "loss": 0.8828,
+      "step": 459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2418193370281274,
+      "learning_rate": 9.80810234541578e-06,
+      "loss": 0.5313,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7006877472369994,
+      "learning_rate": 9.829424307036248e-06,
+      "loss": 0.7956,
+      "step": 461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5340291603571536,
+      "learning_rate": 9.850746268656717e-06,
+      "loss": 0.9739,
+      "step": 462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0323531737641765,
+      "learning_rate": 9.872068230277187e-06,
+      "loss": 0.8223,
+      "step": 463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6875957343673105,
+      "learning_rate": 9.893390191897656e-06,
+      "loss": 0.9563,
+      "step": 464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7233727000747334,
+      "learning_rate": 9.914712153518125e-06,
+      "loss": 0.9839,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9840414369958783,
+      "learning_rate": 9.936034115138594e-06,
+      "loss": 0.9088,
+      "step": 466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9474909918305257,
+      "learning_rate": 9.957356076759063e-06,
+      "loss": 0.9589,
+      "step": 467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7316374245114288,
+      "learning_rate": 9.978678038379532e-06,
+      "loss": 0.866,
+      "step": 468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7791288864569303,
+      "learning_rate": 1e-05,
+      "loss": 0.876,
+      "step": 469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.029855751908816,
+      "learning_rate": 9.999999892569433e-06,
+      "loss": 0.8735,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.071058743064303,
+      "learning_rate": 9.999999570277734e-06,
+      "loss": 0.8676,
+      "step": 471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6990051817975669,
+      "learning_rate": 9.999999033124919e-06,
+      "loss": 0.8824,
+      "step": 472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0738320876907603,
+      "learning_rate": 9.999998281111008e-06,
+      "loss": 0.9662,
+      "step": 473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0561438791209965,
+      "learning_rate": 9.999997314236036e-06,
+      "loss": 0.863,
+      "step": 474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.043342086805317,
+      "learning_rate": 9.999996132500046e-06,
+      "loss": 0.8369,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9549733600665204,
+      "learning_rate": 9.999994735903083e-06,
+      "loss": 0.8157,
+      "step": 476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7765269329081328,
+      "learning_rate": 9.999993124445213e-06,
+      "loss": 0.9937,
+      "step": 477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2833653345750868,
+      "learning_rate": 9.999991298126501e-06,
+      "loss": 0.702,
+      "step": 478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.77411841631966,
+      "learning_rate": 9.999989256947029e-06,
+      "loss": 0.8517,
+      "step": 479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9840782944891566,
+      "learning_rate": 9.999987000906883e-06,
+      "loss": 0.8593,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3824012154595624,
+      "learning_rate": 9.99998453000616e-06,
+      "loss": 0.8934,
+      "step": 481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.37014554788389,
+      "learning_rate": 9.999981844244966e-06,
+      "loss": 1.0631,
+      "step": 482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0150055297976683,
+      "learning_rate": 9.999978943623417e-06,
+      "loss": 0.8053,
+      "step": 483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4901782967272756,
+      "learning_rate": 9.999975828141635e-06,
+      "loss": 0.6791,
+      "step": 484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.741397851786972,
+      "learning_rate": 9.999972497799758e-06,
+      "loss": 0.9771,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.848232326959614,
+      "learning_rate": 9.999968952597926e-06,
+      "loss": 0.9574,
+      "step": 486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.52009546405784,
+      "learning_rate": 9.999965192536295e-06,
+      "loss": 0.8398,
+      "step": 487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.343144065424678,
+      "learning_rate": 9.999961217615022e-06,
+      "loss": 0.6069,
+      "step": 488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1279962355032773,
+      "learning_rate": 9.999957027834282e-06,
+      "loss": 0.9262,
+      "step": 489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.707456681593266,
+      "learning_rate": 9.999952623194252e-06,
+      "loss": 0.8303,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.442154658677076,
+      "learning_rate": 9.99994800369512e-06,
+      "loss": 0.9358,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9453214053479437,
+      "learning_rate": 9.99994316933709e-06,
+      "loss": 0.9086,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3422988213663545,
+      "learning_rate": 9.999938120120364e-06,
+      "loss": 1.0063,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0080631972992244,
+      "learning_rate": 9.999932856045164e-06,
+      "loss": 0.9208,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8878447267785827,
+      "learning_rate": 9.999927377111712e-06,
+      "loss": 0.8724,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8638910330209149,
+      "learning_rate": 9.999921683320247e-06,
+      "loss": 0.9006,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2260230168261748,
+      "learning_rate": 9.999915774671009e-06,
+      "loss": 0.8548,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4385542126141337,
+      "learning_rate": 9.999909651164258e-06,
+      "loss": 0.6194,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0431519100584334,
+      "learning_rate": 9.99990331280025e-06,
+      "loss": 0.8919,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8941679114981533,
+      "learning_rate": 9.999896759579263e-06,
+      "loss": 0.8943,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4061009891458889,
+      "learning_rate": 9.99988999150158e-06,
+      "loss": 0.6418,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0908694108268704,
+      "learning_rate": 9.999883008567487e-06,
+      "loss": 1.0027,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5423295715511747,
+      "learning_rate": 9.999875810777283e-06,
+      "loss": 0.6073,
+      "step": 503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2089224077860423,
+      "learning_rate": 9.999868398131282e-06,
+      "loss": 0.6343,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.213146032017295,
+      "learning_rate": 9.9998607706298e-06,
+      "loss": 0.9032,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7825763153692014,
+      "learning_rate": 9.999852928273166e-06,
+      "loss": 0.8251,
+      "step": 506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2197110164161558,
+      "learning_rate": 9.999844871061718e-06,
+      "loss": 0.7982,
+      "step": 507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6984414725607353,
+      "learning_rate": 9.999836598995799e-06,
+      "loss": 0.9031,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8674548458492914,
+      "learning_rate": 9.999828112075764e-06,
+      "loss": 0.8594,
+      "step": 509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3312285193343945,
+      "learning_rate": 9.999819410301983e-06,
+      "loss": 0.8717,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2062873508788248,
+      "learning_rate": 9.999810493674826e-06,
+      "loss": 0.6732,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.69399001036484,
+      "learning_rate": 9.999801362194676e-06,
+      "loss": 0.9543,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7262548595241949,
+      "learning_rate": 9.999792015861927e-06,
+      "loss": 0.6549,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.14188650280057,
+      "learning_rate": 9.99978245467698e-06,
+      "loss": 0.7797,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.950916236448234,
+      "learning_rate": 9.999772678640245e-06,
+      "loss": 0.8129,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8505380239228624,
+      "learning_rate": 9.999762687752143e-06,
+      "loss": 0.9822,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7671919190752186,
+      "learning_rate": 9.999752482013105e-06,
+      "loss": 0.983,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8418615514429961,
+      "learning_rate": 9.999742061423567e-06,
+      "loss": 0.8524,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9233330873615297,
+      "learning_rate": 9.999731425983975e-06,
+      "loss": 0.7537,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8252935240759913,
+      "learning_rate": 9.999720575694792e-06,
+      "loss": 0.9436,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.052989272592887,
+      "learning_rate": 9.99970951055648e-06,
+      "loss": 0.9029,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4864384010178155,
+      "learning_rate": 9.999698230569515e-06,
+      "loss": 1.0153,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9046650531049187,
+      "learning_rate": 9.999686735734382e-06,
+      "loss": 0.8429,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.986024007336273,
+      "learning_rate": 9.999675026051576e-06,
+      "loss": 0.9225,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.304289981280268,
+      "learning_rate": 9.999663101521599e-06,
+      "loss": 0.8242,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.525269835961098,
+      "learning_rate": 9.999650962144962e-06,
+      "loss": 0.9032,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8846591894242477,
+      "learning_rate": 9.999638607922191e-06,
+      "loss": 0.8749,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7900280134830837,
+      "learning_rate": 9.99962603885381e-06,
+      "loss": 0.8265,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8204857293663168,
+      "learning_rate": 9.999613254940368e-06,
+      "loss": 0.8658,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2287210794591874,
+      "learning_rate": 9.999600256182408e-06,
+      "loss": 0.8962,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.982683164227249,
+      "learning_rate": 9.99958704258049e-06,
+      "loss": 0.7601,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3565944818273077,
+      "learning_rate": 9.999573614135183e-06,
+      "loss": 1.0026,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.888036810147538,
+      "learning_rate": 9.999559970847061e-06,
+      "loss": 0.9216,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.363453851143911,
+      "learning_rate": 9.999546112716715e-06,
+      "loss": 0.8999,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8565098937542688,
+      "learning_rate": 9.999532039744736e-06,
+      "loss": 0.9023,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1639546486703543,
+      "learning_rate": 9.999517751931731e-06,
+      "loss": 0.7927,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2346803888633255,
+      "learning_rate": 9.999503249278315e-06,
+      "loss": 0.91,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9232771490358687,
+      "learning_rate": 9.99948853178511e-06,
+      "loss": 0.8863,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.033101481317618,
+      "learning_rate": 9.999473599452746e-06,
+      "loss": 0.841,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9245369576916302,
+      "learning_rate": 9.999458452281868e-06,
+      "loss": 0.9519,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9500333113282542,
+      "learning_rate": 9.999443090273126e-06,
+      "loss": 0.8506,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9759540321075504,
+      "learning_rate": 9.99942751342718e-06,
+      "loss": 0.8084,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.072732015613899,
+      "learning_rate": 9.9994117217447e-06,
+      "loss": 0.8253,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7824547933034987,
+      "learning_rate": 9.999395715226365e-06,
+      "loss": 0.8468,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3413267209833,
+      "learning_rate": 9.99937949387286e-06,
+      "loss": 0.9591,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4898832771590653,
+      "learning_rate": 9.999363057684885e-06,
+      "loss": 0.8852,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9330236633008158,
+      "learning_rate": 9.999346406663144e-06,
+      "loss": 1.0399,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8964165895591503,
+      "learning_rate": 9.999329540808355e-06,
+      "loss": 0.8581,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2403908477109553,
+      "learning_rate": 9.999312460121242e-06,
+      "loss": 0.822,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9565854227432997,
+      "learning_rate": 9.999295164602536e-06,
+      "loss": 0.8357,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1872248961214273,
+      "learning_rate": 9.999277654252985e-06,
+      "loss": 0.751,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7906398165513877,
+      "learning_rate": 9.999259929073339e-06,
+      "loss": 0.9465,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7832388451141967,
+      "learning_rate": 9.999241989064358e-06,
+      "loss": 0.8417,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.41188370740574,
+      "learning_rate": 9.999223834226817e-06,
+      "loss": 0.5647,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.007560919727563,
+      "learning_rate": 9.999205464561494e-06,
+      "loss": 0.9757,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.327219483103542,
+      "learning_rate": 9.999186880069178e-06,
+      "loss": 0.8852,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6461282846395042,
+      "learning_rate": 9.999168080750666e-06,
+      "loss": 0.8949,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8245903043662073,
+      "learning_rate": 9.99914906660677e-06,
+      "loss": 1.0257,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9246113780867906,
+      "learning_rate": 9.999129837638303e-06,
+      "loss": 0.9393,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2953115501654415,
+      "learning_rate": 9.999110393846097e-06,
+      "loss": 0.6626,
+      "step": 560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8235244828439587,
+      "learning_rate": 9.999090735230981e-06,
+      "loss": 0.9364,
+      "step": 561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7935958180803704,
+      "learning_rate": 9.999070861793803e-06,
+      "loss": 0.8619,
+      "step": 562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4756626360599734,
+      "learning_rate": 9.999050773535414e-06,
+      "loss": 0.9189,
+      "step": 563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.099828728996209,
+      "learning_rate": 9.999030470456684e-06,
+      "loss": 0.9238,
+      "step": 564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8559166873425144,
+      "learning_rate": 9.999009952558478e-06,
+      "loss": 0.8968,
+      "step": 565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.547786765043327,
+      "learning_rate": 9.998989219841684e-06,
+      "loss": 0.9932,
+      "step": 566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2146108943274436,
+      "learning_rate": 9.998968272307187e-06,
+      "loss": 0.6464,
+      "step": 567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2342929524977806,
+      "learning_rate": 9.998947109955892e-06,
+      "loss": 0.9451,
+      "step": 568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9077450595731333,
+      "learning_rate": 9.998925732788706e-06,
+      "loss": 0.7606,
+      "step": 569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7375709285340808,
+      "learning_rate": 9.998904140806547e-06,
+      "loss": 0.8368,
+      "step": 570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1650487226633284,
+      "learning_rate": 9.998882334010345e-06,
+      "loss": 0.8281,
+      "step": 571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1985800637551245,
+      "learning_rate": 9.998860312401035e-06,
+      "loss": 0.8986,
+      "step": 572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.885660611202107,
+      "learning_rate": 9.998838075979566e-06,
+      "loss": 0.8837,
+      "step": 573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7380038434146023,
+      "learning_rate": 9.99881562474689e-06,
+      "loss": 0.8046,
+      "step": 574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8341187383959814,
+      "learning_rate": 9.998792958703975e-06,
+      "loss": 0.959,
+      "step": 575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.194012090239682,
+      "learning_rate": 9.998770077851794e-06,
+      "loss": 0.8476,
+      "step": 576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.257838239656256,
+      "learning_rate": 9.998746982191327e-06,
+      "loss": 0.7439,
+      "step": 577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.074242577767752,
+      "learning_rate": 9.998723671723572e-06,
+      "loss": 0.826,
+      "step": 578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4689499450959003,
+      "learning_rate": 9.998700146449528e-06,
+      "loss": 0.7845,
+      "step": 579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4687789767266457,
+      "learning_rate": 9.998676406370205e-06,
+      "loss": 0.9771,
+      "step": 580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8888461357858384,
+      "learning_rate": 9.998652451486626e-06,
+      "loss": 0.949,
+      "step": 581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.986812209454056,
+      "learning_rate": 9.998628281799816e-06,
+      "loss": 0.9351,
+      "step": 582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7857802850306248,
+      "learning_rate": 9.998603897310818e-06,
+      "loss": 0.931,
+      "step": 583
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5049540405477293,
+      "learning_rate": 9.998579298020676e-06,
+      "loss": 0.6137,
+      "step": 584
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.877074292766484,
+      "learning_rate": 9.998554483930452e-06,
+      "loss": 0.8663,
+      "step": 585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.968457834159462,
+      "learning_rate": 9.998529455041209e-06,
+      "loss": 0.9152,
+      "step": 586
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.561939183176745,
+      "learning_rate": 9.998504211354021e-06,
+      "loss": 0.9191,
+      "step": 587
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7766753327992717,
+      "learning_rate": 9.998478752869976e-06,
+      "loss": 0.9193,
+      "step": 588
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.578134500746126,
+      "learning_rate": 9.998453079590167e-06,
+      "loss": 0.9145,
+      "step": 589
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9494864525103663,
+      "learning_rate": 9.998427191515696e-06,
+      "loss": 0.8579,
+      "step": 590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1572537292308565,
+      "learning_rate": 9.998401088647677e-06,
+      "loss": 0.9748,
+      "step": 591
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8426233428905723,
+      "learning_rate": 9.998374770987232e-06,
+      "loss": 0.9371,
+      "step": 592
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.127911435919586,
+      "learning_rate": 9.99834823853549e-06,
+      "loss": 0.8155,
+      "step": 593
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8332401447954598,
+      "learning_rate": 9.998321491293592e-06,
+      "loss": 0.8749,
+      "step": 594
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.321843127766646,
+      "learning_rate": 9.998294529262688e-06,
+      "loss": 0.8771,
+      "step": 595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.049681357371902,
+      "learning_rate": 9.998267352443938e-06,
+      "loss": 0.7673,
+      "step": 596
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7668113727039019,
+      "learning_rate": 9.998239960838505e-06,
+      "loss": 0.8768,
+      "step": 597
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8563411616767895,
+      "learning_rate": 9.998212354447572e-06,
+      "loss": 0.8269,
+      "step": 598
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3387469366220799,
+      "learning_rate": 9.998184533272321e-06,
+      "loss": 0.7383,
+      "step": 599
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9171698320323378,
+      "learning_rate": 9.99815649731395e-06,
+      "loss": 0.848,
+      "step": 600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.351650211651385,
+      "learning_rate": 9.998128246573663e-06,
+      "loss": 0.8057,
+      "step": 601
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8236541121321594,
+      "learning_rate": 9.998099781052673e-06,
+      "loss": 0.895,
+      "step": 602
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8441627410288801,
+      "learning_rate": 9.998071100752203e-06,
+      "loss": 0.9094,
+      "step": 603
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8603325748291044,
+      "learning_rate": 9.998042205673489e-06,
+      "loss": 0.9504,
+      "step": 604
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1644515315533623,
+      "learning_rate": 9.998013095817768e-06,
+      "loss": 0.9064,
+      "step": 605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.181413257245636,
+      "learning_rate": 9.997983771186293e-06,
+      "loss": 0.9028,
+      "step": 606
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.20516720521565,
+      "learning_rate": 9.997954231780326e-06,
+      "loss": 0.67,
+      "step": 607
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9152085312143734,
+      "learning_rate": 9.997924477601134e-06,
+      "loss": 0.7965,
+      "step": 608
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9555148689522672,
+      "learning_rate": 9.997894508649995e-06,
+      "loss": 0.95,
+      "step": 609
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7994454604327526,
+      "learning_rate": 9.9978643249282e-06,
+      "loss": 0.8093,
+      "step": 610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3773473673931322,
+      "learning_rate": 9.997833926437043e-06,
+      "loss": 0.9482,
+      "step": 611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1625144675443124,
+      "learning_rate": 9.997803313177831e-06,
+      "loss": 0.8604,
+      "step": 612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4141126102539547,
+      "learning_rate": 9.997772485151881e-06,
+      "loss": 0.8924,
+      "step": 613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.130059627992034,
+      "learning_rate": 9.997741442360515e-06,
+      "loss": 0.8992,
+      "step": 614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2546404443418784,
+      "learning_rate": 9.997710184805072e-06,
+      "loss": 0.6699,
+      "step": 615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5415738473359955,
+      "learning_rate": 9.997678712486889e-06,
+      "loss": 0.6879,
+      "step": 616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.960774068408883,
+      "learning_rate": 9.997647025407322e-06,
+      "loss": 0.9305,
+      "step": 617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0102399707780143,
+      "learning_rate": 9.997615123567733e-06,
+      "loss": 0.8982,
+      "step": 618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6341651652021714,
+      "learning_rate": 9.99758300696949e-06,
+      "loss": 0.837,
+      "step": 619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2753492654079968,
+      "learning_rate": 9.997550675613976e-06,
+      "loss": 0.8237,
+      "step": 620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8454514662257493,
+      "learning_rate": 9.997518129502579e-06,
+      "loss": 0.9382,
+      "step": 621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.890501043842892,
+      "learning_rate": 9.997485368636698e-06,
+      "loss": 0.881,
+      "step": 622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0907150442191047,
+      "learning_rate": 9.99745239301774e-06,
+      "loss": 0.916,
+      "step": 623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1569055191141917,
+      "learning_rate": 9.997419202647124e-06,
+      "loss": 0.9162,
+      "step": 624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8384604388325947,
+      "learning_rate": 9.997385797526275e-06,
+      "loss": 0.7748,
+      "step": 625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2748136197903575,
+      "learning_rate": 9.997352177656627e-06,
+      "loss": 0.6534,
+      "step": 626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.947616002043085,
+      "learning_rate": 9.997318343039629e-06,
+      "loss": 0.883,
+      "step": 627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9773218240071553,
+      "learning_rate": 9.99728429367673e-06,
+      "loss": 0.7874,
+      "step": 628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9755231260382262,
+      "learning_rate": 9.997250029569395e-06,
+      "loss": 0.9731,
+      "step": 629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7657876458813815,
+      "learning_rate": 9.997215550719097e-06,
+      "loss": 0.7768,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.822859043489383,
+      "learning_rate": 9.997180857127318e-06,
+      "loss": 0.8128,
+      "step": 631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2883387270719906,
+      "learning_rate": 9.997145948795548e-06,
+      "loss": 0.9035,
+      "step": 632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.314629432459234,
+      "learning_rate": 9.997110825725288e-06,
+      "loss": 0.8763,
+      "step": 633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.019097883770262,
+      "learning_rate": 9.997075487918047e-06,
+      "loss": 0.8826,
+      "step": 634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.027388361654231,
+      "learning_rate": 9.99703993537534e-06,
+      "loss": 0.8425,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7797241615637456,
+      "learning_rate": 9.997004168098703e-06,
+      "loss": 0.8396,
+      "step": 636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6331647266746894,
+      "learning_rate": 9.996968186089664e-06,
+      "loss": 0.8295,
+      "step": 637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8254478240251375,
+      "learning_rate": 9.996931989349775e-06,
+      "loss": 0.8464,
+      "step": 638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8391528165631335,
+      "learning_rate": 9.99689557788059e-06,
+      "loss": 0.9786,
+      "step": 639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3100755742706793,
+      "learning_rate": 9.996858951683672e-06,
+      "loss": 0.6116,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.08902848909447,
+      "learning_rate": 9.996822110760595e-06,
+      "loss": 0.9272,
+      "step": 641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4990328542812967,
+      "learning_rate": 9.996785055112947e-06,
+      "loss": 0.6897,
+      "step": 642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8044390321850872,
+      "learning_rate": 9.996747784742314e-06,
+      "loss": 0.824,
+      "step": 643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4023820412051038,
+      "learning_rate": 9.996710299650302e-06,
+      "loss": 0.774,
+      "step": 644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.112066471302443,
+      "learning_rate": 9.99667259983852e-06,
+      "loss": 0.6152,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0711558538299277,
+      "learning_rate": 9.996634685308586e-06,
+      "loss": 0.8827,
+      "step": 646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0213173789199357,
+      "learning_rate": 9.996596556062133e-06,
+      "loss": 0.8735,
+      "step": 647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3893607800542238,
+      "learning_rate": 9.9965582121008e-06,
+      "loss": 0.693,
+      "step": 648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1759005859514464,
+      "learning_rate": 9.996519653426229e-06,
+      "loss": 0.9742,
+      "step": 649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.930926576667769,
+      "learning_rate": 9.996480880040083e-06,
+      "loss": 0.8795,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6817038569740266,
+      "learning_rate": 9.996441891944023e-06,
+      "loss": 0.9429,
+      "step": 651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1051315325308155,
+      "learning_rate": 9.99640268913973e-06,
+      "loss": 0.8681,
+      "step": 652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.001594932206345,
+      "learning_rate": 9.996363271628886e-06,
+      "loss": 0.8454,
+      "step": 653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8822622227166153,
+      "learning_rate": 9.996323639413185e-06,
+      "loss": 0.9193,
+      "step": 654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0471994857557174,
+      "learning_rate": 9.996283792494328e-06,
+      "loss": 0.9532,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.829948124445125,
+      "learning_rate": 9.99624373087403e-06,
+      "loss": 0.9247,
+      "step": 656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1794593129418312,
+      "learning_rate": 9.996203454554011e-06,
+      "loss": 0.5815,
+      "step": 657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8217623719402236,
+      "learning_rate": 9.996162963536004e-06,
+      "loss": 0.8971,
+      "step": 658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.023500021667296,
+      "learning_rate": 9.996122257821746e-06,
+      "loss": 0.8815,
+      "step": 659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9139237521501618,
+      "learning_rate": 9.996081337412988e-06,
+      "loss": 1.0196,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8257965197928756,
+      "learning_rate": 9.996040202311488e-06,
+      "loss": 0.944,
+      "step": 661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9436293781407097,
+      "learning_rate": 9.995998852519014e-06,
+      "loss": 0.888,
+      "step": 662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7594098641506641,
+      "learning_rate": 9.995957288037343e-06,
+      "loss": 0.7567,
+      "step": 663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7034982379510892,
+      "learning_rate": 9.99591550886826e-06,
+      "loss": 0.8184,
+      "step": 664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.583720658918584,
+      "learning_rate": 9.995873515013562e-06,
+      "loss": 0.7187,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6376002116192396,
+      "learning_rate": 9.99583130647505e-06,
+      "loss": 0.7253,
+      "step": 666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8824572493998297,
+      "learning_rate": 9.995788883254543e-06,
+      "loss": 1.0456,
+      "step": 667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.865682694022239,
+      "learning_rate": 9.995746245353861e-06,
+      "loss": 0.9119,
+      "step": 668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7766826315532833,
+      "learning_rate": 9.995703392774836e-06,
+      "loss": 0.8369,
+      "step": 669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2266456611018084,
+      "learning_rate": 9.99566032551931e-06,
+      "loss": 0.9575,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8517701076190214,
+      "learning_rate": 9.995617043589135e-06,
+      "loss": 0.9326,
+      "step": 671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.885094241360763,
+      "learning_rate": 9.99557354698617e-06,
+      "loss": 0.7931,
+      "step": 672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8541109221791268,
+      "learning_rate": 9.995529835712283e-06,
+      "loss": 0.8807,
+      "step": 673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8406529839981007,
+      "learning_rate": 9.995485909769354e-06,
+      "loss": 0.9639,
+      "step": 674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1516794682593825,
+      "learning_rate": 9.995441769159269e-06,
+      "loss": 0.9289,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8386512472114966,
+      "learning_rate": 9.995397413883928e-06,
+      "loss": 0.9652,
+      "step": 676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9223065965178059,
+      "learning_rate": 9.995352843945232e-06,
+      "loss": 0.7682,
+      "step": 677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.137061966849564,
+      "learning_rate": 9.9953080593451e-06,
+      "loss": 0.8793,
+      "step": 678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6591024777698418,
+      "learning_rate": 9.995263060085456e-06,
+      "loss": 0.7142,
+      "step": 679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2174111052022236,
+      "learning_rate": 9.995217846168232e-06,
+      "loss": 0.8921,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2168807693958246,
+      "learning_rate": 9.995172417595373e-06,
+      "loss": 0.7639,
+      "step": 681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.553876794644122,
+      "learning_rate": 9.995126774368831e-06,
+      "loss": 0.6383,
+      "step": 682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3817032891959682,
+      "learning_rate": 9.995080916490566e-06,
+      "loss": 0.6384,
+      "step": 683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8029306103940768,
+      "learning_rate": 9.99503484396255e-06,
+      "loss": 0.815,
+      "step": 684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4964443712157651,
+      "learning_rate": 9.994988556786763e-06,
+      "loss": 0.6961,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0535280263312106,
+      "learning_rate": 9.99494205496519e-06,
+      "loss": 0.858,
+      "step": 686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.881945164603295,
+      "learning_rate": 9.994895338499835e-06,
+      "loss": 0.8926,
+      "step": 687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7456424610887584,
+      "learning_rate": 9.994848407392703e-06,
+      "loss": 0.9713,
+      "step": 688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1570490750698357,
+      "learning_rate": 9.99480126164581e-06,
+      "loss": 0.7339,
+      "step": 689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6986841369875414,
+      "learning_rate": 9.994753901261183e-06,
+      "loss": 0.8632,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1713813957479187,
+      "learning_rate": 9.994706326240858e-06,
+      "loss": 1.0248,
+      "step": 691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4633606182328527,
+      "learning_rate": 9.994658536586878e-06,
+      "loss": 0.8071,
+      "step": 692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.012494854262561,
+      "learning_rate": 9.994610532301296e-06,
+      "loss": 0.8079,
+      "step": 693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6771593450889113,
+      "learning_rate": 9.994562313386177e-06,
+      "loss": 0.7721,
+      "step": 694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5380237532935117,
+      "learning_rate": 9.99451387984359e-06,
+      "loss": 0.712,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1824695114524535,
+      "learning_rate": 9.99446523167562e-06,
+      "loss": 0.9398,
+      "step": 696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0468938291846492,
+      "learning_rate": 9.994416368884357e-06,
+      "loss": 0.8826,
+      "step": 697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9460627461299531,
+      "learning_rate": 9.994367291471896e-06,
+      "loss": 0.9574,
+      "step": 698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.023541918362746,
+      "learning_rate": 9.994317999440351e-06,
+      "loss": 0.8529,
+      "step": 699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1771766148288494,
+      "learning_rate": 9.99426849279184e-06,
+      "loss": 0.8712,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6950611275772278,
+      "learning_rate": 9.994218771528487e-06,
+      "loss": 0.8066,
+      "step": 701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9672110633565665,
+      "learning_rate": 9.994168835652432e-06,
+      "loss": 0.9571,
+      "step": 702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.60480711530087,
+      "learning_rate": 9.994118685165819e-06,
+      "loss": 0.8818,
+      "step": 703
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2165263995614464,
+      "learning_rate": 9.994068320070805e-06,
+      "loss": 0.675,
+      "step": 704
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1574175825254227,
+      "learning_rate": 9.99401774036955e-06,
+      "loss": 0.9199,
+      "step": 705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.789265059796381,
+      "learning_rate": 9.993966946064233e-06,
+      "loss": 0.9906,
+      "step": 706
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0020925238936753,
+      "learning_rate": 9.993915937157033e-06,
+      "loss": 1.0006,
+      "step": 707
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6877440034390894,
+      "learning_rate": 9.993864713650143e-06,
+      "loss": 0.9133,
+      "step": 708
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8436661861668997,
+      "learning_rate": 9.993813275545764e-06,
+      "loss": 0.837,
+      "step": 709
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.105350370539767,
+      "learning_rate": 9.993761622846107e-06,
+      "loss": 0.9377,
+      "step": 710
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.597758614184637,
+      "learning_rate": 9.99370975555339e-06,
+      "loss": 0.8345,
+      "step": 711
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.925718742431276,
+      "learning_rate": 9.993657673669848e-06,
+      "loss": 0.8309,
+      "step": 712
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.631485705028276,
+      "learning_rate": 9.993605377197711e-06,
+      "loss": 0.7575,
+      "step": 713
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.145666363132522,
+      "learning_rate": 9.99355286613923e-06,
+      "loss": 0.8931,
+      "step": 714
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.08272205188069,
+      "learning_rate": 9.993500140496659e-06,
+      "loss": 0.8159,
+      "step": 715
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3388589972475007,
+      "learning_rate": 9.993447200272269e-06,
+      "loss": 0.8466,
+      "step": 716
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.984471115217179,
+      "learning_rate": 9.993394045468329e-06,
+      "loss": 0.883,
+      "step": 717
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2856197226856483,
+      "learning_rate": 9.993340676087129e-06,
+      "loss": 0.9034,
+      "step": 718
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2718077232629112,
+      "learning_rate": 9.993287092130956e-06,
+      "loss": 0.7161,
+      "step": 719
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9664865906408717,
+      "learning_rate": 9.993233293602119e-06,
+      "loss": 0.8985,
+      "step": 720
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1421178414860333,
+      "learning_rate": 9.993179280502926e-06,
+      "loss": 0.5991,
+      "step": 721
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.15584798801153,
+      "learning_rate": 9.993125052835699e-06,
+      "loss": 0.8461,
+      "step": 722
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1700660511092162,
+      "learning_rate": 9.993070610602766e-06,
+      "loss": 0.9321,
+      "step": 723
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.324199508297681,
+      "learning_rate": 9.993015953806472e-06,
+      "loss": 0.8786,
+      "step": 724
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8904390193689664,
+      "learning_rate": 9.992961082449161e-06,
+      "loss": 0.8173,
+      "step": 725
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.085408306361865,
+      "learning_rate": 9.992905996533193e-06,
+      "loss": 0.8727,
+      "step": 726
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2743168014910276,
+      "learning_rate": 9.992850696060933e-06,
+      "loss": 0.8356,
+      "step": 727
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6964762594232734,
+      "learning_rate": 9.99279518103476e-06,
+      "loss": 0.8096,
+      "step": 728
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.837791692758949,
+      "learning_rate": 9.992739451457058e-06,
+      "loss": 0.7895,
+      "step": 729
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8203526377265355,
+      "learning_rate": 9.992683507330223e-06,
+      "loss": 0.9401,
+      "step": 730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3054727110611006,
+      "learning_rate": 9.992627348656658e-06,
+      "loss": 0.9154,
+      "step": 731
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.355277755590754,
+      "learning_rate": 9.992570975438776e-06,
+      "loss": 0.7213,
+      "step": 732
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6743624380637385,
+      "learning_rate": 9.992514387679e-06,
+      "loss": 0.8561,
+      "step": 733
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7939254361834833,
+      "learning_rate": 9.992457585379764e-06,
+      "loss": 0.8979,
+      "step": 734
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9277691988603773,
+      "learning_rate": 9.992400568543506e-06,
+      "loss": 0.9633,
+      "step": 735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.092680302418741,
+      "learning_rate": 9.992343337172674e-06,
+      "loss": 0.8502,
+      "step": 736
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7823188758133968,
+      "learning_rate": 9.992285891269732e-06,
+      "loss": 0.8893,
+      "step": 737
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.05738972385811,
+      "learning_rate": 9.992228230837148e-06,
+      "loss": 0.8843,
+      "step": 738
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7781939577186576,
+      "learning_rate": 9.992170355877398e-06,
+      "loss": 0.8492,
+      "step": 739
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6656002756858133,
+      "learning_rate": 9.99211226639297e-06,
+      "loss": 0.9012,
+      "step": 740
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3549878939617237,
+      "learning_rate": 9.99205396238636e-06,
+      "loss": 0.4992,
+      "step": 741
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.180188871371648,
+      "learning_rate": 9.991995443860074e-06,
+      "loss": 0.8871,
+      "step": 742
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.38835471118772,
+      "learning_rate": 9.991936710816624e-06,
+      "loss": 0.784,
+      "step": 743
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.39409986410357,
+      "learning_rate": 9.991877763258538e-06,
+      "loss": 0.7039,
+      "step": 744
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8979241847093642,
+      "learning_rate": 9.991818601188345e-06,
+      "loss": 0.8227,
+      "step": 745
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9073500268002206,
+      "learning_rate": 9.99175922460859e-06,
+      "loss": 0.8119,
+      "step": 746
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2295857516190574,
+      "learning_rate": 9.991699633521823e-06,
+      "loss": 0.938,
+      "step": 747
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.004872288984494,
+      "learning_rate": 9.991639827930607e-06,
+      "loss": 0.8889,
+      "step": 748
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3288509794184082,
+      "learning_rate": 9.991579807837511e-06,
+      "loss": 0.8815,
+      "step": 749
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2650300506523804,
+      "learning_rate": 9.991519573245112e-06,
+      "loss": 0.7927,
+      "step": 750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.633285909829132,
+      "learning_rate": 9.991459124156003e-06,
+      "loss": 0.8001,
+      "step": 751
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3154470619695366,
+      "learning_rate": 9.991398460572776e-06,
+      "loss": 0.7699,
+      "step": 752
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.139641325280375,
+      "learning_rate": 9.991337582498042e-06,
+      "loss": 0.7178,
+      "step": 753
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9278302141820551,
+      "learning_rate": 9.991276489934416e-06,
+      "loss": 0.7362,
+      "step": 754
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2343855779864767,
+      "learning_rate": 9.991215182884521e-06,
+      "loss": 0.6939,
+      "step": 755
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9772578586161467,
+      "learning_rate": 9.991153661350996e-06,
+      "loss": 0.9272,
+      "step": 756
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7868731785596128,
+      "learning_rate": 9.991091925336481e-06,
+      "loss": 0.9432,
+      "step": 757
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.342012062004596,
+      "learning_rate": 9.991029974843632e-06,
+      "loss": 0.7664,
+      "step": 758
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.310698311594551,
+      "learning_rate": 9.990967809875107e-06,
+      "loss": 0.9726,
+      "step": 759
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1672511987544527,
+      "learning_rate": 9.990905430433582e-06,
+      "loss": 0.9006,
+      "step": 760
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2192402482447067,
+      "learning_rate": 9.990842836521732e-06,
+      "loss": 0.8403,
+      "step": 761
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1508292719443336,
+      "learning_rate": 9.990780028142253e-06,
+      "loss": 0.8652,
+      "step": 762
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.213935031915199,
+      "learning_rate": 9.990717005297841e-06,
+      "loss": 0.9556,
+      "step": 763
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.352201275739576,
+      "learning_rate": 9.990653767991203e-06,
+      "loss": 0.7482,
+      "step": 764
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3567334849808168,
+      "learning_rate": 9.99059031622506e-06,
+      "loss": 0.7347,
+      "step": 765
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4360118797716104,
+      "learning_rate": 9.990526650002133e-06,
+      "loss": 0.9245,
+      "step": 766
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.421917186426181,
+      "learning_rate": 9.990462769325163e-06,
+      "loss": 0.8476,
+      "step": 767
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.873973378039041,
+      "learning_rate": 9.990398674196896e-06,
+      "loss": 0.7815,
+      "step": 768
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.277689392309156,
+      "learning_rate": 9.99033436462008e-06,
+      "loss": 0.8566,
+      "step": 769
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9678995152691143,
+      "learning_rate": 9.990269840597484e-06,
+      "loss": 0.9549,
+      "step": 770
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0128282141470764,
+      "learning_rate": 9.990205102131879e-06,
+      "loss": 0.833,
+      "step": 771
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9172825640427336,
+      "learning_rate": 9.990140149226047e-06,
+      "loss": 0.9394,
+      "step": 772
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1326649561986764,
+      "learning_rate": 9.99007498188278e-06,
+      "loss": 0.8303,
+      "step": 773
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3890236780254073,
+      "learning_rate": 9.990009600104875e-06,
+      "loss": 0.6357,
+      "step": 774
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9932508128912179,
+      "learning_rate": 9.989944003895146e-06,
+      "loss": 0.9015,
+      "step": 775
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.151013135338263,
+      "learning_rate": 9.98987819325641e-06,
+      "loss": 0.9646,
+      "step": 776
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8113963372226076,
+      "learning_rate": 9.989812168191495e-06,
+      "loss": 0.7464,
+      "step": 777
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4769961759578383,
+      "learning_rate": 9.989745928703237e-06,
+      "loss": 0.8518,
+      "step": 778
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0091734313639433,
+      "learning_rate": 9.989679474794484e-06,
+      "loss": 0.9005,
+      "step": 779
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7252762016481575,
+      "learning_rate": 9.989612806468092e-06,
+      "loss": 0.8538,
+      "step": 780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8227194482345939,
+      "learning_rate": 9.989545923726925e-06,
+      "loss": 0.8469,
+      "step": 781
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.399717128374073,
+      "learning_rate": 9.989478826573857e-06,
+      "loss": 0.9592,
+      "step": 782
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0480760910875198,
+      "learning_rate": 9.989411515011773e-06,
+      "loss": 0.8972,
+      "step": 783
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8866787934643554,
+      "learning_rate": 9.989343989043563e-06,
+      "loss": 0.8185,
+      "step": 784
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8598101268879568,
+      "learning_rate": 9.989276248672131e-06,
+      "loss": 0.9842,
+      "step": 785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9157923704407196,
+      "learning_rate": 9.989208293900386e-06,
+      "loss": 0.8418,
+      "step": 786
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2400816310227174,
+      "learning_rate": 9.98914012473125e-06,
+      "loss": 0.7938,
+      "step": 787
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2083517176407588,
+      "learning_rate": 9.989071741167649e-06,
+      "loss": 0.9783,
+      "step": 788
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9282549858278912,
+      "learning_rate": 9.989003143212526e-06,
+      "loss": 0.8502,
+      "step": 789
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.101904670399808,
+      "learning_rate": 9.988934330868826e-06,
+      "loss": 0.8979,
+      "step": 790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2459278521408255,
+      "learning_rate": 9.988865304139509e-06,
+      "loss": 0.8078,
+      "step": 791
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0379799323641947,
+      "learning_rate": 9.988796063027534e-06,
+      "loss": 0.9341,
+      "step": 792
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.902253124945575,
+      "learning_rate": 9.988726607535885e-06,
+      "loss": 0.9114,
+      "step": 793
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7059436444893277,
+      "learning_rate": 9.988656937667544e-06,
+      "loss": 0.9222,
+      "step": 794
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8406901820130535,
+      "learning_rate": 9.988587053425504e-06,
+      "loss": 0.9905,
+      "step": 795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4475738437489873,
+      "learning_rate": 9.988516954812767e-06,
+      "loss": 0.7726,
+      "step": 796
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.077902433185746,
+      "learning_rate": 9.988446641832346e-06,
+      "loss": 0.9417,
+      "step": 797
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8782893876083433,
+      "learning_rate": 9.988376114487264e-06,
+      "loss": 0.8704,
+      "step": 798
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.713316296575651,
+      "learning_rate": 9.98830537278055e-06,
+      "loss": 0.9155,
+      "step": 799
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.789419890839306,
+      "learning_rate": 9.988234416715246e-06,
+      "loss": 0.7601,
+      "step": 800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.378303833251951,
+      "learning_rate": 9.988163246294398e-06,
+      "loss": 0.9259,
+      "step": 801
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9550569477600979,
+      "learning_rate": 9.988091861521067e-06,
+      "loss": 1.0043,
+      "step": 802
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6506055888115683,
+      "learning_rate": 9.98802026239832e-06,
+      "loss": 0.8317,
+      "step": 803
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.939688572510093,
+      "learning_rate": 9.987948448929232e-06,
+      "loss": 0.9274,
+      "step": 804
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.485569154997922,
+      "learning_rate": 9.98787642111689e-06,
+      "loss": 0.8204,
+      "step": 805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1379840523946974,
+      "learning_rate": 9.987804178964391e-06,
+      "loss": 0.8197,
+      "step": 806
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9728538651533571,
+      "learning_rate": 9.987731722474839e-06,
+      "loss": 0.8467,
+      "step": 807
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3134471811973842,
+      "learning_rate": 9.987659051651345e-06,
+      "loss": 0.8371,
+      "step": 808
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9459071587386059,
+      "learning_rate": 9.987586166497032e-06,
+      "loss": 0.7538,
+      "step": 809
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8460821632049202,
+      "learning_rate": 9.987513067015035e-06,
+      "loss": 0.8211,
+      "step": 810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2897896995198623,
+      "learning_rate": 9.987439753208494e-06,
+      "loss": 0.762,
+      "step": 811
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3585055354678095,
+      "learning_rate": 9.98736622508056e-06,
+      "loss": 0.7368,
+      "step": 812
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.299161258550293,
+      "learning_rate": 9.98729248263439e-06,
+      "loss": 0.7326,
+      "step": 813
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.828019247130325,
+      "learning_rate": 9.987218525873155e-06,
+      "loss": 0.8408,
+      "step": 814
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4024837536180195,
+      "learning_rate": 9.987144354800035e-06,
+      "loss": 0.6325,
+      "step": 815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8550350234151551,
+      "learning_rate": 9.987069969418213e-06,
+      "loss": 0.793,
+      "step": 816
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8867849454321015,
+      "learning_rate": 9.986995369730887e-06,
+      "loss": 0.8626,
+      "step": 817
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2229441233992446,
+      "learning_rate": 9.986920555741267e-06,
+      "loss": 0.8471,
+      "step": 818
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9010436274229472,
+      "learning_rate": 9.98684552745256e-06,
+      "loss": 0.8358,
+      "step": 819
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3752666031258765,
+      "learning_rate": 9.986770284867997e-06,
+      "loss": 0.6929,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.225939810301291,
+      "learning_rate": 9.986694827990807e-06,
+      "loss": 0.9446,
+      "step": 821
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.512227500792467,
+      "learning_rate": 9.986619156824236e-06,
+      "loss": 0.7812,
+      "step": 822
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8453996397657726,
+      "learning_rate": 9.986543271371532e-06,
+      "loss": 0.8715,
+      "step": 823
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5095964600006027,
+      "learning_rate": 9.98646717163596e-06,
+      "loss": 0.7419,
+      "step": 824
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8713718424788457,
+      "learning_rate": 9.986390857620787e-06,
+      "loss": 0.8864,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3502444786305636,
+      "learning_rate": 9.986314329329294e-06,
+      "loss": 0.7462,
+      "step": 826
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3400919488020424,
+      "learning_rate": 9.986237586764768e-06,
+      "loss": 0.7122,
+      "step": 827
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.073973505717456,
+      "learning_rate": 9.98616062993051e-06,
+      "loss": 0.8712,
+      "step": 828
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9741112161842078,
+      "learning_rate": 9.986083458829824e-06,
+      "loss": 0.7895,
+      "step": 829
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1096362372425297,
+      "learning_rate": 9.986006073466027e-06,
+      "loss": 0.8964,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9132360933490484,
+      "learning_rate": 9.985928473842445e-06,
+      "loss": 0.9117,
+      "step": 831
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0610755305980937,
+      "learning_rate": 9.985850659962411e-06,
+      "loss": 0.7975,
+      "step": 832
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2962105255644407,
+      "learning_rate": 9.985772631829272e-06,
+      "loss": 0.8032,
+      "step": 833
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1653515449200595,
+      "learning_rate": 9.985694389446378e-06,
+      "loss": 0.9451,
+      "step": 834
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2990909949703822,
+      "learning_rate": 9.985615932817092e-06,
+      "loss": 0.6967,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9267903132409716,
+      "learning_rate": 9.985537261944787e-06,
+      "loss": 0.8107,
+      "step": 836
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.000333143925124,
+      "learning_rate": 9.985458376832842e-06,
+      "loss": 0.834,
+      "step": 837
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1728865753826563,
+      "learning_rate": 9.985379277484646e-06,
+      "loss": 0.9454,
+      "step": 838
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4094967605213897,
+      "learning_rate": 9.9852999639036e-06,
+      "loss": 0.9421,
+      "step": 839
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.118330990399244,
+      "learning_rate": 9.985220436093112e-06,
+      "loss": 0.8546,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.980865806117518,
+      "learning_rate": 9.985140694056601e-06,
+      "loss": 1.0389,
+      "step": 841
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.690656248958916,
+      "learning_rate": 9.985060737797489e-06,
+      "loss": 0.9451,
+      "step": 842
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3547908857615785,
+      "learning_rate": 9.984980567319216e-06,
+      "loss": 0.8586,
+      "step": 843
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.260232155299939,
+      "learning_rate": 9.984900182625226e-06,
+      "loss": 1.0086,
+      "step": 844
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.885565566296528,
+      "learning_rate": 9.984819583718973e-06,
+      "loss": 0.8416,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8781304238179544,
+      "learning_rate": 9.984738770603921e-06,
+      "loss": 0.824,
+      "step": 846
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1528675449648138,
+      "learning_rate": 9.984657743283543e-06,
+      "loss": 0.6453,
+      "step": 847
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.110623285196561,
+      "learning_rate": 9.984576501761319e-06,
+      "loss": 0.9335,
+      "step": 848
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3883684966164915,
+      "learning_rate": 9.98449504604074e-06,
+      "loss": 0.8015,
+      "step": 849
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.342943443051584,
+      "learning_rate": 9.984413376125312e-06,
+      "loss": 0.8736,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9051184326228061,
+      "learning_rate": 9.984331492018537e-06,
+      "loss": 0.7941,
+      "step": 851
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1047439830120824,
+      "learning_rate": 9.984249393723939e-06,
+      "loss": 0.8642,
+      "step": 852
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9113199504294918,
+      "learning_rate": 9.984167081245044e-06,
+      "loss": 0.9909,
+      "step": 853
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7166967638839492,
+      "learning_rate": 9.984084554585387e-06,
+      "loss": 0.8921,
+      "step": 854
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.011031013265907,
+      "learning_rate": 9.984001813748519e-06,
+      "loss": 0.8519,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9288702937473416,
+      "learning_rate": 9.98391885873799e-06,
+      "loss": 0.8376,
+      "step": 856
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.797426823576584,
+      "learning_rate": 9.98383568955737e-06,
+      "loss": 0.8489,
+      "step": 857
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.090570855058258,
+      "learning_rate": 9.983752306210233e-06,
+      "loss": 0.9157,
+      "step": 858
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8569684062409113,
+      "learning_rate": 9.983668708700156e-06,
+      "loss": 0.9057,
+      "step": 859
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7647662172669907,
+      "learning_rate": 9.983584897030737e-06,
+      "loss": 0.8101,
+      "step": 860
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.130477882786362,
+      "learning_rate": 9.983500871205577e-06,
+      "loss": 0.9524,
+      "step": 861
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9061963198455947,
+      "learning_rate": 9.983416631228284e-06,
+      "loss": 0.9147,
+      "step": 862
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5138033964991418,
+      "learning_rate": 9.98333217710248e-06,
+      "loss": 0.7557,
+      "step": 863
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8644468033975627,
+      "learning_rate": 9.983247508831795e-06,
+      "loss": 0.8216,
+      "step": 864
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3587014628000746,
+      "learning_rate": 9.983162626419866e-06,
+      "loss": 0.8329,
+      "step": 865
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.22414104756627,
+      "learning_rate": 9.98307752987034e-06,
+      "loss": 0.9284,
+      "step": 866
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1087556647400143,
+      "learning_rate": 9.982992219186876e-06,
+      "loss": 0.6253,
+      "step": 867
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.068623698211126,
+      "learning_rate": 9.982906694373136e-06,
+      "loss": 0.8325,
+      "step": 868
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8013353230072524,
+      "learning_rate": 9.9828209554328e-06,
+      "loss": 0.6738,
+      "step": 869
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.862429973255459,
+      "learning_rate": 9.98273500236955e-06,
+      "loss": 0.9048,
+      "step": 870
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5389031230252372,
+      "learning_rate": 9.98264883518708e-06,
+      "loss": 0.5814,
+      "step": 871
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.590158188776279,
+      "learning_rate": 9.98256245388909e-06,
+      "loss": 0.8479,
+      "step": 872
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.453701413011588,
+      "learning_rate": 9.982475858479298e-06,
+      "loss": 0.8478,
+      "step": 873
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9557115210010327,
+      "learning_rate": 9.982389048961421e-06,
+      "loss": 0.8387,
+      "step": 874
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6949052735945678,
+      "learning_rate": 9.98230202533919e-06,
+      "loss": 0.7887,
+      "step": 875
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.255539176672344,
+      "learning_rate": 9.982214787616345e-06,
+      "loss": 0.8297,
+      "step": 876
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.084927867215076,
+      "learning_rate": 9.982127335796631e-06,
+      "loss": 0.8797,
+      "step": 877
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9772548275739261,
+      "learning_rate": 9.982039669883815e-06,
+      "loss": 0.7657,
+      "step": 878
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0065286826397593,
+      "learning_rate": 9.981951789881657e-06,
+      "loss": 0.944,
+      "step": 879
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.111598680390501,
+      "learning_rate": 9.981863695793934e-06,
+      "loss": 0.7497,
+      "step": 880
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9954663781413597,
+      "learning_rate": 9.981775387624433e-06,
+      "loss": 0.7624,
+      "step": 881
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.089461788885941,
+      "learning_rate": 9.98168686537695e-06,
+      "loss": 0.8834,
+      "step": 882
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9486510988779022,
+      "learning_rate": 9.981598129055286e-06,
+      "loss": 0.8574,
+      "step": 883
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8960789235180562,
+      "learning_rate": 9.981509178663256e-06,
+      "loss": 0.8029,
+      "step": 884
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.204238516394484,
+      "learning_rate": 9.981420014204682e-06,
+      "loss": 0.6156,
+      "step": 885
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6089698395975143,
+      "learning_rate": 9.981330635683396e-06,
+      "loss": 0.9101,
+      "step": 886
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3244445384893724,
+      "learning_rate": 9.981241043103239e-06,
+      "loss": 0.7101,
+      "step": 887
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2598740365086454,
+      "learning_rate": 9.981151236468061e-06,
+      "loss": 0.8608,
+      "step": 888
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2461884491780988,
+      "learning_rate": 9.98106121578172e-06,
+      "loss": 0.6231,
+      "step": 889
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9388656045303065,
+      "learning_rate": 9.980970981048086e-06,
+      "loss": 0.7894,
+      "step": 890
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.702964930681584,
+      "learning_rate": 9.980880532271033e-06,
+      "loss": 0.8327,
+      "step": 891
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9418466392145457,
+      "learning_rate": 9.980789869454454e-06,
+      "loss": 0.8967,
+      "step": 892
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.820679616142403,
+      "learning_rate": 9.980698992602241e-06,
+      "loss": 0.8513,
+      "step": 893
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2069915477396282,
+      "learning_rate": 9.980607901718297e-06,
+      "loss": 0.6462,
+      "step": 894
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7944411606545512,
+      "learning_rate": 9.98051659680654e-06,
+      "loss": 0.7591,
+      "step": 895
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0928106519006193,
+      "learning_rate": 9.980425077870895e-06,
+      "loss": 0.8709,
+      "step": 896
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1236849372836397,
+      "learning_rate": 9.98033334491529e-06,
+      "loss": 0.7098,
+      "step": 897
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3309072210285773,
+      "learning_rate": 9.98024139794367e-06,
+      "loss": 0.6746,
+      "step": 898
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9956247363703732,
+      "learning_rate": 9.980149236959986e-06,
+      "loss": 0.9375,
+      "step": 899
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1030150422619616,
+      "learning_rate": 9.980056861968197e-06,
+      "loss": 0.6796,
+      "step": 900
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.173952570730066,
+      "learning_rate": 9.979964272972275e-06,
+      "loss": 0.7976,
+      "step": 901
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8877044718487412,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 0.7039,
+      "step": 902
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7585988298604256,
+      "learning_rate": 9.979778452983949e-06,
+      "loss": 0.8472,
+      "step": 903
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1735637146198314,
+      "learning_rate": 9.979685221999532e-06,
+      "loss": 0.957,
+      "step": 904
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5966411283746693,
+      "learning_rate": 9.979591777026952e-06,
+      "loss": 0.7749,
+      "step": 905
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2460267214338119,
+      "learning_rate": 9.97949811807022e-06,
+      "loss": 0.747,
+      "step": 906
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0370906776909874,
+      "learning_rate": 9.979404245133367e-06,
+      "loss": 0.6944,
+      "step": 907
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6722494023998788,
+      "learning_rate": 9.979310158220422e-06,
+      "loss": 0.7723,
+      "step": 908
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0419533146135644,
+      "learning_rate": 9.97921585733543e-06,
+      "loss": 0.8359,
+      "step": 909
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.026216338032198,
+      "learning_rate": 9.979121342482442e-06,
+      "loss": 0.8469,
+      "step": 910
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9643351719884456,
+      "learning_rate": 9.979026613665523e-06,
+      "loss": 0.8208,
+      "step": 911
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0962294772998105,
+      "learning_rate": 9.978931670888741e-06,
+      "loss": 0.9993,
+      "step": 912
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.316983850642248,
+      "learning_rate": 9.978836514156177e-06,
+      "loss": 0.807,
+      "step": 913
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3938468575207517,
+      "learning_rate": 9.97874114347192e-06,
+      "loss": 0.8751,
+      "step": 914
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8576731274768068,
+      "learning_rate": 9.978645558840066e-06,
+      "loss": 0.8898,
+      "step": 915
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9379765345482602,
+      "learning_rate": 9.978549760264724e-06,
+      "loss": 0.8031,
+      "step": 916
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4886995355455033,
+      "learning_rate": 9.978453747750012e-06,
+      "loss": 0.706,
+      "step": 917
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4024621019890084,
+      "learning_rate": 9.978357521300054e-06,
+      "loss": 0.6803,
+      "step": 918
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.942676982063723,
+      "learning_rate": 9.978261080918988e-06,
+      "loss": 1.0592,
+      "step": 919
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9193585992415372,
+      "learning_rate": 9.978164426610956e-06,
+      "loss": 0.7918,
+      "step": 920
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.041815358966382,
+      "learning_rate": 9.97806755838011e-06,
+      "loss": 0.7398,
+      "step": 921
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8381219544723162,
+      "learning_rate": 9.977970476230615e-06,
+      "loss": 0.9289,
+      "step": 922
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2759327916901066,
+      "learning_rate": 9.977873180166641e-06,
+      "loss": 0.6748,
+      "step": 923
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7084303555706357,
+      "learning_rate": 9.977775670192373e-06,
+      "loss": 0.7575,
+      "step": 924
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0870770433846006,
+      "learning_rate": 9.977677946311995e-06,
+      "loss": 0.9322,
+      "step": 925
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9580293888605134,
+      "learning_rate": 9.977580008529713e-06,
+      "loss": 0.7867,
+      "step": 926
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.158538883623233,
+      "learning_rate": 9.977481856849729e-06,
+      "loss": 0.8527,
+      "step": 927
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8214868371167787,
+      "learning_rate": 9.977383491276267e-06,
+      "loss": 0.9518,
+      "step": 928
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.900852307918768,
+      "learning_rate": 9.977284911813549e-06,
+      "loss": 0.7949,
+      "step": 929
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.390725242736315,
+      "learning_rate": 9.977186118465813e-06,
+      "loss": 0.6303,
+      "step": 930
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7784248368863096,
+      "learning_rate": 9.977087111237307e-06,
+      "loss": 0.8117,
+      "step": 931
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2807839346552186,
+      "learning_rate": 9.976987890132283e-06,
+      "loss": 0.5401,
+      "step": 932
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8345165000965125,
+      "learning_rate": 9.976888455155002e-06,
+      "loss": 0.7967,
+      "step": 933
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4903175010953524,
+      "learning_rate": 9.976788806309742e-06,
+      "loss": 0.8521,
+      "step": 934
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0788557734769637,
+      "learning_rate": 9.976688943600786e-06,
+      "loss": 0.8827,
+      "step": 935
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5883417794454873,
+      "learning_rate": 9.97658886703242e-06,
+      "loss": 0.6478,
+      "step": 936
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6402423460904192,
+      "learning_rate": 9.976488576608945e-06,
+      "loss": 0.8602,
+      "step": 937
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2199316153063893,
+      "learning_rate": 9.976388072334674e-06,
+      "loss": 0.8766,
+      "step": 938
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8428829799709294,
+      "learning_rate": 9.976287354213924e-06,
+      "loss": 0.859,
+      "step": 939
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8746456533400557,
+      "learning_rate": 9.976186422251026e-06,
+      "loss": 0.8489,
+      "step": 940
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9849725445019337,
+      "learning_rate": 9.976085276450312e-06,
+      "loss": 0.8782,
+      "step": 941
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0208775452722354,
+      "learning_rate": 9.975983916816134e-06,
+      "loss": 0.9727,
+      "step": 942
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3372217573939462,
+      "learning_rate": 9.975882343352843e-06,
+      "loss": 0.7354,
+      "step": 943
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9798588234422247,
+      "learning_rate": 9.975780556064806e-06,
+      "loss": 1.046,
+      "step": 944
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.835580672184335,
+      "learning_rate": 9.975678554956397e-06,
+      "loss": 0.8099,
+      "step": 945
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.141774314621427,
+      "learning_rate": 9.975576340032e-06,
+      "loss": 0.8044,
+      "step": 946
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.762413843838386,
+      "learning_rate": 9.975473911296005e-06,
+      "loss": 0.8601,
+      "step": 947
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0245069954965915,
+      "learning_rate": 9.975371268752816e-06,
+      "loss": 0.7001,
+      "step": 948
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.000511279837186,
+      "learning_rate": 9.975268412406842e-06,
+      "loss": 0.8827,
+      "step": 949
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.407660538463852,
+      "learning_rate": 9.975165342262505e-06,
+      "loss": 0.8338,
+      "step": 950
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.027909742330816,
+      "learning_rate": 9.975062058324232e-06,
+      "loss": 0.8428,
+      "step": 951
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5925010707314349,
+      "learning_rate": 9.974958560596464e-06,
+      "loss": 0.6829,
+      "step": 952
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.861085843798142,
+      "learning_rate": 9.974854849083645e-06,
+      "loss": 0.8775,
+      "step": 953
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7486113021978387,
+      "learning_rate": 9.974750923790234e-06,
+      "loss": 0.812,
+      "step": 954
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1927285796767058,
+      "learning_rate": 9.974646784720696e-06,
+      "loss": 0.7162,
+      "step": 955
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8266046652994048,
+      "learning_rate": 9.974542431879507e-06,
+      "loss": 0.911,
+      "step": 956
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6170479787052265,
+      "learning_rate": 9.974437865271152e-06,
+      "loss": 0.8029,
+      "step": 957
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.748929060666404,
+      "learning_rate": 9.97433308490012e-06,
+      "loss": 0.8673,
+      "step": 958
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9824857863148033,
+      "learning_rate": 9.97422809077092e-06,
+      "loss": 1.0105,
+      "step": 959
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0037893414175545,
+      "learning_rate": 9.97412288288806e-06,
+      "loss": 0.816,
+      "step": 960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8087361149000016,
+      "learning_rate": 9.974017461256062e-06,
+      "loss": 0.8733,
+      "step": 961
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.677792284098914,
+      "learning_rate": 9.973911825879454e-06,
+      "loss": 0.8005,
+      "step": 962
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5637676628402644,
+      "learning_rate": 9.97380597676278e-06,
+      "loss": 0.7017,
+      "step": 963
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.693131987710579,
+      "learning_rate": 9.973699913910584e-06,
+      "loss": 0.9043,
+      "step": 964
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7841172734506114,
+      "learning_rate": 9.973593637327427e-06,
+      "loss": 0.835,
+      "step": 965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7230268865417928,
+      "learning_rate": 9.973487147017874e-06,
+      "loss": 0.7716,
+      "step": 966
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8121414369939421,
+      "learning_rate": 9.9733804429865e-06,
+      "loss": 0.8457,
+      "step": 967
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.063060148722983,
+      "learning_rate": 9.973273525237895e-06,
+      "loss": 0.874,
+      "step": 968
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7205897413307498,
+      "learning_rate": 9.97316639377665e-06,
+      "loss": 0.7816,
+      "step": 969
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1695028399917498,
+      "learning_rate": 9.973059048607367e-06,
+      "loss": 0.6885,
+      "step": 970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.818698199502042,
+      "learning_rate": 9.972951489734664e-06,
+      "loss": 0.8639,
+      "step": 971
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6867298734766287,
+      "learning_rate": 9.972843717163159e-06,
+      "loss": 0.7377,
+      "step": 972
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6977440064778972,
+      "learning_rate": 9.972735730897484e-06,
+      "loss": 0.9386,
+      "step": 973
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.535258489073851,
+      "learning_rate": 9.97262753094228e-06,
+      "loss": 0.7885,
+      "step": 974
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7632877764742974,
+      "learning_rate": 9.972519117302195e-06,
+      "loss": 0.7559,
+      "step": 975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.831049316248215,
+      "learning_rate": 9.972410489981892e-06,
+      "loss": 0.8843,
+      "step": 976
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.491609564433192,
+      "learning_rate": 9.972301648986033e-06,
+      "loss": 0.6956,
+      "step": 977
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.224934659879179,
+      "learning_rate": 9.972192594319301e-06,
+      "loss": 0.7023,
+      "step": 978
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1409976212569797,
+      "learning_rate": 9.972083325986377e-06,
+      "loss": 0.9025,
+      "step": 979
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9397006406783557,
+      "learning_rate": 9.97197384399196e-06,
+      "loss": 0.8466,
+      "step": 980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4008509860477942,
+      "learning_rate": 9.971864148340755e-06,
+      "loss": 0.8917,
+      "step": 981
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9137822002880238,
+      "learning_rate": 9.971754239037474e-06,
+      "loss": 0.8219,
+      "step": 982
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.910366882056281,
+      "learning_rate": 9.971644116086841e-06,
+      "loss": 0.8473,
+      "step": 983
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7731956315405575,
+      "learning_rate": 9.971533779493586e-06,
+      "loss": 0.7719,
+      "step": 984
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8488799057516347,
+      "learning_rate": 9.971423229262454e-06,
+      "loss": 0.8207,
+      "step": 985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8405413972035904,
+      "learning_rate": 9.971312465398192e-06,
+      "loss": 0.7139,
+      "step": 986
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7277576701881916,
+      "learning_rate": 9.971201487905563e-06,
+      "loss": 0.9715,
+      "step": 987
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8660348385567558,
+      "learning_rate": 9.971090296789336e-06,
+      "loss": 0.9191,
+      "step": 988
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7665561918784285,
+      "learning_rate": 9.970978892054286e-06,
+      "loss": 0.8727,
+      "step": 989
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8627989498974815,
+      "learning_rate": 9.970867273705203e-06,
+      "loss": 0.8047,
+      "step": 990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8007098720796095,
+      "learning_rate": 9.970755441746882e-06,
+      "loss": 0.9094,
+      "step": 991
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.673938101845992,
+      "learning_rate": 9.970643396184128e-06,
+      "loss": 0.7791,
+      "step": 992
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8319782266884572,
+      "learning_rate": 9.970531137021759e-06,
+      "loss": 0.8492,
+      "step": 993
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.150450691882396,
+      "learning_rate": 9.970418664264596e-06,
+      "loss": 0.9398,
+      "step": 994
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.279578956179492,
+      "learning_rate": 9.970305977917473e-06,
+      "loss": 0.8088,
+      "step": 995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1238482432617267,
+      "learning_rate": 9.970193077985235e-06,
+      "loss": 0.7183,
+      "step": 996
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0843018238374973,
+      "learning_rate": 9.97007996447273e-06,
+      "loss": 0.8437,
+      "step": 997
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7185722835483443,
+      "learning_rate": 9.969966637384818e-06,
+      "loss": 0.8566,
+      "step": 998
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0495719211519723,
+      "learning_rate": 9.969853096726372e-06,
+      "loss": 0.8876,
+      "step": 999
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0787367175346025,
+      "learning_rate": 9.969739342502272e-06,
+      "loss": 0.7913,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6683274164114175,
+      "learning_rate": 9.969625374717401e-06,
+      "loss": 0.8122,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7975530308268264,
+      "learning_rate": 9.969511193376661e-06,
+      "loss": 0.8495,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2842295190592123,
+      "learning_rate": 9.969396798484958e-06,
+      "loss": 0.8685,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8320958363019564,
+      "learning_rate": 9.969282190047207e-06,
+      "loss": 0.8583,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4499232285684056,
+      "learning_rate": 9.969167368068334e-06,
+      "loss": 0.9774,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.054579401619427,
+      "learning_rate": 9.96905233255327e-06,
+      "loss": 0.8134,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.871914649666746,
+      "learning_rate": 9.968937083506962e-06,
+      "loss": 0.8816,
+      "step": 1007
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.616116551976703,
+      "learning_rate": 9.96882162093436e-06,
+      "loss": 0.7576,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.786769016984922,
+      "learning_rate": 9.968705944840428e-06,
+      "loss": 0.8156,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.093636074830123,
+      "learning_rate": 9.968590055230135e-06,
+      "loss": 0.8428,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7131188114889406,
+      "learning_rate": 9.968473952108462e-06,
+      "loss": 0.8634,
+      "step": 1011
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9624358452877249,
+      "learning_rate": 9.968357635480397e-06,
+      "loss": 0.8242,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.049858050690435,
+      "learning_rate": 9.96824110535094e-06,
+      "loss": 0.8621,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8944145038468312,
+      "learning_rate": 9.968124361725098e-06,
+      "loss": 0.936,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.173941716066314,
+      "learning_rate": 9.968007404607887e-06,
+      "loss": 0.9101,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1886126984471943,
+      "learning_rate": 9.967890234004335e-06,
+      "loss": 0.5795,
+      "step": 1016
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7950001306893826,
+      "learning_rate": 9.967772849919474e-06,
+      "loss": 0.7866,
+      "step": 1017
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8068881248097173,
+      "learning_rate": 9.96765525235835e-06,
+      "loss": 0.8853,
+      "step": 1018
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.88939287381751,
+      "learning_rate": 9.967537441326018e-06,
+      "loss": 0.8576,
+      "step": 1019
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.342205527905726,
+      "learning_rate": 9.967419416827537e-06,
+      "loss": 0.901,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7396706029327236,
+      "learning_rate": 9.967301178867981e-06,
+      "loss": 0.8926,
+      "step": 1021
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7941151909148045,
+      "learning_rate": 9.96718272745243e-06,
+      "loss": 0.7804,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0126090061065183,
+      "learning_rate": 9.967064062585977e-06,
+      "loss": 0.8389,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1641449145842926,
+      "learning_rate": 9.966945184273716e-06,
+      "loss": 0.7645,
+      "step": 1024
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1356599045929892,
+      "learning_rate": 9.96682609252076e-06,
+      "loss": 0.697,
+      "step": 1025
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7167243030687345,
+      "learning_rate": 9.966706787332224e-06,
+      "loss": 0.8127,
+      "step": 1026
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7464992447729195,
+      "learning_rate": 9.966587268713237e-06,
+      "loss": 0.7847,
+      "step": 1027
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.22558212555563,
+      "learning_rate": 9.966467536668933e-06,
+      "loss": 0.8558,
+      "step": 1028
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.950149594414034,
+      "learning_rate": 9.966347591204459e-06,
+      "loss": 0.7894,
+      "step": 1029
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.150177130286734,
+      "learning_rate": 9.96622743232497e-06,
+      "loss": 0.7349,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7304101571045156,
+      "learning_rate": 9.966107060035624e-06,
+      "loss": 0.8193,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1771405356205467,
+      "learning_rate": 9.9659864743416e-06,
+      "loss": 0.843,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.060599907005229,
+      "learning_rate": 9.965865675248077e-06,
+      "loss": 0.8062,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8291681456943882,
+      "learning_rate": 9.965744662760246e-06,
+      "loss": 0.7592,
+      "step": 1034
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6955247729861171,
+      "learning_rate": 9.965623436883307e-06,
+      "loss": 0.7814,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.006435927564684,
+      "learning_rate": 9.96550199762247e-06,
+      "loss": 0.9182,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.720105932658694,
+      "learning_rate": 9.965380344982955e-06,
+      "loss": 0.8218,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9063220961203309,
+      "learning_rate": 9.965258478969985e-06,
+      "loss": 0.8501,
+      "step": 1038
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9567694216566442,
+      "learning_rate": 9.965136399588803e-06,
+      "loss": 0.8025,
+      "step": 1039
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9377734927990453,
+      "learning_rate": 9.96501410684465e-06,
+      "loss": 0.7683,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8836125884575197,
+      "learning_rate": 9.964891600742785e-06,
+      "loss": 0.864,
+      "step": 1041
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6861321081219987,
+      "learning_rate": 9.96476888128847e-06,
+      "loss": 0.8366,
+      "step": 1042
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9928050535078465,
+      "learning_rate": 9.964645948486978e-06,
+      "loss": 0.8797,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6624010406674237,
+      "learning_rate": 9.964522802343593e-06,
+      "loss": 0.8235,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7844214889247871,
+      "learning_rate": 9.964399442863609e-06,
+      "loss": 0.7939,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7081864988686224,
+      "learning_rate": 9.96427587005232e-06,
+      "loss": 0.887,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.856224122511381,
+      "learning_rate": 9.964152083915045e-06,
+      "loss": 0.8331,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7428402828969753,
+      "learning_rate": 9.964028084457098e-06,
+      "loss": 0.9432,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8118625075150963,
+      "learning_rate": 9.963903871683806e-06,
+      "loss": 0.9105,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2441480907428741,
+      "learning_rate": 9.963779445600512e-06,
+      "loss": 0.6919,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0036286321729406,
+      "learning_rate": 9.963654806212561e-06,
+      "loss": 0.8125,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6762988995781878,
+      "learning_rate": 9.963529953525307e-06,
+      "loss": 0.7695,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9270458549918934,
+      "learning_rate": 9.963404887544117e-06,
+      "loss": 0.9022,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8688359724060015,
+      "learning_rate": 9.963279608274364e-06,
+      "loss": 0.9123,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7955064167113686,
+      "learning_rate": 9.963154115721433e-06,
+      "loss": 0.8643,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1465590683755353,
+      "learning_rate": 9.963028409890716e-06,
+      "loss": 0.7671,
+      "step": 1056
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9250447948002012,
+      "learning_rate": 9.962902490787616e-06,
+      "loss": 0.7355,
+      "step": 1057
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.177396711907553,
+      "learning_rate": 9.962776358417543e-06,
+      "loss": 0.9742,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7486161657303665,
+      "learning_rate": 9.962650012785917e-06,
+      "loss": 0.9089,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7133215321761532,
+      "learning_rate": 9.962523453898166e-06,
+      "loss": 0.8738,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8855753009100498,
+      "learning_rate": 9.96239668175973e-06,
+      "loss": 0.7587,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8795425078976413,
+      "learning_rate": 9.962269696376059e-06,
+      "loss": 0.7769,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.153146203057806,
+      "learning_rate": 9.962142497752605e-06,
+      "loss": 0.833,
+      "step": 1063
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0684741171614864,
+      "learning_rate": 9.962015085894838e-06,
+      "loss": 0.5552,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9981626744652363,
+      "learning_rate": 9.961887460808232e-06,
+      "loss": 0.8485,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7315063845286376,
+      "learning_rate": 9.96175962249827e-06,
+      "loss": 0.8144,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9318246762161981,
+      "learning_rate": 9.961631570970449e-06,
+      "loss": 0.7964,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.830962658123044,
+      "learning_rate": 9.961503306230267e-06,
+      "loss": 0.8087,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.252614384501908,
+      "learning_rate": 9.961374828283239e-06,
+      "loss": 0.7511,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6512105999009963,
+      "learning_rate": 9.961246137134883e-06,
+      "loss": 0.8023,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9763322521720177,
+      "learning_rate": 9.961117232790734e-06,
+      "loss": 1.001,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2557909147551425,
+      "learning_rate": 9.960988115256328e-06,
+      "loss": 0.9202,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9449521617393313,
+      "learning_rate": 9.960858784537213e-06,
+      "loss": 0.8247,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6098082506821954,
+      "learning_rate": 9.960729240638947e-06,
+      "loss": 0.7918,
+      "step": 1074
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2290267202775178,
+      "learning_rate": 9.9605994835671e-06,
+      "loss": 0.6746,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.289307804201319,
+      "learning_rate": 9.960469513327242e-06,
+      "loss": 0.5949,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9586679977479786,
+      "learning_rate": 9.960339329924964e-06,
+      "loss": 0.8548,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9896911853235957,
+      "learning_rate": 9.960208933365857e-06,
+      "loss": 1.0005,
+      "step": 1078
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.239701426140685,
+      "learning_rate": 9.960078323655524e-06,
+      "loss": 0.9153,
+      "step": 1079
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6778074798589608,
+      "learning_rate": 9.95994750079958e-06,
+      "loss": 0.8241,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6984131057333065,
+      "learning_rate": 9.959816464803647e-06,
+      "loss": 0.8397,
+      "step": 1081
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.580578318219224,
+      "learning_rate": 9.959685215673351e-06,
+      "loss": 0.8101,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.148142743281002,
+      "learning_rate": 9.959553753414338e-06,
+      "loss": 0.6469,
+      "step": 1083
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8359234055077487,
+      "learning_rate": 9.959422078032253e-06,
+      "loss": 0.8411,
+      "step": 1084
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3973353309410772,
+      "learning_rate": 9.959290189532757e-06,
+      "loss": 0.7059,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.872475245581424,
+      "learning_rate": 9.959158087921516e-06,
+      "loss": 0.7361,
+      "step": 1086
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8250981468235667,
+      "learning_rate": 9.95902577320421e-06,
+      "loss": 0.945,
+      "step": 1087
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8282439142323703,
+      "learning_rate": 9.95889324538652e-06,
+      "loss": 0.847,
+      "step": 1088
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.19283757078478,
+      "learning_rate": 9.958760504474144e-06,
+      "loss": 0.7695,
+      "step": 1089
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6819908907470535,
+      "learning_rate": 9.958627550472784e-06,
+      "loss": 0.8471,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9594292694127602,
+      "learning_rate": 9.958494383388155e-06,
+      "loss": 0.8089,
+      "step": 1091
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0752861435024603,
+      "learning_rate": 9.958361003225979e-06,
+      "loss": 0.9241,
+      "step": 1092
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8535411751119064,
+      "learning_rate": 9.958227409991988e-06,
+      "loss": 0.9344,
+      "step": 1093
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8712161049261242,
+      "learning_rate": 9.958093603691923e-06,
+      "loss": 0.8238,
+      "step": 1094
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5847947916320533,
+      "learning_rate": 9.957959584331533e-06,
+      "loss": 0.9027,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9790953187203055,
+      "learning_rate": 9.957825351916579e-06,
+      "loss": 0.9111,
+      "step": 1096
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.079109589458921,
+      "learning_rate": 9.957690906452827e-06,
+      "loss": 0.8171,
+      "step": 1097
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0972295787892623,
+      "learning_rate": 9.957556247946057e-06,
+      "loss": 0.7889,
+      "step": 1098
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1066829804596297,
+      "learning_rate": 9.957421376402053e-06,
+      "loss": 0.9054,
+      "step": 1099
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4426636964979616,
+      "learning_rate": 9.95728629182661e-06,
+      "loss": 0.8141,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7891156680080065,
+      "learning_rate": 9.957150994225535e-06,
+      "loss": 0.8426,
+      "step": 1101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.709878012563874,
+      "learning_rate": 9.957015483604643e-06,
+      "loss": 0.8648,
+      "step": 1102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.423770132144771,
+      "learning_rate": 9.956879759969755e-06,
+      "loss": 0.8754,
+      "step": 1103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9904314911082257,
+      "learning_rate": 9.956743823326704e-06,
+      "loss": 0.8223,
+      "step": 1104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0806599323114727,
+      "learning_rate": 9.956607673681334e-06,
+      "loss": 0.7762,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.63885893242105,
+      "learning_rate": 9.956471311039491e-06,
+      "loss": 0.686,
+      "step": 1106
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9995222004719966,
+      "learning_rate": 9.956334735407037e-06,
+      "loss": 0.8484,
+      "step": 1107
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6713130734501718,
+      "learning_rate": 9.95619794678984e-06,
+      "loss": 0.7383,
+      "step": 1108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.041398058568926,
+      "learning_rate": 9.956060945193781e-06,
+      "loss": 0.5712,
+      "step": 1109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6256787020399694,
+      "learning_rate": 9.955923730624745e-06,
+      "loss": 0.9558,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6292484336401332,
+      "learning_rate": 9.95578630308863e-06,
+      "loss": 0.6726,
+      "step": 1111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0655600125545166,
+      "learning_rate": 9.95564866259134e-06,
+      "loss": 0.8677,
+      "step": 1112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8382200139870986,
+      "learning_rate": 9.95551080913879e-06,
+      "loss": 0.7141,
+      "step": 1113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8839358449032233,
+      "learning_rate": 9.955372742736903e-06,
+      "loss": 0.8652,
+      "step": 1114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7254573080349913,
+      "learning_rate": 9.955234463391612e-06,
+      "loss": 0.9058,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.136549969045551,
+      "learning_rate": 9.955095971108863e-06,
+      "loss": 0.8312,
+      "step": 1116
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7712750982488081,
+      "learning_rate": 9.954957265894602e-06,
+      "loss": 0.8789,
+      "step": 1117
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7276590565482255,
+      "learning_rate": 9.954818347754795e-06,
+      "loss": 0.8104,
+      "step": 1118
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6494922104622474,
+      "learning_rate": 9.954679216695406e-06,
+      "loss": 0.7994,
+      "step": 1119
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6480736765663972,
+      "learning_rate": 9.954539872722417e-06,
+      "loss": 0.8598,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7197088719959939,
+      "learning_rate": 9.954400315841817e-06,
+      "loss": 0.8049,
+      "step": 1121
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9431192214321726,
+      "learning_rate": 9.9542605460596e-06,
+      "loss": 0.7897,
+      "step": 1122
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9093407529910724,
+      "learning_rate": 9.954120563381772e-06,
+      "loss": 0.8767,
+      "step": 1123
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0338564359893074,
+      "learning_rate": 9.953980367814354e-06,
+      "loss": 0.924,
+      "step": 1124
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7384517905885664,
+      "learning_rate": 9.953839959363363e-06,
+      "loss": 0.7915,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9043602187999469,
+      "learning_rate": 9.953699338034838e-06,
+      "loss": 0.8748,
+      "step": 1126
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9884249758581214,
+      "learning_rate": 9.953558503834819e-06,
+      "loss": 0.8493,
+      "step": 1127
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.991673472286364,
+      "learning_rate": 9.95341745676936e-06,
+      "loss": 0.8255,
+      "step": 1128
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0808321616606693,
+      "learning_rate": 9.953276196844519e-06,
+      "loss": 0.8758,
+      "step": 1129
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9701381164654657,
+      "learning_rate": 9.95313472406637e-06,
+      "loss": 0.8803,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.771759628664673,
+      "learning_rate": 9.95299303844099e-06,
+      "loss": 0.8841,
+      "step": 1131
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4439826822800206,
+      "learning_rate": 9.952851139974468e-06,
+      "loss": 0.7061,
+      "step": 1132
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7009212918737677,
+      "learning_rate": 9.952709028672902e-06,
+      "loss": 0.8601,
+      "step": 1133
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8081590330370654,
+      "learning_rate": 9.9525667045424e-06,
+      "loss": 0.7403,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.146917577172074,
+      "learning_rate": 9.952424167589076e-06,
+      "loss": 0.5944,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0813653977620827,
+      "learning_rate": 9.952281417819057e-06,
+      "loss": 0.7373,
+      "step": 1136
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.797810019959103,
+      "learning_rate": 9.952138455238474e-06,
+      "loss": 0.9108,
+      "step": 1137
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.975515580222026,
+      "learning_rate": 9.951995279853472e-06,
+      "loss": 0.7592,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8223189465772147,
+      "learning_rate": 9.951851891670206e-06,
+      "loss": 0.883,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7113094387371202,
+      "learning_rate": 9.951708290694837e-06,
+      "loss": 0.8312,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.733067214484412,
+      "learning_rate": 9.951564476933534e-06,
+      "loss": 0.762,
+      "step": 1141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.970359330783292,
+      "learning_rate": 9.951420450392476e-06,
+      "loss": 0.9173,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.118603015043401,
+      "learning_rate": 9.951276211077854e-06,
+      "loss": 0.9146,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.180785630099805,
+      "learning_rate": 9.951131758995866e-06,
+      "loss": 0.8274,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7056730071413257,
+      "learning_rate": 9.950987094152721e-06,
+      "loss": 0.8062,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.200052130336348,
+      "learning_rate": 9.950842216554633e-06,
+      "loss": 0.7999,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7332646410340071,
+      "learning_rate": 9.95069712620783e-06,
+      "loss": 0.8426,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8893758943328447,
+      "learning_rate": 9.950551823118544e-06,
+      "loss": 0.8589,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8856055521545518,
+      "learning_rate": 9.950406307293023e-06,
+      "loss": 0.8924,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2452835637817152,
+      "learning_rate": 9.950260578737517e-06,
+      "loss": 0.9407,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.877620774421519,
+      "learning_rate": 9.950114637458288e-06,
+      "loss": 0.8561,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.129084836637377,
+      "learning_rate": 9.949968483461609e-06,
+      "loss": 0.6333,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5846154889326585,
+      "learning_rate": 9.94982211675376e-06,
+      "loss": 0.8063,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6807363057527573,
+      "learning_rate": 9.949675537341031e-06,
+      "loss": 0.8265,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0045633365576143,
+      "learning_rate": 9.949528745229721e-06,
+      "loss": 0.9627,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8740163507759398,
+      "learning_rate": 9.949381740426139e-06,
+      "loss": 0.8096,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8456857718961026,
+      "learning_rate": 9.949234522936597e-06,
+      "loss": 0.7509,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2401797625588498,
+      "learning_rate": 9.949087092767428e-06,
+      "loss": 0.8696,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2790437176751823,
+      "learning_rate": 9.948939449924964e-06,
+      "loss": 0.8176,
+      "step": 1159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9070320135417569,
+      "learning_rate": 9.94879159441555e-06,
+      "loss": 0.7913,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8214681099865917,
+      "learning_rate": 9.948643526245538e-06,
+      "loss": 0.7996,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7286822807603723,
+      "learning_rate": 9.948495245421294e-06,
+      "loss": 0.8296,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8176000011691316,
+      "learning_rate": 9.948346751949188e-06,
+      "loss": 0.8132,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1964874899729496,
+      "learning_rate": 9.948198045835601e-06,
+      "loss": 0.8135,
+      "step": 1164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7057192029972899,
+      "learning_rate": 9.948049127086925e-06,
+      "loss": 0.7496,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7808535552315217,
+      "learning_rate": 9.947899995709556e-06,
+      "loss": 0.8763,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6894598274179446,
+      "learning_rate": 9.947750651709907e-06,
+      "loss": 0.8904,
+      "step": 1167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6622431880238475,
+      "learning_rate": 9.947601095094392e-06,
+      "loss": 0.7801,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8478661834939902,
+      "learning_rate": 9.94745132586944e-06,
+      "loss": 0.8345,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9378740043109548,
+      "learning_rate": 9.947301344041485e-06,
+      "loss": 0.7962,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0870422523955563,
+      "learning_rate": 9.947151149616972e-06,
+      "loss": 0.952,
+      "step": 1171
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9903449125026103,
+      "learning_rate": 9.947000742602358e-06,
+      "loss": 0.9323,
+      "step": 1172
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6952214422155676,
+      "learning_rate": 9.946850123004103e-06,
+      "loss": 0.7773,
+      "step": 1173
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8570962445308332,
+      "learning_rate": 9.946699290828683e-06,
+      "loss": 0.9441,
+      "step": 1174
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.04431403930665,
+      "learning_rate": 9.946548246082577e-06,
+      "loss": 0.8767,
+      "step": 1175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.736322344657312,
+      "learning_rate": 9.946396988772275e-06,
+      "loss": 0.9322,
+      "step": 1176
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7231537946147846,
+      "learning_rate": 9.94624551890428e-06,
+      "loss": 0.8282,
+      "step": 1177
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4685161654494967,
+      "learning_rate": 9.946093836485097e-06,
+      "loss": 0.6456,
+      "step": 1178
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.373178295261869,
+      "learning_rate": 9.94594194152125e-06,
+      "loss": 0.9599,
+      "step": 1179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.857510184109176,
+      "learning_rate": 9.945789834019259e-06,
+      "loss": 0.8764,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8495171695956671,
+      "learning_rate": 9.945637513985664e-06,
+      "loss": 0.7284,
+      "step": 1181
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.027897319440806,
+      "learning_rate": 9.945484981427011e-06,
+      "loss": 0.8141,
+      "step": 1182
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1638972328347097,
+      "learning_rate": 9.945332236349857e-06,
+      "loss": 0.8469,
+      "step": 1183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8884818397656877,
+      "learning_rate": 9.945179278760759e-06,
+      "loss": 0.8208,
+      "step": 1184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8780897756059232,
+      "learning_rate": 9.945026108666296e-06,
+      "loss": 0.9119,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6691407319813298,
+      "learning_rate": 9.944872726073048e-06,
+      "loss": 0.7763,
+      "step": 1186
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.650095961580117,
+      "learning_rate": 9.944719130987606e-06,
+      "loss": 0.7931,
+      "step": 1187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1056222448065713,
+      "learning_rate": 9.94456532341657e-06,
+      "loss": 0.8376,
+      "step": 1188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.804883942945445,
+      "learning_rate": 9.94441130336655e-06,
+      "loss": 0.8838,
+      "step": 1189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6630992815579686,
+      "learning_rate": 9.944257070844165e-06,
+      "loss": 0.7796,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.042904426908836,
+      "learning_rate": 9.944102625856042e-06,
+      "loss": 0.7466,
+      "step": 1191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8217490260921263,
+      "learning_rate": 9.943947968408816e-06,
+      "loss": 0.8671,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7606584926055688,
+      "learning_rate": 9.943793098509138e-06,
+      "loss": 0.8206,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.270857794789571,
+      "learning_rate": 9.943638016163658e-06,
+      "loss": 0.7296,
+      "step": 1194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.579142821708759,
+      "learning_rate": 9.943482721379045e-06,
+      "loss": 0.6859,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.771883907319787,
+      "learning_rate": 9.943327214161968e-06,
+      "loss": 0.8563,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7105220887921968,
+      "learning_rate": 9.943171494519111e-06,
+      "loss": 0.7886,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8518241891216782,
+      "learning_rate": 9.943015562457166e-06,
+      "loss": 0.7186,
+      "step": 1198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.697858318099271,
+      "learning_rate": 9.942859417982833e-06,
+      "loss": 0.8829,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6549476699032168,
+      "learning_rate": 9.942703061102823e-06,
+      "loss": 0.8535,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7407702554525897,
+      "learning_rate": 9.942546491823856e-06,
+      "loss": 0.8417,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3125545757382415,
+      "learning_rate": 9.942389710152657e-06,
+      "loss": 0.7979,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8572370239169624,
+      "learning_rate": 9.942232716095965e-06,
+      "loss": 0.7865,
+      "step": 1203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9019648688410253,
+      "learning_rate": 9.942075509660527e-06,
+      "loss": 0.7853,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1651045448186808,
+      "learning_rate": 9.941918090853096e-06,
+      "loss": 0.8186,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4665144324870583,
+      "learning_rate": 9.941760459680441e-06,
+      "loss": 0.8985,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.888355233529363,
+      "learning_rate": 9.94160261614933e-06,
+      "loss": 0.8449,
+      "step": 1207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4643800697706622,
+      "learning_rate": 9.941444560266551e-06,
+      "loss": 0.745,
+      "step": 1208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6347786074708148,
+      "learning_rate": 9.941286292038894e-06,
+      "loss": 0.7816,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7472102885222045,
+      "learning_rate": 9.941127811473162e-06,
+      "loss": 0.9218,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7325372773457748,
+      "learning_rate": 9.94096911857616e-06,
+      "loss": 0.8514,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6067460582156685,
+      "learning_rate": 9.940810213354713e-06,
+      "loss": 0.8279,
+      "step": 1212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.059609612729524,
+      "learning_rate": 9.940651095815646e-06,
+      "loss": 0.8126,
+      "step": 1213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8334769026198419,
+      "learning_rate": 9.940491765965798e-06,
+      "loss": 0.8059,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7281942990278178,
+      "learning_rate": 9.940332223812017e-06,
+      "loss": 0.7246,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9318934483027073,
+      "learning_rate": 9.940172469361158e-06,
+      "loss": 0.9159,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8468889559021218,
+      "learning_rate": 9.940012502620085e-06,
+      "loss": 0.8069,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1670939686222483,
+      "learning_rate": 9.939852323595671e-06,
+      "loss": 0.6638,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6440016421637116,
+      "learning_rate": 9.939691932294804e-06,
+      "loss": 0.8148,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.728244478924005,
+      "learning_rate": 9.939531328724371e-06,
+      "loss": 0.8515,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3910510021089197,
+      "learning_rate": 9.939370512891277e-06,
+      "loss": 0.8846,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.304764878860346,
+      "learning_rate": 9.93920948480243e-06,
+      "loss": 0.9542,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.321939585265632,
+      "learning_rate": 9.939048244464753e-06,
+      "loss": 0.6852,
+      "step": 1223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7757123804078412,
+      "learning_rate": 9.938886791885172e-06,
+      "loss": 0.7933,
+      "step": 1224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.884231193929429,
+      "learning_rate": 9.938725127070628e-06,
+      "loss": 1.0089,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.285358765525892,
+      "learning_rate": 9.938563250028063e-06,
+      "loss": 0.9035,
+      "step": 1226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5720588134151878,
+      "learning_rate": 9.93840116076444e-06,
+      "loss": 0.8588,
+      "step": 1227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1146453107534917,
+      "learning_rate": 9.938238859286719e-06,
+      "loss": 0.8266,
+      "step": 1228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1208024460045727,
+      "learning_rate": 9.938076345601875e-06,
+      "loss": 0.7499,
+      "step": 1229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5084893318083537,
+      "learning_rate": 9.937913619716895e-06,
+      "loss": 0.8264,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.053285915287363,
+      "learning_rate": 9.937750681638766e-06,
+      "loss": 0.8128,
+      "step": 1231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.752569629158935,
+      "learning_rate": 9.937587531374497e-06,
+      "loss": 0.9078,
+      "step": 1232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8502304307281303,
+      "learning_rate": 9.937424168931094e-06,
+      "loss": 0.8633,
+      "step": 1233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.648908311308415,
+      "learning_rate": 9.937260594315578e-06,
+      "loss": 0.8152,
+      "step": 1234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.799125025800523,
+      "learning_rate": 9.937096807534978e-06,
+      "loss": 0.8039,
+      "step": 1235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7792300446778797,
+      "learning_rate": 9.936932808596332e-06,
+      "loss": 0.8478,
+      "step": 1236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.755169766969406,
+      "learning_rate": 9.936768597506689e-06,
+      "loss": 0.8906,
+      "step": 1237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9792252676389148,
+      "learning_rate": 9.936604174273105e-06,
+      "loss": 0.8387,
+      "step": 1238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9477325467228723,
+      "learning_rate": 9.936439538902644e-06,
+      "loss": 0.8299,
+      "step": 1239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0929654793215033,
+      "learning_rate": 9.936274691402384e-06,
+      "loss": 0.8991,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0552432216488576,
+      "learning_rate": 9.936109631779404e-06,
+      "loss": 0.6721,
+      "step": 1241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.299207432071052,
+      "learning_rate": 9.935944360040803e-06,
+      "loss": 0.9499,
+      "step": 1242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.054947384153624,
+      "learning_rate": 9.935778876193678e-06,
+      "loss": 0.843,
+      "step": 1243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9355174124932324,
+      "learning_rate": 9.935613180245143e-06,
+      "loss": 0.8219,
+      "step": 1244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7090922107185544,
+      "learning_rate": 9.935447272202317e-06,
+      "loss": 0.9498,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8978480664770025,
+      "learning_rate": 9.935281152072329e-06,
+      "loss": 0.8777,
+      "step": 1246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3033882188567416,
+      "learning_rate": 9.93511481986232e-06,
+      "loss": 0.6237,
+      "step": 1247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9882473509517713,
+      "learning_rate": 9.934948275579435e-06,
+      "loss": 0.9542,
+      "step": 1248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8918630683289377,
+      "learning_rate": 9.934781519230832e-06,
+      "loss": 0.8817,
+      "step": 1249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8849308053141147,
+      "learning_rate": 9.934614550823679e-06,
+      "loss": 0.9621,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.471858306126375,
+      "learning_rate": 9.934447370365146e-06,
+      "loss": 0.7725,
+      "step": 1251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.55361551134109,
+      "learning_rate": 9.93427997786242e-06,
+      "loss": 0.7533,
+      "step": 1252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9482958792816842,
+      "learning_rate": 9.934112373322695e-06,
+      "loss": 0.8534,
+      "step": 1253
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7236764821017587,
+      "learning_rate": 9.933944556753173e-06,
+      "loss": 0.8065,
+      "step": 1254
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9850134918410502,
+      "learning_rate": 9.933776528161063e-06,
+      "loss": 0.9414,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8887481832740134,
+      "learning_rate": 9.933608287553589e-06,
+      "loss": 0.9431,
+      "step": 1256
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9729325884054443,
+      "learning_rate": 9.93343983493798e-06,
+      "loss": 0.8487,
+      "step": 1257
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0852388301934957,
+      "learning_rate": 9.93327117032147e-06,
+      "loss": 0.8613,
+      "step": 1258
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.033902025628635,
+      "learning_rate": 9.933102293711314e-06,
+      "loss": 0.9317,
+      "step": 1259
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6208211591458714,
+      "learning_rate": 9.932933205114766e-06,
+      "loss": 0.8079,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.849300730038279,
+      "learning_rate": 9.932763904539091e-06,
+      "loss": 0.8519,
+      "step": 1261
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7792151473330329,
+      "learning_rate": 9.932594391991566e-06,
+      "loss": 0.8792,
+      "step": 1262
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.836139910922359,
+      "learning_rate": 9.932424667479475e-06,
+      "loss": 0.8668,
+      "step": 1263
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6593235553132566,
+      "learning_rate": 9.932254731010108e-06,
+      "loss": 0.8739,
+      "step": 1264
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.814073524516948,
+      "learning_rate": 9.932084582590773e-06,
+      "loss": 0.8598,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4147875010476154,
+      "learning_rate": 9.93191422222878e-06,
+      "loss": 0.8235,
+      "step": 1266
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9226058002225437,
+      "learning_rate": 9.931743649931446e-06,
+      "loss": 0.9068,
+      "step": 1267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9721608202363181,
+      "learning_rate": 9.931572865706105e-06,
+      "loss": 0.8422,
+      "step": 1268
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6294496560513423,
+      "learning_rate": 9.931401869560096e-06,
+      "loss": 0.8164,
+      "step": 1269
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.769300393772311,
+      "learning_rate": 9.931230661500764e-06,
+      "loss": 0.8474,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7732922883180933,
+      "learning_rate": 9.93105924153547e-06,
+      "loss": 0.8004,
+      "step": 1271
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8743997298654085,
+      "learning_rate": 9.930887609671577e-06,
+      "loss": 0.8516,
+      "step": 1272
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.492070963441159,
+      "learning_rate": 9.930715765916462e-06,
+      "loss": 0.6794,
+      "step": 1273
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7431794988119353,
+      "learning_rate": 9.93054371027751e-06,
+      "loss": 0.8803,
+      "step": 1274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9472870261270554,
+      "learning_rate": 9.930371442762112e-06,
+      "loss": 0.9055,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.859992132759118,
+      "learning_rate": 9.930198963377672e-06,
+      "loss": 0.8112,
+      "step": 1276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1916452429156723,
+      "learning_rate": 9.930026272131605e-06,
+      "loss": 0.862,
+      "step": 1277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.684656613896876,
+      "learning_rate": 9.929853369031328e-06,
+      "loss": 0.7656,
+      "step": 1278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0791926974582267,
+      "learning_rate": 9.929680254084273e-06,
+      "loss": 0.8238,
+      "step": 1279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5922331486613546,
+      "learning_rate": 9.929506927297878e-06,
+      "loss": 0.7049,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6832788827192633,
+      "learning_rate": 9.929333388679593e-06,
+      "loss": 0.848,
+      "step": 1281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6643653664677325,
+      "learning_rate": 9.929159638236872e-06,
+      "loss": 0.8974,
+      "step": 1282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2868247835937647,
+      "learning_rate": 9.928985675977185e-06,
+      "loss": 0.8876,
+      "step": 1283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2365035684590604,
+      "learning_rate": 9.928811501908006e-06,
+      "loss": 0.7231,
+      "step": 1284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8614786277880297,
+      "learning_rate": 9.928637116036818e-06,
+      "loss": 0.8535,
+      "step": 1285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1606113690378461,
+      "learning_rate": 9.928462518371119e-06,
+      "loss": 0.6827,
+      "step": 1286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8565953262968755,
+      "learning_rate": 9.928287708918408e-06,
+      "loss": 0.8657,
+      "step": 1287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9125182502507692,
+      "learning_rate": 9.928112687686197e-06,
+      "loss": 0.8379,
+      "step": 1288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.279248047811361,
+      "learning_rate": 9.92793745468201e-06,
+      "loss": 0.8974,
+      "step": 1289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1832995819301844,
+      "learning_rate": 9.927762009913377e-06,
+      "loss": 0.6873,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1193848747363337,
+      "learning_rate": 9.927586353387832e-06,
+      "loss": 0.8778,
+      "step": 1291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0681169645754134,
+      "learning_rate": 9.92741048511293e-06,
+      "loss": 0.6356,
+      "step": 1292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0456664109231468,
+      "learning_rate": 9.927234405096224e-06,
+      "loss": 0.6536,
+      "step": 1293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.062040044868049,
+      "learning_rate": 9.927058113345282e-06,
+      "loss": 0.8273,
+      "step": 1294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1752117881463104,
+      "learning_rate": 9.92688160986768e-06,
+      "loss": 0.8481,
+      "step": 1295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.157006766708858,
+      "learning_rate": 9.926704894671004e-06,
+      "loss": 0.8285,
+      "step": 1296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9188902946491246,
+      "learning_rate": 9.926527967762845e-06,
+      "loss": 0.8482,
+      "step": 1297
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.483324452733649,
+      "learning_rate": 9.926350829150806e-06,
+      "loss": 0.8093,
+      "step": 1298
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8463512118920504,
+      "learning_rate": 9.926173478842502e-06,
+      "loss": 0.9543,
+      "step": 1299
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8005451370839682,
+      "learning_rate": 9.925995916845553e-06,
+      "loss": 0.8163,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9302808364466162,
+      "learning_rate": 9.925818143167587e-06,
+      "loss": 0.919,
+      "step": 1301
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0664406789779135,
+      "learning_rate": 9.925640157816246e-06,
+      "loss": 0.842,
+      "step": 1302
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0931305088495855,
+      "learning_rate": 9.925461960799177e-06,
+      "loss": 0.833,
+      "step": 1303
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9729422132277379,
+      "learning_rate": 9.925283552124039e-06,
+      "loss": 0.7876,
+      "step": 1304
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7518797875145316,
+      "learning_rate": 9.925104931798496e-06,
+      "loss": 0.9063,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.025318091207668,
+      "learning_rate": 9.924926099830226e-06,
+      "loss": 0.8486,
+      "step": 1306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.189873548371559,
+      "learning_rate": 9.924747056226913e-06,
+      "loss": 0.9113,
+      "step": 1307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7826282975837797,
+      "learning_rate": 9.92456780099625e-06,
+      "loss": 0.8505,
+      "step": 1308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8789150078374255,
+      "learning_rate": 9.924388334145943e-06,
+      "loss": 0.8273,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.081822595666377,
+      "learning_rate": 9.9242086556837e-06,
+      "loss": 0.9072,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5710206917290088,
+      "learning_rate": 9.924028765617247e-06,
+      "loss": 0.7621,
+      "step": 1311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1840452087852706,
+      "learning_rate": 9.92384866395431e-06,
+      "loss": 0.9243,
+      "step": 1312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9825933668818028,
+      "learning_rate": 9.923668350702628e-06,
+      "loss": 0.9101,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9329435531201478,
+      "learning_rate": 9.923487825869955e-06,
+      "loss": 0.8553,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0690181629275273,
+      "learning_rate": 9.923307089464043e-06,
+      "loss": 0.7795,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7378320377744878,
+      "learning_rate": 9.92312614149266e-06,
+      "loss": 0.7238,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.128910200673605,
+      "learning_rate": 9.922944981963583e-06,
+      "loss": 0.9089,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.022772055037916,
+      "learning_rate": 9.922763610884596e-06,
+      "loss": 0.7496,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9337504610412255,
+      "learning_rate": 9.922582028263495e-06,
+      "loss": 0.8572,
+      "step": 1319
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.471987987332869,
+      "learning_rate": 9.922400234108079e-06,
+      "loss": 0.8489,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3434664325085575,
+      "learning_rate": 9.922218228426162e-06,
+      "loss": 0.6541,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8403334207229296,
+      "learning_rate": 9.922036011225564e-06,
+      "loss": 0.8544,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6739998424782703,
+      "learning_rate": 9.92185358251412e-06,
+      "loss": 0.8176,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2318168181958007,
+      "learning_rate": 9.921670942299664e-06,
+      "loss": 0.6997,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8166334230216417,
+      "learning_rate": 9.921488090590047e-06,
+      "loss": 0.9122,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1475332624846812,
+      "learning_rate": 9.921305027393125e-06,
+      "loss": 0.8464,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9549319718086786,
+      "learning_rate": 9.921121752716767e-06,
+      "loss": 0.8515,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1738536747138344,
+      "learning_rate": 9.920938266568847e-06,
+      "loss": 0.8476,
+      "step": 1328
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7240807412496548,
+      "learning_rate": 9.92075456895725e-06,
+      "loss": 0.7595,
+      "step": 1329
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8189657867840243,
+      "learning_rate": 9.92057065988987e-06,
+      "loss": 0.8826,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7354464822065723,
+      "learning_rate": 9.92038653937461e-06,
+      "loss": 0.834,
+      "step": 1331
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1953914193237694,
+      "learning_rate": 9.920202207419382e-06,
+      "loss": 0.8961,
+      "step": 1332
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5621840513148992,
+      "learning_rate": 9.920017664032109e-06,
+      "loss": 0.7082,
+      "step": 1333
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7435248553104554,
+      "learning_rate": 9.919832909220717e-06,
+      "loss": 0.8838,
+      "step": 1334
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2497683552278884,
+      "learning_rate": 9.91964794299315e-06,
+      "loss": 0.7359,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0385408941554655,
+      "learning_rate": 9.919462765357353e-06,
+      "loss": 0.8389,
+      "step": 1336
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9620511668116878,
+      "learning_rate": 9.919277376321284e-06,
+      "loss": 0.7981,
+      "step": 1337
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6402121372631697,
+      "learning_rate": 9.91909177589291e-06,
+      "loss": 0.9225,
+      "step": 1338
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.953388511635238,
+      "learning_rate": 9.91890596408021e-06,
+      "loss": 0.7499,
+      "step": 1339
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.827034445269916,
+      "learning_rate": 9.918719940891165e-06,
+      "loss": 0.8091,
+      "step": 1340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8948858272949036,
+      "learning_rate": 9.918533706333767e-06,
+      "loss": 0.8756,
+      "step": 1341
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7824607866696955,
+      "learning_rate": 9.918347260416024e-06,
+      "loss": 0.8134,
+      "step": 1342
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9400675957168045,
+      "learning_rate": 9.918160603145943e-06,
+      "loss": 0.9399,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8775540668019939,
+      "learning_rate": 9.917973734531549e-06,
+      "loss": 0.7939,
+      "step": 1344
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1007913520111436,
+      "learning_rate": 9.917786654580872e-06,
+      "loss": 0.5854,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7623818106315514,
+      "learning_rate": 9.917599363301947e-06,
+      "loss": 0.7997,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9709117965301681,
+      "learning_rate": 9.917411860702828e-06,
+      "loss": 0.752,
+      "step": 1347
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9947218724338465,
+      "learning_rate": 9.917224146791568e-06,
+      "loss": 0.9064,
+      "step": 1348
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8936611707807964,
+      "learning_rate": 9.917036221576235e-06,
+      "loss": 0.7787,
+      "step": 1349
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8525264778789023,
+      "learning_rate": 9.916848085064906e-06,
+      "loss": 0.8861,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2491377478309393,
+      "learning_rate": 9.916659737265664e-06,
+      "loss": 0.782,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8892858492092695,
+      "learning_rate": 9.916471178186602e-06,
+      "loss": 0.8166,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7773557788564074,
+      "learning_rate": 9.916282407835826e-06,
+      "loss": 0.7816,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6769365985434508,
+      "learning_rate": 9.916093426221445e-06,
+      "loss": 0.79,
+      "step": 1354
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.255998927526379,
+      "learning_rate": 9.91590423335158e-06,
+      "loss": 0.8015,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9966863549711815,
+      "learning_rate": 9.915714829234361e-06,
+      "loss": 0.9219,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7710913952841285,
+      "learning_rate": 9.91552521387793e-06,
+      "loss": 0.797,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9282160177262861,
+      "learning_rate": 9.915335387290432e-06,
+      "loss": 0.9372,
+      "step": 1358
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6393685567909435,
+      "learning_rate": 9.915145349480027e-06,
+      "loss": 0.7718,
+      "step": 1359
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9046638213014757,
+      "learning_rate": 9.914955100454878e-06,
+      "loss": 0.8652,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1197306597342835,
+      "learning_rate": 9.914764640223162e-06,
+      "loss": 0.6563,
+      "step": 1361
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7870514599876353,
+      "learning_rate": 9.914573968793064e-06,
+      "loss": 0.8482,
+      "step": 1362
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5206735623734042,
+      "learning_rate": 9.914383086172778e-06,
+      "loss": 0.7891,
+      "step": 1363
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9835048639772361,
+      "learning_rate": 9.914191992370504e-06,
+      "loss": 0.8079,
+      "step": 1364
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5821898497504496,
+      "learning_rate": 9.914000687394457e-06,
+      "loss": 0.6998,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.051010074107109,
+      "learning_rate": 9.913809171252856e-06,
+      "loss": 0.9633,
+      "step": 1366
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6268259900585307,
+      "learning_rate": 9.913617443953931e-06,
+      "loss": 0.7997,
+      "step": 1367
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3405791950881238,
+      "learning_rate": 9.913425505505923e-06,
+      "loss": 0.7514,
+      "step": 1368
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8674255013889063,
+      "learning_rate": 9.913233355917075e-06,
+      "loss": 0.9508,
+      "step": 1369
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8565802986408513,
+      "learning_rate": 9.913040995195651e-06,
+      "loss": 0.7681,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6231354050121172,
+      "learning_rate": 9.91284842334991e-06,
+      "loss": 0.8517,
+      "step": 1371
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.753742928916686,
+      "learning_rate": 9.912655640388134e-06,
+      "loss": 0.7325,
+      "step": 1372
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0120485108787791,
+      "learning_rate": 9.912462646318602e-06,
+      "loss": 0.6905,
+      "step": 1373
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5315349419539366,
+      "learning_rate": 9.91226944114961e-06,
+      "loss": 0.8219,
+      "step": 1374
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9343763089687365,
+      "learning_rate": 9.91207602488946e-06,
+      "loss": 0.7702,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7499383526132486,
+      "learning_rate": 9.911882397546464e-06,
+      "loss": 0.7754,
+      "step": 1376
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.299170557081601,
+      "learning_rate": 9.91168855912894e-06,
+      "loss": 0.6902,
+      "step": 1377
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8477223092612163,
+      "learning_rate": 9.911494509645222e-06,
+      "loss": 0.8829,
+      "step": 1378
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6801660442999375,
+      "learning_rate": 9.911300249103646e-06,
+      "loss": 0.8275,
+      "step": 1379
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9297591725646084,
+      "learning_rate": 9.91110577751256e-06,
+      "loss": 0.8944,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6445381394149834,
+      "learning_rate": 9.91091109488032e-06,
+      "loss": 0.8665,
+      "step": 1381
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6991856909893128,
+      "learning_rate": 9.910716201215294e-06,
+      "loss": 0.8854,
+      "step": 1382
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.92477116297879,
+      "learning_rate": 9.910521096525856e-06,
+      "loss": 0.6908,
+      "step": 1383
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6930185691859423,
+      "learning_rate": 9.910325780820391e-06,
+      "loss": 0.7996,
+      "step": 1384
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6899399245900577,
+      "learning_rate": 9.910130254107288e-06,
+      "loss": 0.8224,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5719281477112372,
+      "learning_rate": 9.909934516394957e-06,
+      "loss": 0.799,
+      "step": 1386
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2308153362535634,
+      "learning_rate": 9.909738567691803e-06,
+      "loss": 0.9314,
+      "step": 1387
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.105360896145239,
+      "learning_rate": 9.909542408006247e-06,
+      "loss": 0.8906,
+      "step": 1388
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0355339064616347,
+      "learning_rate": 9.90934603734672e-06,
+      "loss": 0.8173,
+      "step": 1389
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5918560095692735,
+      "learning_rate": 9.909149455721661e-06,
+      "loss": 0.7735,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1888569980743129,
+      "learning_rate": 9.908952663139518e-06,
+      "loss": 0.7021,
+      "step": 1391
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7605802014275833,
+      "learning_rate": 9.908755659608743e-06,
+      "loss": 0.9161,
+      "step": 1392
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5268835945050288,
+      "learning_rate": 9.908558445137807e-06,
+      "loss": 0.7462,
+      "step": 1393
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.050862309956157,
+      "learning_rate": 9.908361019735181e-06,
+      "loss": 0.8852,
+      "step": 1394
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7767128984218867,
+      "learning_rate": 9.908163383409353e-06,
+      "loss": 0.815,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.647509787776577,
+      "learning_rate": 9.907965536168811e-06,
+      "loss": 0.9333,
+      "step": 1396
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8877944858630722,
+      "learning_rate": 9.907767478022061e-06,
+      "loss": 0.8939,
+      "step": 1397
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7094078159615624,
+      "learning_rate": 9.90756920897761e-06,
+      "loss": 0.7934,
+      "step": 1398
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.261084757681306,
+      "learning_rate": 9.907370729043984e-06,
+      "loss": 0.7926,
+      "step": 1399
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8529337511840747,
+      "learning_rate": 9.907172038229706e-06,
+      "loss": 0.7545,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6323342413527124,
+      "learning_rate": 9.906973136543317e-06,
+      "loss": 0.7048,
+      "step": 1401
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.821228163959321,
+      "learning_rate": 9.906774023993362e-06,
+      "loss": 0.7556,
+      "step": 1402
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0631907658977724,
+      "learning_rate": 9.906574700588403e-06,
+      "loss": 0.645,
+      "step": 1403
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8795825642939965,
+      "learning_rate": 9.906375166336998e-06,
+      "loss": 0.8838,
+      "step": 1404
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7646580635803575,
+      "learning_rate": 9.906175421247728e-06,
+      "loss": 0.8648,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0806098191610953,
+      "learning_rate": 9.905975465329172e-06,
+      "loss": 0.742,
+      "step": 1406
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.664232043965964,
+      "learning_rate": 9.905775298589923e-06,
+      "loss": 0.7908,
+      "step": 1407
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9922589185059985,
+      "learning_rate": 9.905574921038586e-06,
+      "loss": 0.8113,
+      "step": 1408
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8765485219848985,
+      "learning_rate": 9.905374332683768e-06,
+      "loss": 0.6792,
+      "step": 1409
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9922915163428547,
+      "learning_rate": 9.90517353353409e-06,
+      "loss": 0.8945,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0193731310423377,
+      "learning_rate": 9.904972523598183e-06,
+      "loss": 0.7579,
+      "step": 1411
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2783861053966215,
+      "learning_rate": 9.90477130288468e-06,
+      "loss": 0.8142,
+      "step": 1412
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7378081781648096,
+      "learning_rate": 9.904569871402233e-06,
+      "loss": 0.9451,
+      "step": 1413
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8597008561234478,
+      "learning_rate": 9.904368229159494e-06,
+      "loss": 0.8693,
+      "step": 1414
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6323468359510072,
+      "learning_rate": 9.904166376165129e-06,
+      "loss": 0.8939,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1727995751462805,
+      "learning_rate": 9.903964312427813e-06,
+      "loss": 0.7231,
+      "step": 1416
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0143167896166303,
+      "learning_rate": 9.903762037956228e-06,
+      "loss": 0.7765,
+      "step": 1417
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9423068462580177,
+      "learning_rate": 9.903559552759069e-06,
+      "loss": 0.8122,
+      "step": 1418
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2378016077651521,
+      "learning_rate": 9.903356856845035e-06,
+      "loss": 0.6664,
+      "step": 1419
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0724244821578033,
+      "learning_rate": 9.903153950222834e-06,
+      "loss": 0.8974,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9031303904866814,
+      "learning_rate": 9.90295083290119e-06,
+      "loss": 0.8661,
+      "step": 1421
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2048185349604812,
+      "learning_rate": 9.90274750488883e-06,
+      "loss": 0.6942,
+      "step": 1422
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.814747231290849,
+      "learning_rate": 9.902543966194489e-06,
+      "loss": 0.7677,
+      "step": 1423
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6860708513426574,
+      "learning_rate": 9.902340216826915e-06,
+      "loss": 0.783,
+      "step": 1424
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8811567486578284,
+      "learning_rate": 9.902136256794865e-06,
+      "loss": 0.9112,
+      "step": 1425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8384834224378597,
+      "learning_rate": 9.901932086107103e-06,
+      "loss": 0.7274,
+      "step": 1426
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6774185723838744,
+      "learning_rate": 9.901727704772401e-06,
+      "loss": 0.8254,
+      "step": 1427
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0418331356226624,
+      "learning_rate": 9.901523112799543e-06,
+      "loss": 0.8591,
+      "step": 1428
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7983556508603458,
+      "learning_rate": 9.90131831019732e-06,
+      "loss": 0.8397,
+      "step": 1429
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7344591310023798,
+      "learning_rate": 9.901113296974533e-06,
+      "loss": 0.6868,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9739026632570376,
+      "learning_rate": 9.900908073139995e-06,
+      "loss": 0.9957,
+      "step": 1431
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.56115341495953,
+      "learning_rate": 9.90070263870252e-06,
+      "loss": 0.8218,
+      "step": 1432
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6669113013292567,
+      "learning_rate": 9.900496993670937e-06,
+      "loss": 0.7957,
+      "step": 1433
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9779613266100036,
+      "learning_rate": 9.900291138054086e-06,
+      "loss": 0.9809,
+      "step": 1434
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8954833196730778,
+      "learning_rate": 9.90008507186081e-06,
+      "loss": 0.7572,
+      "step": 1435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7212482073373874,
+      "learning_rate": 9.899878795099966e-06,
+      "loss": 0.8517,
+      "step": 1436
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8369062490568675,
+      "learning_rate": 9.899672307780419e-06,
+      "loss": 0.9275,
+      "step": 1437
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8192145645134121,
+      "learning_rate": 9.89946560991104e-06,
+      "loss": 0.833,
+      "step": 1438
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7612066714836148,
+      "learning_rate": 9.899258701500712e-06,
+      "loss": 0.9038,
+      "step": 1439
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7274541842801079,
+      "learning_rate": 9.899051582558325e-06,
+      "loss": 0.7225,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9065033325133083,
+      "learning_rate": 9.898844253092782e-06,
+      "loss": 0.878,
+      "step": 1441
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8980104772225064,
+      "learning_rate": 9.898636713112992e-06,
+      "loss": 0.7337,
+      "step": 1442
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8143275623560826,
+      "learning_rate": 9.89842896262787e-06,
+      "loss": 0.8863,
+      "step": 1443
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.138221552568561,
+      "learning_rate": 9.89822100164635e-06,
+      "loss": 0.8174,
+      "step": 1444
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8568096334258604,
+      "learning_rate": 9.898012830177361e-06,
+      "loss": 0.8022,
+      "step": 1445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.818588153136762,
+      "learning_rate": 9.897804448229854e-06,
+      "loss": 0.7504,
+      "step": 1446
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9561340368634106,
+      "learning_rate": 9.89759585581278e-06,
+      "loss": 0.817,
+      "step": 1447
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7757957272088243,
+      "learning_rate": 9.897387052935106e-06,
+      "loss": 0.8045,
+      "step": 1448
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.767528036163758,
+      "learning_rate": 9.897178039605803e-06,
+      "loss": 0.8293,
+      "step": 1449
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0966492997192474,
+      "learning_rate": 9.896968815833853e-06,
+      "loss": 0.9824,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.662614016305948,
+      "learning_rate": 9.896759381628247e-06,
+      "loss": 0.8091,
+      "step": 1451
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7573411580837344,
+      "learning_rate": 9.896549736997985e-06,
+      "loss": 0.8897,
+      "step": 1452
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.917692233621924,
+      "learning_rate": 9.896339881952076e-06,
+      "loss": 0.842,
+      "step": 1453
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.821144146199315,
+      "learning_rate": 9.896129816499535e-06,
+      "loss": 0.836,
+      "step": 1454
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.392357727367577,
+      "learning_rate": 9.895919540649395e-06,
+      "loss": 0.7876,
+      "step": 1455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9932313428606578,
+      "learning_rate": 9.895709054410686e-06,
+      "loss": 0.7121,
+      "step": 1456
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.584969801699579,
+      "learning_rate": 9.895498357792456e-06,
+      "loss": 0.905,
+      "step": 1457
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5755696878133048,
+      "learning_rate": 9.89528745080376e-06,
+      "loss": 0.7352,
+      "step": 1458
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6648219109791047,
+      "learning_rate": 9.89507633345366e-06,
+      "loss": 0.8196,
+      "step": 1459
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1133076420551264,
+      "learning_rate": 9.894865005751227e-06,
+      "loss": 0.6096,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7674696551213906,
+      "learning_rate": 9.894653467705542e-06,
+      "loss": 0.7524,
+      "step": 1461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3395455579143685,
+      "learning_rate": 9.894441719325698e-06,
+      "loss": 0.7849,
+      "step": 1462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7112363319768542,
+      "learning_rate": 9.894229760620793e-06,
+      "loss": 0.8866,
+      "step": 1463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.62856563035511,
+      "learning_rate": 9.894017591599934e-06,
+      "loss": 0.8667,
+      "step": 1464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.682036247516913,
+      "learning_rate": 9.89380521227224e-06,
+      "loss": 0.8932,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.70439278125174,
+      "learning_rate": 9.893592622646838e-06,
+      "loss": 0.7909,
+      "step": 1466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9108908002783742,
+      "learning_rate": 9.893379822732863e-06,
+      "loss": 0.8322,
+      "step": 1467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8554190086882316,
+      "learning_rate": 9.893166812539456e-06,
+      "loss": 0.9379,
+      "step": 1468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8742907533442885,
+      "learning_rate": 9.892953592075776e-06,
+      "loss": 0.7839,
+      "step": 1469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.533645455119599,
+      "learning_rate": 9.892740161350981e-06,
+      "loss": 0.7618,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0309235320493864,
+      "learning_rate": 9.892526520374246e-06,
+      "loss": 0.7137,
+      "step": 1471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6521035656128533,
+      "learning_rate": 9.892312669154749e-06,
+      "loss": 0.7934,
+      "step": 1472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9899723041613826,
+      "learning_rate": 9.892098607701682e-06,
+      "loss": 0.8112,
+      "step": 1473
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8933859860712505,
+      "learning_rate": 9.891884336024242e-06,
+      "loss": 0.8235,
+      "step": 1474
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6607378664281356,
+      "learning_rate": 9.891669854131636e-06,
+      "loss": 0.8556,
+      "step": 1475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.153002962743788,
+      "learning_rate": 9.891455162033085e-06,
+      "loss": 0.846,
+      "step": 1476
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8500620370347718,
+      "learning_rate": 9.891240259737809e-06,
+      "loss": 0.7925,
+      "step": 1477
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9522350909284643,
+      "learning_rate": 9.891025147255047e-06,
+      "loss": 0.812,
+      "step": 1478
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.060447281086279,
+      "learning_rate": 9.890809824594041e-06,
+      "loss": 0.9617,
+      "step": 1479
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9011027016842805,
+      "learning_rate": 9.890594291764044e-06,
+      "loss": 0.7559,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.608835334032645,
+      "learning_rate": 9.89037854877432e-06,
+      "loss": 0.7546,
+      "step": 1481
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7296441347636373,
+      "learning_rate": 9.890162595634138e-06,
+      "loss": 0.8323,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.343134273942839,
+      "learning_rate": 9.889946432352776e-06,
+      "loss": 0.6697,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3026952114627317,
+      "learning_rate": 9.889730058939529e-06,
+      "loss": 0.7915,
+      "step": 1484
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1868469283618894,
+      "learning_rate": 9.889513475403689e-06,
+      "loss": 0.7585,
+      "step": 1485
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7187416350294957,
+      "learning_rate": 9.889296681754565e-06,
+      "loss": 0.7475,
+      "step": 1486
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9384992963500463,
+      "learning_rate": 9.889079678001474e-06,
+      "loss": 0.7914,
+      "step": 1487
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8444901085826855,
+      "learning_rate": 9.88886246415374e-06,
+      "loss": 0.8985,
+      "step": 1488
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8192452325183548,
+      "learning_rate": 9.8886450402207e-06,
+      "loss": 0.8244,
+      "step": 1489
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5001306338644644,
+      "learning_rate": 9.888427406211694e-06,
+      "loss": 0.8325,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7907937128552296,
+      "learning_rate": 9.888209562136074e-06,
+      "loss": 0.8197,
+      "step": 1491
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1472488607343034,
+      "learning_rate": 9.887991508003205e-06,
+      "loss": 0.6541,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5353678359463208,
+      "learning_rate": 9.887773243822454e-06,
+      "loss": 0.7744,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7627174401956225,
+      "learning_rate": 9.8875547696032e-06,
+      "loss": 0.7583,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.133054089887048,
+      "learning_rate": 9.887336085354834e-06,
+      "loss": 0.7364,
+      "step": 1495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2545649424760434,
+      "learning_rate": 9.887117191086749e-06,
+      "loss": 0.6458,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8189946982819316,
+      "learning_rate": 9.886898086808356e-06,
+      "loss": 0.8462,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.076293440618704,
+      "learning_rate": 9.886678772529069e-06,
+      "loss": 0.7277,
+      "step": 1498
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.328621581685798,
+      "learning_rate": 9.88645924825831e-06,
+      "loss": 0.6814,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8627646247556637,
+      "learning_rate": 9.886239514005515e-06,
+      "loss": 0.9209,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7346252731882787,
+      "learning_rate": 9.886019569780128e-06,
+      "loss": 0.8042,
+      "step": 1501
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6402664943971328,
+      "learning_rate": 9.885799415591594e-06,
+      "loss": 0.8671,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7461675325630062,
+      "learning_rate": 9.88557905144938e-06,
+      "loss": 0.8169,
+      "step": 1503
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6513999954261223,
+      "learning_rate": 9.885358477362956e-06,
+      "loss": 0.7608,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1865714580392606,
+      "learning_rate": 9.885137693341795e-06,
+      "loss": 0.743,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7722760757807778,
+      "learning_rate": 9.884916699395389e-06,
+      "loss": 0.7308,
+      "step": 1506
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.941382044454953,
+      "learning_rate": 9.884695495533232e-06,
+      "loss": 0.8767,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8239110594149008,
+      "learning_rate": 9.884474081764833e-06,
+      "loss": 1.0077,
+      "step": 1508
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2881906447995946,
+      "learning_rate": 9.8842524580997e-06,
+      "loss": 0.597,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.527086266148255,
+      "learning_rate": 9.884030624547364e-06,
+      "loss": 0.8175,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7602572891800812,
+      "learning_rate": 9.883808581117355e-06,
+      "loss": 0.8903,
+      "step": 1511
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9820292790084832,
+      "learning_rate": 9.883586327819214e-06,
+      "loss": 0.8318,
+      "step": 1512
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1521934015671196,
+      "learning_rate": 9.883363864662491e-06,
+      "loss": 0.7859,
+      "step": 1513
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1201223556222375,
+      "learning_rate": 9.883141191656748e-06,
+      "loss": 0.7161,
+      "step": 1514
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6720863508543438,
+      "learning_rate": 9.882918308811554e-06,
+      "loss": 0.7507,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9582897871436786,
+      "learning_rate": 9.882695216136485e-06,
+      "loss": 0.7755,
+      "step": 1516
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.725990374047224,
+      "learning_rate": 9.882471913641126e-06,
+      "loss": 0.8191,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8598545880288202,
+      "learning_rate": 9.882248401335076e-06,
+      "loss": 0.9555,
+      "step": 1518
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8115442217471416,
+      "learning_rate": 9.88202467922794e-06,
+      "loss": 0.8386,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6308141137704422,
+      "learning_rate": 9.881800747329328e-06,
+      "loss": 0.7831,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7419835132528072,
+      "learning_rate": 9.881576605648869e-06,
+      "loss": 0.7872,
+      "step": 1521
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3200026790508197,
+      "learning_rate": 9.881352254196189e-06,
+      "loss": 0.8438,
+      "step": 1522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8917432566344892,
+      "learning_rate": 9.881127692980932e-06,
+      "loss": 0.806,
+      "step": 1523
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2913867677852429,
+      "learning_rate": 9.880902922012747e-06,
+      "loss": 0.7051,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7015219144209428,
+      "learning_rate": 9.880677941301293e-06,
+      "loss": 0.879,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5680175146608883,
+      "learning_rate": 9.880452750856239e-06,
+      "loss": 0.8087,
+      "step": 1526
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6279236909528492,
+      "learning_rate": 9.880227350687259e-06,
+      "loss": 0.7523,
+      "step": 1527
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6142204701582539,
+      "learning_rate": 9.880001740804043e-06,
+      "loss": 0.8218,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5382399887429192,
+      "learning_rate": 9.879775921216284e-06,
+      "loss": 0.7968,
+      "step": 1529
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.997095961955695,
+      "learning_rate": 9.879549891933684e-06,
+      "loss": 0.8406,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7611273615517515,
+      "learning_rate": 9.879323652965957e-06,
+      "loss": 0.8234,
+      "step": 1531
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7118886072412993,
+      "learning_rate": 9.879097204322828e-06,
+      "loss": 0.7632,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7154070957181857,
+      "learning_rate": 9.878870546014025e-06,
+      "loss": 0.6397,
+      "step": 1533
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8424141687874778,
+      "learning_rate": 9.87864367804929e-06,
+      "loss": 0.9727,
+      "step": 1534
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9790670492827833,
+      "learning_rate": 9.87841660043837e-06,
+      "loss": 0.672,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4605950440180269,
+      "learning_rate": 9.878189313191023e-06,
+      "loss": 0.7972,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9125360803470726,
+      "learning_rate": 9.877961816317018e-06,
+      "loss": 0.7569,
+      "step": 1537
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9087398499023305,
+      "learning_rate": 9.87773410982613e-06,
+      "loss": 0.8983,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.104265455965984,
+      "learning_rate": 9.877506193728144e-06,
+      "loss": 0.685,
+      "step": 1539
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8663903377524094,
+      "learning_rate": 9.877278068032852e-06,
+      "loss": 0.8346,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7873124989061402,
+      "learning_rate": 9.877049732750062e-06,
+      "loss": 0.8066,
+      "step": 1541
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8666793763013765,
+      "learning_rate": 9.876821187889581e-06,
+      "loss": 0.984,
+      "step": 1542
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8615786647600425,
+      "learning_rate": 9.876592433461234e-06,
+      "loss": 0.8901,
+      "step": 1543
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8111378410878587,
+      "learning_rate": 9.876363469474848e-06,
+      "loss": 0.7414,
+      "step": 1544
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.082931107682366,
+      "learning_rate": 9.876134295940264e-06,
+      "loss": 0.8406,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0628068808456024,
+      "learning_rate": 9.875904912867328e-06,
+      "loss": 0.9154,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.972769680412449,
+      "learning_rate": 9.875675320265903e-06,
+      "loss": 0.8973,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3670779900623944,
+      "learning_rate": 9.875445518145848e-06,
+      "loss": 0.8248,
+      "step": 1548
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.550693655214365,
+      "learning_rate": 9.87521550651704e-06,
+      "loss": 0.7131,
+      "step": 1549
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7209305241215198,
+      "learning_rate": 9.874985285389364e-06,
+      "loss": 0.856,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8524628870633515,
+      "learning_rate": 9.874754854772714e-06,
+      "loss": 0.8511,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8294774945317052,
+      "learning_rate": 9.874524214676993e-06,
+      "loss": 0.8368,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.029685999979293,
+      "learning_rate": 9.874293365112109e-06,
+      "loss": 0.7698,
+      "step": 1553
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6216516621713712,
+      "learning_rate": 9.874062306087983e-06,
+      "loss": 0.7768,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6435463890239816,
+      "learning_rate": 9.873831037614544e-06,
+      "loss": 0.814,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4982606612344096,
+      "learning_rate": 9.873599559701732e-06,
+      "loss": 0.8511,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6355360986665766,
+      "learning_rate": 9.873367872359492e-06,
+      "loss": 0.7484,
+      "step": 1557
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.252368086033878,
+      "learning_rate": 9.873135975597781e-06,
+      "loss": 0.6261,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.560617047964374,
+      "learning_rate": 9.872903869426564e-06,
+      "loss": 0.8134,
+      "step": 1559
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7687686413665376,
+      "learning_rate": 9.872671553855817e-06,
+      "loss": 0.9793,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2244340385042465,
+      "learning_rate": 9.872439028895518e-06,
+      "loss": 0.8381,
+      "step": 1561
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5289434732008473,
+      "learning_rate": 9.872206294555665e-06,
+      "loss": 0.8125,
+      "step": 1562
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3314071778929186,
+      "learning_rate": 9.871973350846255e-06,
+      "loss": 0.7812,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7448551486421322,
+      "learning_rate": 9.8717401977773e-06,
+      "loss": 0.8525,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6672520950955498,
+      "learning_rate": 9.87150683535882e-06,
+      "loss": 0.8295,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4615155218206644,
+      "learning_rate": 9.871273263600842e-06,
+      "loss": 0.7887,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4188039376758548,
+      "learning_rate": 9.871039482513402e-06,
+      "loss": 0.9291,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7468416591483236,
+      "learning_rate": 9.870805492106546e-06,
+      "loss": 0.857,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7066684953184381,
+      "learning_rate": 9.870571292390331e-06,
+      "loss": 0.7909,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.579531738407528,
+      "learning_rate": 9.87033688337482e-06,
+      "loss": 0.7987,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.731949133593372,
+      "learning_rate": 9.870102265070087e-06,
+      "loss": 0.8758,
+      "step": 1571
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1862850603683097,
+      "learning_rate": 9.869867437486214e-06,
+      "loss": 0.7175,
+      "step": 1572
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.057999674703341,
+      "learning_rate": 9.86963240063329e-06,
+      "loss": 0.64,
+      "step": 1573
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0311486533138536,
+      "learning_rate": 9.869397154521418e-06,
+      "loss": 0.5848,
+      "step": 1574
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.612159910107449,
+      "learning_rate": 9.869161699160704e-06,
+      "loss": 0.9297,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8327529298505656,
+      "learning_rate": 9.868926034561268e-06,
+      "loss": 0.7429,
+      "step": 1576
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8382028885636172,
+      "learning_rate": 9.868690160733237e-06,
+      "loss": 0.7866,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.131894346804339,
+      "learning_rate": 9.868454077686746e-06,
+      "loss": 0.8234,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0347043470620414,
+      "learning_rate": 9.868217785431942e-06,
+      "loss": 0.8088,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.716657595601275,
+      "learning_rate": 9.867981283978975e-06,
+      "loss": 0.8185,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7277417997800353,
+      "learning_rate": 9.867744573338012e-06,
+      "loss": 0.9946,
+      "step": 1581
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2119153936981333,
+      "learning_rate": 9.867507653519225e-06,
+      "loss": 0.5793,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8766175804739937,
+      "learning_rate": 9.867270524532792e-06,
+      "loss": 0.872,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.622822067188611,
+      "learning_rate": 9.867033186388906e-06,
+      "loss": 0.8582,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8384484589672807,
+      "learning_rate": 9.866795639097763e-06,
+      "loss": 0.8372,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6617087114797817,
+      "learning_rate": 9.866557882669574e-06,
+      "loss": 0.795,
+      "step": 1586
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.596079842246663,
+      "learning_rate": 9.866319917114554e-06,
+      "loss": 0.821,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.054938458641251,
+      "learning_rate": 9.866081742442928e-06,
+      "loss": 0.9086,
+      "step": 1588
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0417276877327826,
+      "learning_rate": 9.865843358664933e-06,
+      "loss": 0.5907,
+      "step": 1589
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6459437829315857,
+      "learning_rate": 9.865604765790812e-06,
+      "loss": 0.7597,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.819588589373226,
+      "learning_rate": 9.865365963830818e-06,
+      "loss": 0.8116,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.646120624555309,
+      "learning_rate": 9.865126952795212e-06,
+      "loss": 0.8111,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.181053696863792,
+      "learning_rate": 9.864887732694266e-06,
+      "loss": 0.6727,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9816255731814918,
+      "learning_rate": 9.86464830353826e-06,
+      "loss": 0.7907,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8341963390755711,
+      "learning_rate": 9.864408665337482e-06,
+      "loss": 0.9639,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.224333029079672,
+      "learning_rate": 9.86416881810223e-06,
+      "loss": 0.7533,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8872883608629814,
+      "learning_rate": 9.86392876184281e-06,
+      "loss": 0.8458,
+      "step": 1597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6887712534673247,
+      "learning_rate": 9.86368849656954e-06,
+      "loss": 0.7016,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9804638712612757,
+      "learning_rate": 9.863448022292742e-06,
+      "loss": 0.7915,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9168702384776515,
+      "learning_rate": 9.863207339022753e-06,
+      "loss": 0.7631,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8374878036974485,
+      "learning_rate": 9.862966446769912e-06,
+      "loss": 0.9313,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.634009077646799,
+      "learning_rate": 9.862725345544573e-06,
+      "loss": 0.8203,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6893345630723349,
+      "learning_rate": 9.862484035357095e-06,
+      "loss": 0.7632,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8900901257824676,
+      "learning_rate": 9.86224251621785e-06,
+      "loss": 0.856,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8395987859701515,
+      "learning_rate": 9.862000788137215e-06,
+      "loss": 0.8231,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1255419791308228,
+      "learning_rate": 9.861758851125578e-06,
+      "loss": 0.7383,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.279683786724696,
+      "learning_rate": 9.861516705193334e-06,
+      "loss": 0.6296,
+      "step": 1607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0845557142683044,
+      "learning_rate": 9.861274350350891e-06,
+      "loss": 0.8061,
+      "step": 1608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.091180055609136,
+      "learning_rate": 9.861031786608663e-06,
+      "loss": 0.8514,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.053650144772192,
+      "learning_rate": 9.860789013977074e-06,
+      "loss": 0.7693,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8971239344347746,
+      "learning_rate": 9.860546032466553e-06,
+      "loss": 0.8727,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7184828761167674,
+      "learning_rate": 9.860302842087545e-06,
+      "loss": 0.8608,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3632571212865428,
+      "learning_rate": 9.8600594428505e-06,
+      "loss": 0.6922,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.441978453277159,
+      "learning_rate": 9.859815834765875e-06,
+      "loss": 0.7729,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.126755609056438,
+      "learning_rate": 9.859572017844143e-06,
+      "loss": 0.777,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.272884416089138,
+      "learning_rate": 9.859327992095777e-06,
+      "loss": 0.735,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2320068725385431,
+      "learning_rate": 9.859083757531265e-06,
+      "loss": 0.6672,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1613439746447083,
+      "learning_rate": 9.858839314161101e-06,
+      "loss": 0.633,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0595938762170203,
+      "learning_rate": 9.858594661995792e-06,
+      "loss": 0.7861,
+      "step": 1619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2923299157557473,
+      "learning_rate": 9.85834980104585e-06,
+      "loss": 0.7505,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6893819813616935,
+      "learning_rate": 9.858104731321796e-06,
+      "loss": 0.7773,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.942840270350131,
+      "learning_rate": 9.857859452834161e-06,
+      "loss": 0.9445,
+      "step": 1622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9712501428949742,
+      "learning_rate": 9.857613965593488e-06,
+      "loss": 0.8336,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.188768759976188,
+      "learning_rate": 9.857368269610325e-06,
+      "loss": 0.9203,
+      "step": 1624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8041428513482816,
+      "learning_rate": 9.857122364895229e-06,
+      "loss": 0.7644,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9901294016627296,
+      "learning_rate": 9.856876251458767e-06,
+      "loss": 0.8387,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6811290101289131,
+      "learning_rate": 9.856629929311516e-06,
+      "loss": 0.7949,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2624450939836203,
+      "learning_rate": 9.85638339846406e-06,
+      "loss": 0.6907,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2995015123802272,
+      "learning_rate": 9.856136658926993e-06,
+      "loss": 0.743,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.71394565422585,
+      "learning_rate": 9.855889710710919e-06,
+      "loss": 0.8033,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2991428784546026,
+      "learning_rate": 9.85564255382645e-06,
+      "loss": 0.6778,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2863897360883527,
+      "learning_rate": 9.855395188284206e-06,
+      "loss": 0.7154,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8633216314247645,
+      "learning_rate": 9.85514761409482e-06,
+      "loss": 0.8432,
+      "step": 1633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.248700566482203,
+      "learning_rate": 9.854899831268926e-06,
+      "loss": 0.9017,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5491238983097144,
+      "learning_rate": 9.854651839817174e-06,
+      "loss": 0.7616,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.999364540960768,
+      "learning_rate": 9.854403639750219e-06,
+      "loss": 0.7479,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8881106069065838,
+      "learning_rate": 9.854155231078728e-06,
+      "loss": 0.8392,
+      "step": 1637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8996575662268036,
+      "learning_rate": 9.853906613813378e-06,
+      "loss": 0.8445,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5717188699142512,
+      "learning_rate": 9.85365778796485e-06,
+      "loss": 0.8266,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1159322250500678,
+      "learning_rate": 9.853408753543835e-06,
+      "loss": 0.6253,
+      "step": 1640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.788425188126182,
+      "learning_rate": 9.85315951056104e-06,
+      "loss": 0.8499,
+      "step": 1641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8129541196972268,
+      "learning_rate": 9.852910059027172e-06,
+      "loss": 0.8993,
+      "step": 1642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7448328879459891,
+      "learning_rate": 9.852660398952949e-06,
+      "loss": 0.7931,
+      "step": 1643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9317020735607655,
+      "learning_rate": 9.852410530349102e-06,
+      "loss": 0.8615,
+      "step": 1644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.241400702515855,
+      "learning_rate": 9.852160453226367e-06,
+      "loss": 0.7741,
+      "step": 1645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.011411728611347,
+      "learning_rate": 9.851910167595493e-06,
+      "loss": 0.7381,
+      "step": 1646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6876406081963473,
+      "learning_rate": 9.851659673467232e-06,
+      "loss": 0.8489,
+      "step": 1647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2321936560886158,
+      "learning_rate": 9.85140897085235e-06,
+      "loss": 0.6838,
+      "step": 1648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.742788849936382,
+      "learning_rate": 9.851158059761617e-06,
+      "loss": 0.7993,
+      "step": 1649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7116664557612358,
+      "learning_rate": 9.850906940205822e-06,
+      "loss": 0.8431,
+      "step": 1650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7351685169609903,
+      "learning_rate": 9.85065561219575e-06,
+      "loss": 0.8753,
+      "step": 1651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0287767993910326,
+      "learning_rate": 9.850404075742204e-06,
+      "loss": 0.9427,
+      "step": 1652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1727471517445585,
+      "learning_rate": 9.850152330855992e-06,
+      "loss": 0.8827,
+      "step": 1653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0565009951687012,
+      "learning_rate": 9.849900377547933e-06,
+      "loss": 0.6078,
+      "step": 1654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.753136397964492,
+      "learning_rate": 9.849648215828853e-06,
+      "loss": 0.8424,
+      "step": 1655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4958785074990484,
+      "learning_rate": 9.849395845709588e-06,
+      "loss": 0.7322,
+      "step": 1656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.071324977330242,
+      "learning_rate": 9.849143267200985e-06,
+      "loss": 0.93,
+      "step": 1657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1336759061664117,
+      "learning_rate": 9.848890480313894e-06,
+      "loss": 0.7892,
+      "step": 1658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9149866443370438,
+      "learning_rate": 9.848637485059183e-06,
+      "loss": 0.7599,
+      "step": 1659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8000688820944921,
+      "learning_rate": 9.848384281447718e-06,
+      "loss": 0.8468,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9373377770554177,
+      "learning_rate": 9.848130869490384e-06,
+      "loss": 0.7973,
+      "step": 1661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7933617181620163,
+      "learning_rate": 9.847877249198069e-06,
+      "loss": 0.8865,
+      "step": 1662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8826556474533827,
+      "learning_rate": 9.84762342058167e-06,
+      "loss": 0.7037,
+      "step": 1663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.963664127386394,
+      "learning_rate": 9.8473693836521e-06,
+      "loss": 0.9246,
+      "step": 1664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9787742028013942,
+      "learning_rate": 9.84711513842027e-06,
+      "loss": 0.7711,
+      "step": 1665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4999310336544918,
+      "learning_rate": 9.846860684897107e-06,
+      "loss": 0.7262,
+      "step": 1666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8910041437815854,
+      "learning_rate": 9.846606023093545e-06,
+      "loss": 0.6842,
+      "step": 1667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8924471569595156,
+      "learning_rate": 9.84635115302053e-06,
+      "loss": 0.8083,
+      "step": 1668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7297373037199524,
+      "learning_rate": 9.846096074689012e-06,
+      "loss": 0.8894,
+      "step": 1669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2211009948687168,
+      "learning_rate": 9.845840788109953e-06,
+      "loss": 0.7218,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8428990355968093,
+      "learning_rate": 9.845585293294321e-06,
+      "loss": 0.8073,
+      "step": 1671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7499028627959903,
+      "learning_rate": 9.845329590253098e-06,
+      "loss": 0.7472,
+      "step": 1672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0859341473755177,
+      "learning_rate": 9.845073678997275e-06,
+      "loss": 0.5335,
+      "step": 1673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.168105092291629,
+      "learning_rate": 9.844817559537841e-06,
+      "loss": 0.7428,
+      "step": 1674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.070602925292618,
+      "learning_rate": 9.84456123188581e-06,
+      "loss": 0.7418,
+      "step": 1675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6272064713390155,
+      "learning_rate": 9.84430469605219e-06,
+      "loss": 0.7819,
+      "step": 1676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.907110495546705,
+      "learning_rate": 9.84404795204801e-06,
+      "loss": 0.7935,
+      "step": 1677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9725970082731272,
+      "learning_rate": 9.843790999884301e-06,
+      "loss": 0.8281,
+      "step": 1678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8694363994196308,
+      "learning_rate": 9.843533839572105e-06,
+      "loss": 0.8579,
+      "step": 1679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9789805082514178,
+      "learning_rate": 9.843276471122473e-06,
+      "loss": 0.8116,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5564046223609482,
+      "learning_rate": 9.843018894546466e-06,
+      "loss": 0.8446,
+      "step": 1681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8813003032904176,
+      "learning_rate": 9.84276110985515e-06,
+      "loss": 0.8281,
+      "step": 1682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6970460225587696,
+      "learning_rate": 9.842503117059604e-06,
+      "loss": 0.8033,
+      "step": 1683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.728669038868628,
+      "learning_rate": 9.842244916170913e-06,
+      "loss": 0.7662,
+      "step": 1684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.660145171916095,
+      "learning_rate": 9.841986507200174e-06,
+      "loss": 0.6949,
+      "step": 1685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7247375801658213,
+      "learning_rate": 9.841727890158493e-06,
+      "loss": 0.8722,
+      "step": 1686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.216517692682316,
+      "learning_rate": 9.84146906505698e-06,
+      "loss": 0.8788,
+      "step": 1687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.947132880126373,
+      "learning_rate": 9.841210031906758e-06,
+      "loss": 0.7654,
+      "step": 1688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0360590896492723,
+      "learning_rate": 9.840950790718959e-06,
+      "loss": 0.9168,
+      "step": 1689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8604777161920698,
+      "learning_rate": 9.840691341504725e-06,
+      "loss": 0.7859,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7968873629785391,
+      "learning_rate": 9.8404316842752e-06,
+      "loss": 0.7674,
+      "step": 1691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7761598891766273,
+      "learning_rate": 9.84017181904155e-06,
+      "loss": 0.8557,
+      "step": 1692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8251252150333965,
+      "learning_rate": 9.839911745814933e-06,
+      "loss": 0.863,
+      "step": 1693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1752359632984275,
+      "learning_rate": 9.83965146460653e-06,
+      "loss": 0.8334,
+      "step": 1694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2833246814604524,
+      "learning_rate": 9.839390975427526e-06,
+      "loss": 0.6584,
+      "step": 1695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9765656493001582,
+      "learning_rate": 9.839130278289115e-06,
+      "loss": 0.8928,
+      "step": 1696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6297515463630572,
+      "learning_rate": 9.838869373202498e-06,
+      "loss": 0.861,
+      "step": 1697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2953430880350811,
+      "learning_rate": 9.838608260178885e-06,
+      "loss": 0.5244,
+      "step": 1698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8435249781474057,
+      "learning_rate": 9.838346939229501e-06,
+      "loss": 0.7783,
+      "step": 1699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.709785362166408,
+      "learning_rate": 9.838085410365573e-06,
+      "loss": 0.8033,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6482781744383583,
+      "learning_rate": 9.83782367359834e-06,
+      "loss": 0.7975,
+      "step": 1701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6766290895838845,
+      "learning_rate": 9.837561728939048e-06,
+      "loss": 0.7892,
+      "step": 1702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.609457537182532,
+      "learning_rate": 9.837299576398958e-06,
+      "loss": 0.7097,
+      "step": 1703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9933107692120393,
+      "learning_rate": 9.83703721598933e-06,
+      "loss": 0.8677,
+      "step": 1704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.805070656606186,
+      "learning_rate": 9.83677464772144e-06,
+      "loss": 0.7714,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7074014805543665,
+      "learning_rate": 9.83651187160657e-06,
+      "loss": 1.0001,
+      "step": 1706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7012910436321882,
+      "learning_rate": 9.836248887656016e-06,
+      "loss": 0.8988,
+      "step": 1707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9584340404381269,
+      "learning_rate": 9.835985695881076e-06,
+      "loss": 0.9905,
+      "step": 1708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5590739762066963,
+      "learning_rate": 9.835722296293058e-06,
+      "loss": 0.7263,
+      "step": 1709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1136662133475175,
+      "learning_rate": 9.835458688903286e-06,
+      "loss": 0.7087,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0332698042161392,
+      "learning_rate": 9.835194873723085e-06,
+      "loss": 0.9416,
+      "step": 1711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7716268458913351,
+      "learning_rate": 9.834930850763789e-06,
+      "loss": 0.8691,
+      "step": 1712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4248429281884507,
+      "learning_rate": 9.834666620036748e-06,
+      "loss": 0.8012,
+      "step": 1713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.996986580506235,
+      "learning_rate": 9.834402181553314e-06,
+      "loss": 0.9592,
+      "step": 1714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.080420960489194,
+      "learning_rate": 9.834137535324852e-06,
+      "loss": 0.81,
+      "step": 1715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7799342674932126,
+      "learning_rate": 9.833872681362735e-06,
+      "loss": 0.8586,
+      "step": 1716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8173624598078355,
+      "learning_rate": 9.833607619678343e-06,
+      "loss": 0.8202,
+      "step": 1717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.058423958521397,
+      "learning_rate": 9.833342350283067e-06,
+      "loss": 0.5861,
+      "step": 1718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.148923638221597,
+      "learning_rate": 9.833076873188303e-06,
+      "loss": 0.8039,
+      "step": 1719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0271895508864315,
+      "learning_rate": 9.832811188405464e-06,
+      "loss": 0.8214,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5089542806310963,
+      "learning_rate": 9.832545295945964e-06,
+      "loss": 0.7743,
+      "step": 1721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9651946731498058,
+      "learning_rate": 9.83227919582123e-06,
+      "loss": 0.8602,
+      "step": 1722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6669305893498365,
+      "learning_rate": 9.832012888042699e-06,
+      "loss": 0.8458,
+      "step": 1723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7605287035296722,
+      "learning_rate": 9.831746372621811e-06,
+      "loss": 0.8033,
+      "step": 1724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.133790915868189,
+      "learning_rate": 9.831479649570021e-06,
+      "loss": 0.711,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.90959755514957,
+      "learning_rate": 9.83121271889879e-06,
+      "loss": 0.893,
+      "step": 1726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0185400361257724,
+      "learning_rate": 9.830945580619588e-06,
+      "loss": 0.8937,
+      "step": 1727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8944718738322552,
+      "learning_rate": 9.830678234743896e-06,
+      "loss": 0.9249,
+      "step": 1728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.028059203992863,
+      "learning_rate": 9.830410681283203e-06,
+      "loss": 0.8187,
+      "step": 1729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5469024270981304,
+      "learning_rate": 9.830142920249004e-06,
+      "loss": 0.8362,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3111309238425592,
+      "learning_rate": 9.829874951652807e-06,
+      "loss": 0.6882,
+      "step": 1731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.854660596088181,
+      "learning_rate": 9.829606775506124e-06,
+      "loss": 0.9348,
+      "step": 1732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8696073381467901,
+      "learning_rate": 9.829338391820484e-06,
+      "loss": 0.8556,
+      "step": 1733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5579891782057143,
+      "learning_rate": 9.829069800607418e-06,
+      "loss": 0.8432,
+      "step": 1734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2244142776626887,
+      "learning_rate": 9.828801001878468e-06,
+      "loss": 0.6477,
+      "step": 1735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9101803711551752,
+      "learning_rate": 9.828531995645183e-06,
+      "loss": 0.744,
+      "step": 1736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.714990074673956,
+      "learning_rate": 9.828262781919124e-06,
+      "loss": 0.7201,
+      "step": 1737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7125607692962614,
+      "learning_rate": 9.827993360711863e-06,
+      "loss": 0.8251,
+      "step": 1738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.698739725951539,
+      "learning_rate": 9.827723732034972e-06,
+      "loss": 0.7556,
+      "step": 1739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0023109370033114,
+      "learning_rate": 9.827453895900043e-06,
+      "loss": 0.7606,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8216166297778003,
+      "learning_rate": 9.827183852318666e-06,
+      "loss": 0.834,
+      "step": 1741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7228762348893771,
+      "learning_rate": 9.82691360130245e-06,
+      "loss": 0.7157,
+      "step": 1742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3230376050984205,
+      "learning_rate": 9.826643142863006e-06,
+      "loss": 0.9208,
+      "step": 1743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.243602722148661,
+      "learning_rate": 9.826372477011956e-06,
+      "loss": 0.773,
+      "step": 1744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.207229220623213,
+      "learning_rate": 9.826101603760934e-06,
+      "loss": 0.6266,
+      "step": 1745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8601962525068036,
+      "learning_rate": 9.825830523121575e-06,
+      "loss": 0.8046,
+      "step": 1746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.205442602445377,
+      "learning_rate": 9.825559235105532e-06,
+      "loss": 0.8001,
+      "step": 1747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9987320795236818,
+      "learning_rate": 9.82528773972446e-06,
+      "loss": 0.7606,
+      "step": 1748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1718318718487057,
+      "learning_rate": 9.825016036990029e-06,
+      "loss": 0.8034,
+      "step": 1749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.929024812603161,
+      "learning_rate": 9.824744126913914e-06,
+      "loss": 0.7815,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.922969720330231,
+      "learning_rate": 9.824472009507796e-06,
+      "loss": 0.7862,
+      "step": 1751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1876827909765917,
+      "learning_rate": 9.824199684783373e-06,
+      "loss": 0.664,
+      "step": 1752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2054506601610613,
+      "learning_rate": 9.823927152752344e-06,
+      "loss": 0.6873,
+      "step": 1753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.178155429705447,
+      "learning_rate": 9.823654413426424e-06,
+      "loss": 0.6977,
+      "step": 1754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8031031121857664,
+      "learning_rate": 9.82338146681733e-06,
+      "loss": 0.7594,
+      "step": 1755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8744463380646224,
+      "learning_rate": 9.823108312936791e-06,
+      "loss": 0.6967,
+      "step": 1756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.955960310303873,
+      "learning_rate": 9.822834951796547e-06,
+      "loss": 0.8956,
+      "step": 1757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7400997017283202,
+      "learning_rate": 9.822561383408345e-06,
+      "loss": 0.7435,
+      "step": 1758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8838733051412981,
+      "learning_rate": 9.822287607783938e-06,
+      "loss": 0.9908,
+      "step": 1759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.139752353483334,
+      "learning_rate": 9.822013624935093e-06,
+      "loss": 0.778,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7770656151801019,
+      "learning_rate": 9.821739434873586e-06,
+      "loss": 0.7733,
+      "step": 1761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.314096622523756,
+      "learning_rate": 9.821465037611194e-06,
+      "loss": 0.6633,
+      "step": 1762
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.044282270978045,
+      "learning_rate": 9.821190433159711e-06,
+      "loss": 0.7098,
+      "step": 1763
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2608689217493898,
+      "learning_rate": 9.820915621530939e-06,
+      "loss": 0.7498,
+      "step": 1764
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8565843532237862,
+      "learning_rate": 9.820640602736687e-06,
+      "loss": 0.8096,
+      "step": 1765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.198863744181698,
+      "learning_rate": 9.82036537678877e-06,
+      "loss": 0.8178,
+      "step": 1766
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9379067435045014,
+      "learning_rate": 9.820089943699018e-06,
+      "loss": 0.8835,
+      "step": 1767
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5845616914693914,
+      "learning_rate": 9.819814303479268e-06,
+      "loss": 0.7895,
+      "step": 1768
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5432976246284793,
+      "learning_rate": 9.81953845614136e-06,
+      "loss": 0.8448,
+      "step": 1769
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7583891015726432,
+      "learning_rate": 9.819262401697154e-06,
+      "loss": 0.8889,
+      "step": 1770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8177523419268133,
+      "learning_rate": 9.818986140158507e-06,
+      "loss": 0.829,
+      "step": 1771
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9111990118027706,
+      "learning_rate": 9.818709671537295e-06,
+      "loss": 0.8007,
+      "step": 1772
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8833575446538715,
+      "learning_rate": 9.818432995845394e-06,
+      "loss": 0.9143,
+      "step": 1773
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.096791969524371,
+      "learning_rate": 9.818156113094699e-06,
+      "loss": 0.7846,
+      "step": 1774
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7619156620819765,
+      "learning_rate": 9.817879023297103e-06,
+      "loss": 0.8393,
+      "step": 1775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1913024432970611,
+      "learning_rate": 9.817601726464516e-06,
+      "loss": 0.7559,
+      "step": 1776
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8766213157273055,
+      "learning_rate": 9.817324222608853e-06,
+      "loss": 0.8473,
+      "step": 1777
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.834256608081246,
+      "learning_rate": 9.817046511742042e-06,
+      "loss": 0.8188,
+      "step": 1778
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6668535210755349,
+      "learning_rate": 9.816768593876012e-06,
+      "loss": 0.7308,
+      "step": 1779
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.780655153448778,
+      "learning_rate": 9.816490469022708e-06,
+      "loss": 0.7561,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6812472846642448,
+      "learning_rate": 9.816212137194081e-06,
+      "loss": 0.8947,
+      "step": 1781
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.929748659961367,
+      "learning_rate": 9.815933598402095e-06,
+      "loss": 0.8249,
+      "step": 1782
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5913248751441194,
+      "learning_rate": 9.815654852658714e-06,
+      "loss": 0.8181,
+      "step": 1783
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7205682148662496,
+      "learning_rate": 9.81537589997592e-06,
+      "loss": 0.7515,
+      "step": 1784
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6409669098077229,
+      "learning_rate": 9.815096740365698e-06,
+      "loss": 0.8741,
+      "step": 1785
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8533105456300105,
+      "learning_rate": 9.814817373840046e-06,
+      "loss": 0.8518,
+      "step": 1786
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3173945554078283,
+      "learning_rate": 9.814537800410968e-06,
+      "loss": 0.7125,
+      "step": 1787
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8609313417165405,
+      "learning_rate": 9.814258020090478e-06,
+      "loss": 0.7789,
+      "step": 1788
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2405640060580496,
+      "learning_rate": 9.8139780328906e-06,
+      "loss": 0.7763,
+      "step": 1789
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7226107308362757,
+      "learning_rate": 9.813697838823364e-06,
+      "loss": 0.7517,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9583931834062502,
+      "learning_rate": 9.813417437900812e-06,
+      "loss": 0.7208,
+      "step": 1791
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9814807900179658,
+      "learning_rate": 9.81313683013499e-06,
+      "loss": 0.7961,
+      "step": 1792
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7175864425958118,
+      "learning_rate": 9.81285601553796e-06,
+      "loss": 0.8288,
+      "step": 1793
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2103035777574274,
+      "learning_rate": 9.812574994121791e-06,
+      "loss": 0.8818,
+      "step": 1794
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.599058507573942,
+      "learning_rate": 9.812293765898555e-06,
+      "loss": 0.9495,
+      "step": 1795
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1695764204870116,
+      "learning_rate": 9.812012330880337e-06,
+      "loss": 0.9222,
+      "step": 1796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9260174011112348,
+      "learning_rate": 9.811730689079233e-06,
+      "loss": 0.8111,
+      "step": 1797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8625285218463534,
+      "learning_rate": 9.811448840507345e-06,
+      "loss": 0.8265,
+      "step": 1798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9022510691110273,
+      "learning_rate": 9.811166785176785e-06,
+      "loss": 0.8076,
+      "step": 1799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.014959950639391,
+      "learning_rate": 9.810884523099673e-06,
+      "loss": 0.642,
+      "step": 1800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1467799411975215,
+      "learning_rate": 9.810602054288138e-06,
+      "loss": 0.6952,
+      "step": 1801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1442141550548524,
+      "learning_rate": 9.81031937875432e-06,
+      "loss": 0.9191,
+      "step": 1802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2959514694938346,
+      "learning_rate": 9.810036496510364e-06,
+      "loss": 0.847,
+      "step": 1803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8374607696026053,
+      "learning_rate": 9.809753407568427e-06,
+      "loss": 0.7392,
+      "step": 1804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1087746655958166,
+      "learning_rate": 9.809470111940675e-06,
+      "loss": 0.5751,
+      "step": 1805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8602158383889322,
+      "learning_rate": 9.809186609639281e-06,
+      "loss": 0.7908,
+      "step": 1806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8956618009260922,
+      "learning_rate": 9.808902900676429e-06,
+      "loss": 0.8684,
+      "step": 1807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7129799301434812,
+      "learning_rate": 9.808618985064307e-06,
+      "loss": 0.8114,
+      "step": 1808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0858027719982593,
+      "learning_rate": 9.80833486281512e-06,
+      "loss": 0.7188,
+      "step": 1809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7186842441664092,
+      "learning_rate": 9.808050533941074e-06,
+      "loss": 0.7775,
+      "step": 1810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8063801397164843,
+      "learning_rate": 9.807765998454388e-06,
+      "loss": 0.7396,
+      "step": 1811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2514128912237323,
+      "learning_rate": 9.80748125636729e-06,
+      "loss": 0.8522,
+      "step": 1812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.972058599344271,
+      "learning_rate": 9.807196307692015e-06,
+      "loss": 0.7983,
+      "step": 1813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7966883220615584,
+      "learning_rate": 9.80691115244081e-06,
+      "loss": 0.7989,
+      "step": 1814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7890176210755544,
+      "learning_rate": 9.806625790625925e-06,
+      "loss": 0.6857,
+      "step": 1815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7697677514789163,
+      "learning_rate": 9.806340222259627e-06,
+      "loss": 0.7563,
+      "step": 1816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0245319789120564,
+      "learning_rate": 9.806054447354185e-06,
+      "loss": 0.8251,
+      "step": 1817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9142457269637616,
+      "learning_rate": 9.805768465921878e-06,
+      "loss": 0.8652,
+      "step": 1818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.010303123179374,
+      "learning_rate": 9.805482277974999e-06,
+      "loss": 0.832,
+      "step": 1819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.574871483682648,
+      "learning_rate": 9.805195883525844e-06,
+      "loss": 0.7959,
+      "step": 1820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8889014702536737,
+      "learning_rate": 9.804909282586718e-06,
+      "loss": 0.85,
+      "step": 1821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6608769400467485,
+      "learning_rate": 9.804622475169943e-06,
+      "loss": 0.8345,
+      "step": 1822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.240933758910767,
+      "learning_rate": 9.804335461287835e-06,
+      "loss": 0.9061,
+      "step": 1823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1178902916209217,
+      "learning_rate": 9.804048240952736e-06,
+      "loss": 0.8767,
+      "step": 1824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.534330983255257,
+      "learning_rate": 9.803760814176983e-06,
+      "loss": 0.8165,
+      "step": 1825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8749749683617705,
+      "learning_rate": 9.80347318097293e-06,
+      "loss": 0.882,
+      "step": 1826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7654379796227768,
+      "learning_rate": 9.803185341352936e-06,
+      "loss": 0.8305,
+      "step": 1827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4443250550198345,
+      "learning_rate": 9.802897295329373e-06,
+      "loss": 0.6797,
+      "step": 1828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.043285092539442,
+      "learning_rate": 9.802609042914614e-06,
+      "loss": 0.8042,
+      "step": 1829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.161348790834212,
+      "learning_rate": 9.80232058412105e-06,
+      "loss": 0.7937,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7668593267608976,
+      "learning_rate": 9.802031918961073e-06,
+      "loss": 0.8316,
+      "step": 1831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7841692435147742,
+      "learning_rate": 9.801743047447092e-06,
+      "loss": 0.7608,
+      "step": 1832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.842145548503316,
+      "learning_rate": 9.801453969591517e-06,
+      "loss": 0.8627,
+      "step": 1833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8074514444064038,
+      "learning_rate": 9.80116468540677e-06,
+      "loss": 0.7655,
+      "step": 1834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0468432274955624,
+      "learning_rate": 9.800875194905286e-06,
+      "loss": 0.768,
+      "step": 1835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3491792288691102,
+      "learning_rate": 9.800585498099501e-06,
+      "loss": 0.6815,
+      "step": 1836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8893925597094798,
+      "learning_rate": 9.800295595001868e-06,
+      "loss": 0.7917,
+      "step": 1837
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9633434760246673,
+      "learning_rate": 9.800005485624842e-06,
+      "loss": 0.7304,
+      "step": 1838
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9496529944414416,
+      "learning_rate": 9.79971516998089e-06,
+      "loss": 0.8066,
+      "step": 1839
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2351360846258221,
+      "learning_rate": 9.799424648082486e-06,
+      "loss": 0.5966,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.87301174829259,
+      "learning_rate": 9.799133919942117e-06,
+      "loss": 0.8775,
+      "step": 1841
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.943960934038455,
+      "learning_rate": 9.798842985572274e-06,
+      "loss": 0.8132,
+      "step": 1842
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8701323017698972,
+      "learning_rate": 9.798551844985463e-06,
+      "loss": 0.7955,
+      "step": 1843
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1653604797426733,
+      "learning_rate": 9.79826049819419e-06,
+      "loss": 0.8314,
+      "step": 1844
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.560558956987241,
+      "learning_rate": 9.797968945210978e-06,
+      "loss": 0.7885,
+      "step": 1845
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8907489005073947,
+      "learning_rate": 9.797677186048356e-06,
+      "loss": 0.782,
+      "step": 1846
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1311320630000146,
+      "learning_rate": 9.797385220718859e-06,
+      "loss": 0.7392,
+      "step": 1847
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.417831434729377,
+      "learning_rate": 9.797093049235034e-06,
+      "loss": 0.7795,
+      "step": 1848
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7592251825683891,
+      "learning_rate": 9.796800671609436e-06,
+      "loss": 0.7979,
+      "step": 1849
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9488034057342767,
+      "learning_rate": 9.796508087854632e-06,
+      "loss": 0.7725,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9065435216712652,
+      "learning_rate": 9.796215297983193e-06,
+      "loss": 0.8495,
+      "step": 1851
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7397782313382604,
+      "learning_rate": 9.7959223020077e-06,
+      "loss": 0.6927,
+      "step": 1852
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9641909980529364,
+      "learning_rate": 9.795629099940746e-06,
+      "loss": 0.7313,
+      "step": 1853
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9569051998442442,
+      "learning_rate": 9.795335691794929e-06,
+      "loss": 0.8757,
+      "step": 1854
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0093985285475235,
+      "learning_rate": 9.795042077582856e-06,
+      "loss": 0.9522,
+      "step": 1855
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7920285256575785,
+      "learning_rate": 9.794748257317148e-06,
+      "loss": 0.9726,
+      "step": 1856
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5940075898775445,
+      "learning_rate": 9.794454231010427e-06,
+      "loss": 0.8065,
+      "step": 1857
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8479396704867441,
+      "learning_rate": 9.794159998675331e-06,
+      "loss": 0.8359,
+      "step": 1858
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.936330084504555,
+      "learning_rate": 9.793865560324503e-06,
+      "loss": 0.8382,
+      "step": 1859
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6273926882558125,
+      "learning_rate": 9.793570915970594e-06,
+      "loss": 0.6975,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.121665012875407,
+      "learning_rate": 9.793276065626267e-06,
+      "loss": 0.8644,
+      "step": 1861
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7732245397207878,
+      "learning_rate": 9.792981009304192e-06,
+      "loss": 0.9585,
+      "step": 1862
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.475127394911683,
+      "learning_rate": 9.79268574701705e-06,
+      "loss": 0.7833,
+      "step": 1863
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.862276674626242,
+      "learning_rate": 9.792390278777527e-06,
+      "loss": 0.8259,
+      "step": 1864
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3020872058896664,
+      "learning_rate": 9.79209460459832e-06,
+      "loss": 0.8417,
+      "step": 1865
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7308624758389182,
+      "learning_rate": 9.791798724492136e-06,
+      "loss": 0.6263,
+      "step": 1866
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0040412490992265,
+      "learning_rate": 9.791502638471689e-06,
+      "loss": 0.8712,
+      "step": 1867
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5935914986383584,
+      "learning_rate": 9.7912063465497e-06,
+      "loss": 0.692,
+      "step": 1868
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.029514894747793,
+      "learning_rate": 9.790909848738907e-06,
+      "loss": 0.7306,
+      "step": 1869
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6883287740738802,
+      "learning_rate": 9.790613145052046e-06,
+      "loss": 0.7699,
+      "step": 1870
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.614271408992112,
+      "learning_rate": 9.790316235501869e-06,
+      "loss": 0.8344,
+      "step": 1871
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8339328266933184,
+      "learning_rate": 9.790019120101135e-06,
+      "loss": 0.8862,
+      "step": 1872
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6796016085381986,
+      "learning_rate": 9.789721798862612e-06,
+      "loss": 0.8675,
+      "step": 1873
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2894606065381837,
+      "learning_rate": 9.789424271799075e-06,
+      "loss": 0.7493,
+      "step": 1874
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7364110667801784,
+      "learning_rate": 9.789126538923312e-06,
+      "loss": 0.7648,
+      "step": 1875
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7567186028222286,
+      "learning_rate": 9.788828600248114e-06,
+      "loss": 0.8612,
+      "step": 1876
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0001038725300893,
+      "learning_rate": 9.788530455786288e-06,
+      "loss": 0.7416,
+      "step": 1877
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5871703300173456,
+      "learning_rate": 9.78823210555064e-06,
+      "loss": 0.8724,
+      "step": 1878
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5975129477034045,
+      "learning_rate": 9.787933549553996e-06,
+      "loss": 0.858,
+      "step": 1879
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.110218543414174,
+      "learning_rate": 9.787634787809185e-06,
+      "loss": 0.966,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7023259474623758,
+      "learning_rate": 9.787335820329045e-06,
+      "loss": 0.763,
+      "step": 1881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5147806991941728,
+      "learning_rate": 9.787036647126422e-06,
+      "loss": 0.6923,
+      "step": 1882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6528546241348752,
+      "learning_rate": 9.786737268214172e-06,
+      "loss": 0.6258,
+      "step": 1883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.882454452868324,
+      "learning_rate": 9.786437683605161e-06,
+      "loss": 0.8096,
+      "step": 1884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2279417183313726,
+      "learning_rate": 9.786137893312265e-06,
+      "loss": 0.7258,
+      "step": 1885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9694380916595062,
+      "learning_rate": 9.785837897348363e-06,
+      "loss": 0.8419,
+      "step": 1886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8822257579807196,
+      "learning_rate": 9.785537695726348e-06,
+      "loss": 0.7504,
+      "step": 1887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8153349122443978,
+      "learning_rate": 9.78523728845912e-06,
+      "loss": 0.8242,
+      "step": 1888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1641016951115513,
+      "learning_rate": 9.78493667555959e-06,
+      "loss": 0.7426,
+      "step": 1889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.776697821791942,
+      "learning_rate": 9.784635857040672e-06,
+      "loss": 0.8062,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.309104683273938,
+      "learning_rate": 9.784334832915297e-06,
+      "loss": 0.7577,
+      "step": 1891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.040068105150651,
+      "learning_rate": 9.784033603196397e-06,
+      "loss": 0.8124,
+      "step": 1892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4888765611335133,
+      "learning_rate": 9.78373216789692e-06,
+      "loss": 0.8673,
+      "step": 1893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.007794325250908,
+      "learning_rate": 9.783430527029818e-06,
+      "loss": 0.8081,
+      "step": 1894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8838010616863758,
+      "learning_rate": 9.783128680608052e-06,
+      "loss": 0.8356,
+      "step": 1895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6754874317745556,
+      "learning_rate": 9.782826628644596e-06,
+      "loss": 0.837,
+      "step": 1896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3085171578073125,
+      "learning_rate": 9.782524371152425e-06,
+      "loss": 0.7094,
+      "step": 1897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7211708759264848,
+      "learning_rate": 9.782221908144531e-06,
+      "loss": 0.8903,
+      "step": 1898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8597573282771211,
+      "learning_rate": 9.781919239633912e-06,
+      "loss": 0.8555,
+      "step": 1899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2042796084444556,
+      "learning_rate": 9.781616365633574e-06,
+      "loss": 0.684,
+      "step": 1900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3011935184349899,
+      "learning_rate": 9.78131328615653e-06,
+      "loss": 0.691,
+      "step": 1901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3037867370905944,
+      "learning_rate": 9.781010001215806e-06,
+      "loss": 0.7358,
+      "step": 1902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8164615877684513,
+      "learning_rate": 9.780706510824436e-06,
+      "loss": 0.8252,
+      "step": 1903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0013309467915237,
+      "learning_rate": 9.780402814995458e-06,
+      "loss": 0.8205,
+      "step": 1904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.940596157714832,
+      "learning_rate": 9.780098913741926e-06,
+      "loss": 0.844,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8276050297227862,
+      "learning_rate": 9.779794807076896e-06,
+      "loss": 0.8691,
+      "step": 1906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.05868469092829,
+      "learning_rate": 9.77949049501344e-06,
+      "loss": 0.8062,
+      "step": 1907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0501513275311343,
+      "learning_rate": 9.779185977564633e-06,
+      "loss": 0.8034,
+      "step": 1908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.140755179141967,
+      "learning_rate": 9.77888125474356e-06,
+      "loss": 0.6504,
+      "step": 1909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.22134955491644,
+      "learning_rate": 9.778576326563315e-06,
+      "loss": 0.7965,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6843391789140887,
+      "learning_rate": 9.778271193037003e-06,
+      "loss": 0.822,
+      "step": 1911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.006908654813507,
+      "learning_rate": 9.777965854177737e-06,
+      "loss": 0.9528,
+      "step": 1912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.641640278002267,
+      "learning_rate": 9.777660309998638e-06,
+      "loss": 0.9052,
+      "step": 1913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1151784544302092,
+      "learning_rate": 9.777354560512835e-06,
+      "loss": 0.6447,
+      "step": 1914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7790564075019442,
+      "learning_rate": 9.777048605733465e-06,
+      "loss": 0.7018,
+      "step": 1915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.156522451876192,
+      "learning_rate": 9.776742445673679e-06,
+      "loss": 0.7895,
+      "step": 1916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7508137538853363,
+      "learning_rate": 9.77643608034663e-06,
+      "loss": 0.8351,
+      "step": 1917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2642870117437,
+      "learning_rate": 9.776129509765487e-06,
+      "loss": 0.5999,
+      "step": 1918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0221017287427228,
+      "learning_rate": 9.77582273394342e-06,
+      "loss": 0.9027,
+      "step": 1919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8590396078454885,
+      "learning_rate": 9.775515752893613e-06,
+      "loss": 0.7526,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8344432447632473,
+      "learning_rate": 9.775208566629262e-06,
+      "loss": 0.8054,
+      "step": 1921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.772544170239367,
+      "learning_rate": 9.774901175163562e-06,
+      "loss": 0.9252,
+      "step": 1922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8012523091831327,
+      "learning_rate": 9.774593578509724e-06,
+      "loss": 0.8584,
+      "step": 1923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6109021944217852,
+      "learning_rate": 9.774285776680967e-06,
+      "loss": 0.7986,
+      "step": 1924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7694495007992763,
+      "learning_rate": 9.773977769690517e-06,
+      "loss": 0.7703,
+      "step": 1925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6041698666273827,
+      "learning_rate": 9.773669557551609e-06,
+      "loss": 0.9065,
+      "step": 1926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.250645676343402,
+      "learning_rate": 9.77336114027749e-06,
+      "loss": 0.8706,
+      "step": 1927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7456225094603224,
+      "learning_rate": 9.773052517881412e-06,
+      "loss": 0.9755,
+      "step": 1928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.139397113443179,
+      "learning_rate": 9.772743690376636e-06,
+      "loss": 0.6819,
+      "step": 1929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6311502085925025,
+      "learning_rate": 9.772434657776435e-06,
+      "loss": 0.7091,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7461289504521966,
+      "learning_rate": 9.772125420094087e-06,
+      "loss": 0.8426,
+      "step": 1931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9923551734983236,
+      "learning_rate": 9.771815977342882e-06,
+      "loss": 0.8961,
+      "step": 1932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6970105276897458,
+      "learning_rate": 9.771506329536118e-06,
+      "loss": 0.9152,
+      "step": 1933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2073562044532276,
+      "learning_rate": 9.7711964766871e-06,
+      "loss": 0.8751,
+      "step": 1934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6804711888560573,
+      "learning_rate": 9.770886418809142e-06,
+      "loss": 0.7791,
+      "step": 1935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6371910597736457,
+      "learning_rate": 9.77057615591557e-06,
+      "loss": 0.7255,
+      "step": 1936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5644606191792516,
+      "learning_rate": 9.770265688019718e-06,
+      "loss": 0.7377,
+      "step": 1937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9964738847464893,
+      "learning_rate": 9.769955015134923e-06,
+      "loss": 0.6985,
+      "step": 1938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1392119547556048,
+      "learning_rate": 9.76964413727454e-06,
+      "loss": 0.7991,
+      "step": 1939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7485706715748126,
+      "learning_rate": 9.769333054451923e-06,
+      "loss": 0.77,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.555521985304844,
+      "learning_rate": 9.769021766680444e-06,
+      "loss": 0.6587,
+      "step": 1941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6895334477539519,
+      "learning_rate": 9.768710273973479e-06,
+      "loss": 0.792,
+      "step": 1942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5849214434518761,
+      "learning_rate": 9.768398576344411e-06,
+      "loss": 0.7979,
+      "step": 1943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9530523577947072,
+      "learning_rate": 9.768086673806638e-06,
+      "loss": 0.788,
+      "step": 1944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.929427397119708,
+      "learning_rate": 9.767774566373563e-06,
+      "loss": 0.8083,
+      "step": 1945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7901734279488288,
+      "learning_rate": 9.767462254058593e-06,
+      "loss": 0.7812,
+      "step": 1946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8912042287421675,
+      "learning_rate": 9.767149736875156e-06,
+      "loss": 0.8096,
+      "step": 1947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1581314745672002,
+      "learning_rate": 9.766837014836675e-06,
+      "loss": 0.8168,
+      "step": 1948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.111843196085903,
+      "learning_rate": 9.766524087956592e-06,
+      "loss": 0.899,
+      "step": 1949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5908467059958504,
+      "learning_rate": 9.766210956248354e-06,
+      "loss": 0.903,
+      "step": 1950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9205210204908272,
+      "learning_rate": 9.765897619725416e-06,
+      "loss": 0.714,
+      "step": 1951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.097039258408706,
+      "learning_rate": 9.765584078401242e-06,
+      "loss": 0.7997,
+      "step": 1952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6143468538285877,
+      "learning_rate": 9.765270332289307e-06,
+      "loss": 0.7124,
+      "step": 1953
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8174592737970643,
+      "learning_rate": 9.764956381403095e-06,
+      "loss": 0.8298,
+      "step": 1954
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.659538793393965,
+      "learning_rate": 9.764642225756093e-06,
+      "loss": 0.8383,
+      "step": 1955
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6562265377439342,
+      "learning_rate": 9.764327865361805e-06,
+      "loss": 0.7549,
+      "step": 1956
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9097093267713265,
+      "learning_rate": 9.764013300233736e-06,
+      "loss": 0.9435,
+      "step": 1957
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1666242549208719,
+      "learning_rate": 9.763698530385407e-06,
+      "loss": 0.6868,
+      "step": 1958
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.60164166618835,
+      "learning_rate": 9.76338355583034e-06,
+      "loss": 0.7046,
+      "step": 1959
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9864944135002323,
+      "learning_rate": 9.763068376582075e-06,
+      "loss": 0.5799,
+      "step": 1960
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6546889376266767,
+      "learning_rate": 9.762752992654155e-06,
+      "loss": 0.8045,
+      "step": 1961
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.664730505503271,
+      "learning_rate": 9.76243740406013e-06,
+      "loss": 0.713,
+      "step": 1962
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9057983081184025,
+      "learning_rate": 9.762121610813564e-06,
+      "loss": 0.8441,
+      "step": 1963
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4097564883427132,
+      "learning_rate": 9.761805612928025e-06,
+      "loss": 0.7168,
+      "step": 1964
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6815176998070602,
+      "learning_rate": 9.761489410417095e-06,
+      "loss": 0.8178,
+      "step": 1965
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.633659063346217,
+      "learning_rate": 9.761173003294362e-06,
+      "loss": 0.8258,
+      "step": 1966
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1272259931403927,
+      "learning_rate": 9.76085639157342e-06,
+      "loss": 0.8655,
+      "step": 1967
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.703119738951521,
+      "learning_rate": 9.760539575267875e-06,
+      "loss": 0.9154,
+      "step": 1968
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7578859439951509,
+      "learning_rate": 9.760222554391343e-06,
+      "loss": 0.8675,
+      "step": 1969
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8553046221452247,
+      "learning_rate": 9.759905328957446e-06,
+      "loss": 0.777,
+      "step": 1970
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2337413019800967,
+      "learning_rate": 9.759587898979816e-06,
+      "loss": 0.7993,
+      "step": 1971
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8373204336877675,
+      "learning_rate": 9.759270264472097e-06,
+      "loss": 0.8578,
+      "step": 1972
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9393864474088767,
+      "learning_rate": 9.75895242544793e-06,
+      "loss": 0.7588,
+      "step": 1973
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6235634070944327,
+      "learning_rate": 9.758634381920982e-06,
+      "loss": 0.6742,
+      "step": 1974
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4693076570073966,
+      "learning_rate": 9.758316133904914e-06,
+      "loss": 0.8299,
+      "step": 1975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0919877774433693,
+      "learning_rate": 9.757997681413407e-06,
+      "loss": 0.6487,
+      "step": 1976
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.145886105931977,
+      "learning_rate": 9.757679024460142e-06,
+      "loss": 0.768,
+      "step": 1977
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8025286777567104,
+      "learning_rate": 9.757360163058814e-06,
+      "loss": 0.8422,
+      "step": 1978
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.776912606321288,
+      "learning_rate": 9.757041097223123e-06,
+      "loss": 0.8521,
+      "step": 1979
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.07495943164237,
+      "learning_rate": 9.756721826966783e-06,
+      "loss": 0.8116,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9019284533028553,
+      "learning_rate": 9.756402352303513e-06,
+      "loss": 0.7802,
+      "step": 1981
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5326263491838052,
+      "learning_rate": 9.756082673247039e-06,
+      "loss": 0.7751,
+      "step": 1982
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3438333047096254,
+      "learning_rate": 9.755762789811101e-06,
+      "loss": 0.6197,
+      "step": 1983
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7382397317437694,
+      "learning_rate": 9.755442702009443e-06,
+      "loss": 0.844,
+      "step": 1984
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9249174664028679,
+      "learning_rate": 9.755122409855822e-06,
+      "loss": 0.8421,
+      "step": 1985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.642790987636424,
+      "learning_rate": 9.754801913364002e-06,
+      "loss": 0.8221,
+      "step": 1986
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7277070444562286,
+      "learning_rate": 9.754481212547752e-06,
+      "loss": 0.8263,
+      "step": 1987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5596095137924404,
+      "learning_rate": 9.754160307420858e-06,
+      "loss": 0.7459,
+      "step": 1988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5128880347561853,
+      "learning_rate": 9.753839197997105e-06,
+      "loss": 0.9099,
+      "step": 1989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6830632998246573,
+      "learning_rate": 9.753517884290298e-06,
+      "loss": 0.8415,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6520529347959831,
+      "learning_rate": 9.753196366314237e-06,
+      "loss": 0.8546,
+      "step": 1991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.757607925809137,
+      "learning_rate": 9.752874644082743e-06,
+      "loss": 0.805,
+      "step": 1992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8454763185560357,
+      "learning_rate": 9.752552717609643e-06,
+      "loss": 0.8195,
+      "step": 1993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.038724353588399,
+      "learning_rate": 9.752230586908767e-06,
+      "loss": 0.7846,
+      "step": 1994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7174396140033652,
+      "learning_rate": 9.751908251993956e-06,
+      "loss": 0.7799,
+      "step": 1995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.725467802860383,
+      "learning_rate": 9.751585712879067e-06,
+      "loss": 0.7259,
+      "step": 1996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5727006336836105,
+      "learning_rate": 9.751262969577958e-06,
+      "loss": 0.7794,
+      "step": 1997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7912569641297438,
+      "learning_rate": 9.750940022104497e-06,
+      "loss": 0.8085,
+      "step": 1998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8121963712154114,
+      "learning_rate": 9.75061687047256e-06,
+      "loss": 0.8893,
+      "step": 1999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6220922352579445,
+      "learning_rate": 9.750293514696038e-06,
+      "loss": 0.7192,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2107619077999368,
+      "learning_rate": 9.749969954788825e-06,
+      "loss": 0.7773,
+      "step": 2001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8756221518363319,
+      "learning_rate": 9.749646190764823e-06,
+      "loss": 0.828,
+      "step": 2002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9712918429962818,
+      "learning_rate": 9.749322222637947e-06,
+      "loss": 0.884,
+      "step": 2003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.686045867962633,
+      "learning_rate": 9.748998050422117e-06,
+      "loss": 0.9214,
+      "step": 2004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6850619060008805,
+      "learning_rate": 9.748673674131264e-06,
+      "loss": 0.8904,
+      "step": 2005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8241697565625037,
+      "learning_rate": 9.748349093779327e-06,
+      "loss": 0.87,
+      "step": 2006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9157344817873305,
+      "learning_rate": 9.748024309380256e-06,
+      "loss": 0.8214,
+      "step": 2007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.703108067674488,
+      "learning_rate": 9.747699320948006e-06,
+      "loss": 0.8757,
+      "step": 2008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1449331010844817,
+      "learning_rate": 9.747374128496541e-06,
+      "loss": 0.902,
+      "step": 2009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8222127377192596,
+      "learning_rate": 9.747048732039837e-06,
+      "loss": 0.8507,
+      "step": 2010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1308579108555021,
+      "learning_rate": 9.746723131591878e-06,
+      "loss": 0.7068,
+      "step": 2011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8809876675953765,
+      "learning_rate": 9.746397327166653e-06,
+      "loss": 0.8885,
+      "step": 2012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.88885445744319,
+      "learning_rate": 9.746071318778165e-06,
+      "loss": 0.7806,
+      "step": 2013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.131067596602769,
+      "learning_rate": 9.745745106440422e-06,
+      "loss": 0.8668,
+      "step": 2014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9600808770004676,
+      "learning_rate": 9.745418690167444e-06,
+      "loss": 0.8425,
+      "step": 2015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1545147291859243,
+      "learning_rate": 9.745092069973254e-06,
+      "loss": 0.7711,
+      "step": 2016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0294488774037873,
+      "learning_rate": 9.74476524587189e-06,
+      "loss": 0.7816,
+      "step": 2017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5759538769200259,
+      "learning_rate": 9.744438217877396e-06,
+      "loss": 0.8712,
+      "step": 2018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.293079142869131,
+      "learning_rate": 9.744110986003826e-06,
+      "loss": 0.7437,
+      "step": 2019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.891918767055586,
+      "learning_rate": 9.743783550265242e-06,
+      "loss": 0.8373,
+      "step": 2020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7852929017959525,
+      "learning_rate": 9.743455910675712e-06,
+      "loss": 0.7964,
+      "step": 2021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7059305954029618,
+      "learning_rate": 9.74312806724932e-06,
+      "loss": 0.8725,
+      "step": 2022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4962284947350004,
+      "learning_rate": 9.74280002000015e-06,
+      "loss": 0.7058,
+      "step": 2023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.798225606460176,
+      "learning_rate": 9.742471768942299e-06,
+      "loss": 0.7774,
+      "step": 2024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4142230080608107,
+      "learning_rate": 9.742143314089877e-06,
+      "loss": 0.7815,
+      "step": 2025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5858770471490127,
+      "learning_rate": 9.741814655456994e-06,
+      "loss": 0.8367,
+      "step": 2026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7068815350779858,
+      "learning_rate": 9.741485793057774e-06,
+      "loss": 0.7876,
+      "step": 2027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5737199705792606,
+      "learning_rate": 9.74115672690635e-06,
+      "loss": 0.8116,
+      "step": 2028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8939515633021011,
+      "learning_rate": 9.740827457016863e-06,
+      "loss": 0.774,
+      "step": 2029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4906435169225278,
+      "learning_rate": 9.74049798340346e-06,
+      "loss": 0.8424,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5990729357996543,
+      "learning_rate": 9.740168306080303e-06,
+      "loss": 0.8469,
+      "step": 2031
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8372030507545318,
+      "learning_rate": 9.739838425061557e-06,
+      "loss": 0.849,
+      "step": 2032
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7624923233412266,
+      "learning_rate": 9.739508340361397e-06,
+      "loss": 0.8658,
+      "step": 2033
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8004371337547591,
+      "learning_rate": 9.739178051994008e-06,
+      "loss": 0.7235,
+      "step": 2034
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8008453007986394,
+      "learning_rate": 9.738847559973584e-06,
+      "loss": 0.7674,
+      "step": 2035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0527529458123865,
+      "learning_rate": 9.738516864314324e-06,
+      "loss": 0.6086,
+      "step": 2036
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0890225583533097,
+      "learning_rate": 9.738185965030444e-06,
+      "loss": 0.6266,
+      "step": 2037
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6962455356747483,
+      "learning_rate": 9.73785486213616e-06,
+      "loss": 0.8146,
+      "step": 2038
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.968827804929509,
+      "learning_rate": 9.7375235556457e-06,
+      "loss": 0.8626,
+      "step": 2039
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2153405702982902,
+      "learning_rate": 9.7371920455733e-06,
+      "loss": 0.6156,
+      "step": 2040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.210333380248047,
+      "learning_rate": 9.736860331933212e-06,
+      "loss": 0.8346,
+      "step": 2041
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7422141831234499,
+      "learning_rate": 9.736528414739681e-06,
+      "loss": 0.8093,
+      "step": 2042
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.233205023394286,
+      "learning_rate": 9.736196294006978e-06,
+      "loss": 0.6739,
+      "step": 2043
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.609377032852449,
+      "learning_rate": 9.735863969749373e-06,
+      "loss": 0.7796,
+      "step": 2044
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8330132201505052,
+      "learning_rate": 9.735531441981145e-06,
+      "loss": 0.8966,
+      "step": 2045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0985234920992761,
+      "learning_rate": 9.735198710716585e-06,
+      "loss": 0.7183,
+      "step": 2046
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0794584221387067,
+      "learning_rate": 9.734865775969988e-06,
+      "loss": 0.6855,
+      "step": 2047
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1955052782437745,
+      "learning_rate": 9.734532637755665e-06,
+      "loss": 0.8295,
+      "step": 2048
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0788438132827796,
+      "learning_rate": 9.734199296087932e-06,
+      "loss": 0.6974,
+      "step": 2049
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9790678796737929,
+      "learning_rate": 9.73386575098111e-06,
+      "loss": 0.6102,
+      "step": 2050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.777367922022324,
+      "learning_rate": 9.733532002449533e-06,
+      "loss": 0.8286,
+      "step": 2051
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.85138692958752,
+      "learning_rate": 9.733198050507544e-06,
+      "loss": 0.7377,
+      "step": 2052
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0280542136520285,
+      "learning_rate": 9.732863895169496e-06,
+      "loss": 0.8844,
+      "step": 2053
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8922240205381156,
+      "learning_rate": 9.732529536449741e-06,
+      "loss": 0.7832,
+      "step": 2054
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3270394367385205,
+      "learning_rate": 9.732194974362657e-06,
+      "loss": 0.9634,
+      "step": 2055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8753622923580315,
+      "learning_rate": 9.731860208922612e-06,
+      "loss": 0.7965,
+      "step": 2056
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6261150353633826,
+      "learning_rate": 9.731525240143998e-06,
+      "loss": 0.8729,
+      "step": 2057
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.245659816385571,
+      "learning_rate": 9.731190068041205e-06,
+      "loss": 0.7473,
+      "step": 2058
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9446516914302308,
+      "learning_rate": 9.730854692628637e-06,
+      "loss": 0.78,
+      "step": 2059
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9229000760691017,
+      "learning_rate": 9.730519113920708e-06,
+      "loss": 0.924,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1115702935902205,
+      "learning_rate": 9.730183331931838e-06,
+      "loss": 0.6102,
+      "step": 2061
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8237516601655799,
+      "learning_rate": 9.729847346676455e-06,
+      "loss": 0.888,
+      "step": 2062
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.795401908862819,
+      "learning_rate": 9.729511158168999e-06,
+      "loss": 0.749,
+      "step": 2063
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6589285479094653,
+      "learning_rate": 9.729174766423912e-06,
+      "loss": 0.8767,
+      "step": 2064
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1730113535285016,
+      "learning_rate": 9.728838171455655e-06,
+      "loss": 0.7492,
+      "step": 2065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1313790630073308,
+      "learning_rate": 9.72850137327869e-06,
+      "loss": 0.7167,
+      "step": 2066
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7564025888377155,
+      "learning_rate": 9.728164371907489e-06,
+      "loss": 0.7978,
+      "step": 2067
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8512733599495503,
+      "learning_rate": 9.727827167356537e-06,
+      "loss": 0.8856,
+      "step": 2068
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1302509880789096,
+      "learning_rate": 9.72748975964032e-06,
+      "loss": 0.7782,
+      "step": 2069
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1067706291796857,
+      "learning_rate": 9.72715214877334e-06,
+      "loss": 0.8005,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9575051226898166,
+      "learning_rate": 9.726814334770106e-06,
+      "loss": 0.7989,
+      "step": 2071
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1361682534229836,
+      "learning_rate": 9.72647631764513e-06,
+      "loss": 0.7365,
+      "step": 2072
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.661915269521918,
+      "learning_rate": 9.72613809741294e-06,
+      "loss": 0.8028,
+      "step": 2073
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8410590226687573,
+      "learning_rate": 9.725799674088072e-06,
+      "loss": 0.8303,
+      "step": 2074
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7410793819911208,
+      "learning_rate": 9.725461047685067e-06,
+      "loss": 0.8115,
+      "step": 2075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7930834719181117,
+      "learning_rate": 9.725122218218477e-06,
+      "loss": 0.7957,
+      "step": 2076
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5875563664002392,
+      "learning_rate": 9.72478318570286e-06,
+      "loss": 0.7785,
+      "step": 2077
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9697281587536024,
+      "learning_rate": 9.724443950152789e-06,
+      "loss": 0.8105,
+      "step": 2078
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2434629264080046,
+      "learning_rate": 9.724104511582838e-06,
+      "loss": 0.855,
+      "step": 2079
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8621720954862966,
+      "learning_rate": 9.723764870007596e-06,
+      "loss": 0.7866,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6444669733428114,
+      "learning_rate": 9.723425025441657e-06,
+      "loss": 0.8268,
+      "step": 2081
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0064653843755194,
+      "learning_rate": 9.723084977899624e-06,
+      "loss": 0.7344,
+      "step": 2082
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.686114549380172,
+      "learning_rate": 9.722744727396111e-06,
+      "loss": 0.8201,
+      "step": 2083
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.018168465775486,
+      "learning_rate": 9.72240427394574e-06,
+      "loss": 0.8217,
+      "step": 2084
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7510366602859624,
+      "learning_rate": 9.72206361756314e-06,
+      "loss": 0.8633,
+      "step": 2085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.716640552022623,
+      "learning_rate": 9.721722758262948e-06,
+      "loss": 0.733,
+      "step": 2086
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8610569193208832,
+      "learning_rate": 9.721381696059815e-06,
+      "loss": 0.8787,
+      "step": 2087
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.753397923846987,
+      "learning_rate": 9.721040430968393e-06,
+      "loss": 0.858,
+      "step": 2088
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4389446743761107,
+      "learning_rate": 9.720698963003351e-06,
+      "loss": 0.6845,
+      "step": 2089
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4320063168878538,
+      "learning_rate": 9.720357292179363e-06,
+      "loss": 0.8223,
+      "step": 2090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7413432393033872,
+      "learning_rate": 9.720015418511108e-06,
+      "loss": 0.9091,
+      "step": 2091
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8447578296961658,
+      "learning_rate": 9.719673342013277e-06,
+      "loss": 0.8901,
+      "step": 2092
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.49549165999107,
+      "learning_rate": 9.719331062700572e-06,
+      "loss": 0.7697,
+      "step": 2093
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7125659873652135,
+      "learning_rate": 9.7189885805877e-06,
+      "loss": 0.9122,
+      "step": 2094
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8137477710382734,
+      "learning_rate": 9.71864589568938e-06,
+      "loss": 0.7048,
+      "step": 2095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9011930121378977,
+      "learning_rate": 9.718303008020338e-06,
+      "loss": 0.7874,
+      "step": 2096
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.342306380600963,
+      "learning_rate": 9.717959917595305e-06,
+      "loss": 0.8583,
+      "step": 2097
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5934805618907664,
+      "learning_rate": 9.71761662442903e-06,
+      "loss": 0.6422,
+      "step": 2098
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1209929159261405,
+      "learning_rate": 9.717273128536259e-06,
+      "loss": 0.544,
+      "step": 2099
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9472053617899854,
+      "learning_rate": 9.716929429931757e-06,
+      "loss": 0.7731,
+      "step": 2100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6103263446906533,
+      "learning_rate": 9.716585528630291e-06,
+      "loss": 0.7062,
+      "step": 2101
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5905218785425557,
+      "learning_rate": 9.716241424646642e-06,
+      "loss": 0.8444,
+      "step": 2102
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7388149807886875,
+      "learning_rate": 9.715897117995594e-06,
+      "loss": 0.7966,
+      "step": 2103
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8488476065411366,
+      "learning_rate": 9.715552608691944e-06,
+      "loss": 0.7976,
+      "step": 2104
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2474025318581248,
+      "learning_rate": 9.715207896750499e-06,
+      "loss": 0.6234,
+      "step": 2105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8104684362805654,
+      "learning_rate": 9.714862982186067e-06,
+      "loss": 0.7972,
+      "step": 2106
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7062249231481945,
+      "learning_rate": 9.714517865013473e-06,
+      "loss": 0.7849,
+      "step": 2107
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8530940104867815,
+      "learning_rate": 9.714172545247544e-06,
+      "loss": 0.8139,
+      "step": 2108
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.07843523046722,
+      "learning_rate": 9.713827022903124e-06,
+      "loss": 0.7805,
+      "step": 2109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.751164496933533,
+      "learning_rate": 9.71348129799506e-06,
+      "loss": 0.8884,
+      "step": 2110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6365536031961752,
+      "learning_rate": 9.713135370538204e-06,
+      "loss": 0.7778,
+      "step": 2111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.182836183035815,
+      "learning_rate": 9.712789240547426e-06,
+      "loss": 0.8907,
+      "step": 2112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7834481143563627,
+      "learning_rate": 9.712442908037599e-06,
+      "loss": 0.7408,
+      "step": 2113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.149001733880411,
+      "learning_rate": 9.712096373023603e-06,
+      "loss": 0.7709,
+      "step": 2114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5930176188602134,
+      "learning_rate": 9.711749635520333e-06,
+      "loss": 0.7231,
+      "step": 2115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2432134673413369,
+      "learning_rate": 9.711402695542689e-06,
+      "loss": 0.7099,
+      "step": 2116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5209268736147443,
+      "learning_rate": 9.711055553105575e-06,
+      "loss": 0.764,
+      "step": 2117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1943647328833007,
+      "learning_rate": 9.710708208223914e-06,
+      "loss": 0.699,
+      "step": 2118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1271211378334978,
+      "learning_rate": 9.710360660912629e-06,
+      "loss": 0.7069,
+      "step": 2119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5703115429798498,
+      "learning_rate": 9.710012911186657e-06,
+      "loss": 0.7777,
+      "step": 2120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6930776025393108,
+      "learning_rate": 9.70966495906094e-06,
+      "loss": 0.8335,
+      "step": 2121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8323802185104376,
+      "learning_rate": 9.70931680455043e-06,
+      "loss": 0.9131,
+      "step": 2122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1472993692966984,
+      "learning_rate": 9.708968447670088e-06,
+      "loss": 0.822,
+      "step": 2123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5382976281342173,
+      "learning_rate": 9.708619888434887e-06,
+      "loss": 0.9076,
+      "step": 2124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0623012233721782,
+      "learning_rate": 9.708271126859801e-06,
+      "loss": 0.7199,
+      "step": 2125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6383908156734646,
+      "learning_rate": 9.70792216295982e-06,
+      "loss": 0.6069,
+      "step": 2126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6436896962798002,
+      "learning_rate": 9.707572996749936e-06,
+      "loss": 0.8542,
+      "step": 2127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9931847758761303,
+      "learning_rate": 9.707223628245157e-06,
+      "loss": 0.9467,
+      "step": 2128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.704068921377075,
+      "learning_rate": 9.706874057460497e-06,
+      "loss": 0.8399,
+      "step": 2129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7775969431344216,
+      "learning_rate": 9.706524284410973e-06,
+      "loss": 0.8325,
+      "step": 2130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.221541233617533,
+      "learning_rate": 9.70617430911162e-06,
+      "loss": 0.7222,
+      "step": 2131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0425088745091073,
+      "learning_rate": 9.705824131577476e-06,
+      "loss": 0.793,
+      "step": 2132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.452801921883894,
+      "learning_rate": 9.70547375182359e-06,
+      "loss": 0.8669,
+      "step": 2133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5866098976332557,
+      "learning_rate": 9.705123169865016e-06,
+      "loss": 0.7471,
+      "step": 2134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0502090878066173,
+      "learning_rate": 9.70477238571682e-06,
+      "loss": 0.8849,
+      "step": 2135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9860763889147477,
+      "learning_rate": 9.704421399394077e-06,
+      "loss": 0.8201,
+      "step": 2136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7981154181041545,
+      "learning_rate": 9.704070210911868e-06,
+      "loss": 0.8116,
+      "step": 2137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9476571125324342,
+      "learning_rate": 9.703718820285288e-06,
+      "loss": 0.7907,
+      "step": 2138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.813924896422818,
+      "learning_rate": 9.703367227529432e-06,
+      "loss": 0.8094,
+      "step": 2139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6383647250196023,
+      "learning_rate": 9.703015432659413e-06,
+      "loss": 0.7569,
+      "step": 2140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1118546646588792,
+      "learning_rate": 9.702663435690344e-06,
+      "loss": 0.7674,
+      "step": 2141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6112490126318049,
+      "learning_rate": 9.702311236637357e-06,
+      "loss": 0.7839,
+      "step": 2142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.709079523288405,
+      "learning_rate": 9.701958835515582e-06,
+      "loss": 0.8253,
+      "step": 2143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4214797680080002,
+      "learning_rate": 9.701606232340165e-06,
+      "loss": 0.6308,
+      "step": 2144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3948169680133944,
+      "learning_rate": 9.701253427126257e-06,
+      "loss": 0.8913,
+      "step": 2145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9971634675523084,
+      "learning_rate": 9.700900419889018e-06,
+      "loss": 0.8561,
+      "step": 2146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8796501817084723,
+      "learning_rate": 9.700547210643618e-06,
+      "loss": 0.8195,
+      "step": 2147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7582350682182022,
+      "learning_rate": 9.700193799405238e-06,
+      "loss": 0.845,
+      "step": 2148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2094828096052874,
+      "learning_rate": 9.699840186189061e-06,
+      "loss": 0.6443,
+      "step": 2149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0929361423240327,
+      "learning_rate": 9.699486371010285e-06,
+      "loss": 0.6585,
+      "step": 2150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6037702893392285,
+      "learning_rate": 9.69913235388411e-06,
+      "loss": 0.8319,
+      "step": 2151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.876397093157488,
+      "learning_rate": 9.698778134825755e-06,
+      "loss": 0.8803,
+      "step": 2152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4557799461012304,
+      "learning_rate": 9.69842371385044e-06,
+      "loss": 0.9302,
+      "step": 2153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1450882907410111,
+      "learning_rate": 9.698069090973391e-06,
+      "loss": 0.6614,
+      "step": 2154
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6155903157574731,
+      "learning_rate": 9.697714266209853e-06,
+      "loss": 0.7355,
+      "step": 2155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7455095504921767,
+      "learning_rate": 9.697359239575069e-06,
+      "loss": 0.8077,
+      "step": 2156
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8733208425235095,
+      "learning_rate": 9.697004011084298e-06,
+      "loss": 0.755,
+      "step": 2157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.937767712274827,
+      "learning_rate": 9.696648580752803e-06,
+      "loss": 0.782,
+      "step": 2158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8138653932112567,
+      "learning_rate": 9.696292948595857e-06,
+      "loss": 0.8026,
+      "step": 2159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6472395671653484,
+      "learning_rate": 9.695937114628746e-06,
+      "loss": 0.7835,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.850153491861609,
+      "learning_rate": 9.695581078866759e-06,
+      "loss": 0.7868,
+      "step": 2161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.829799161567906,
+      "learning_rate": 9.695224841325194e-06,
+      "loss": 0.9028,
+      "step": 2162
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6640752391006002,
+      "learning_rate": 9.694868402019362e-06,
+      "loss": 0.7899,
+      "step": 2163
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2166236678057674,
+      "learning_rate": 9.694511760964578e-06,
+      "loss": 0.5672,
+      "step": 2164
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8698344213057951,
+      "learning_rate": 9.694154918176168e-06,
+      "loss": 0.8298,
+      "step": 2165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.720348304139328,
+      "learning_rate": 9.693797873669466e-06,
+      "loss": 0.7836,
+      "step": 2166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.742744717544489,
+      "learning_rate": 9.693440627459817e-06,
+      "loss": 0.8427,
+      "step": 2167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6783054862394318,
+      "learning_rate": 9.693083179562573e-06,
+      "loss": 0.8788,
+      "step": 2168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0720464008399655,
+      "learning_rate": 9.69272552999309e-06,
+      "loss": 0.7831,
+      "step": 2169
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.692578059531601,
+      "learning_rate": 9.69236767876674e-06,
+      "loss": 0.8758,
+      "step": 2170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6284297810419266,
+      "learning_rate": 9.692009625898902e-06,
+      "loss": 0.6593,
+      "step": 2171
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6201731149810843,
+      "learning_rate": 9.69165137140496e-06,
+      "loss": 0.6608,
+      "step": 2172
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7563216545483271,
+      "learning_rate": 9.69129291530031e-06,
+      "loss": 0.9336,
+      "step": 2173
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0865709247286617,
+      "learning_rate": 9.690934257600353e-06,
+      "loss": 0.9159,
+      "step": 2174
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6835331269277722,
+      "learning_rate": 9.690575398320504e-06,
+      "loss": 0.8276,
+      "step": 2175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.209933461639406,
+      "learning_rate": 9.690216337476188e-06,
+      "loss": 0.9469,
+      "step": 2176
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8677077289054334,
+      "learning_rate": 9.689857075082828e-06,
+      "loss": 0.7619,
+      "step": 2177
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1075042562484738,
+      "learning_rate": 9.689497611155862e-06,
+      "loss": 0.585,
+      "step": 2178
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6970421927690456,
+      "learning_rate": 9.689137945710742e-06,
+      "loss": 0.7891,
+      "step": 2179
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.988169712609248,
+      "learning_rate": 9.688778078762922e-06,
+      "loss": 0.7604,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.996298751145542,
+      "learning_rate": 9.688418010327862e-06,
+      "loss": 0.8424,
+      "step": 2181
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5400946330113616,
+      "learning_rate": 9.688057740421042e-06,
+      "loss": 0.6531,
+      "step": 2182
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7938835171474754,
+      "learning_rate": 9.68769726905794e-06,
+      "loss": 0.7497,
+      "step": 2183
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6327240895500519,
+      "learning_rate": 9.687336596254045e-06,
+      "loss": 0.8483,
+      "step": 2184
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6631316691495759,
+      "learning_rate": 9.686975722024856e-06,
+      "loss": 0.7734,
+      "step": 2185
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7689716057476104,
+      "learning_rate": 9.686614646385883e-06,
+      "loss": 0.7354,
+      "step": 2186
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.981335356087586,
+      "learning_rate": 9.68625336935264e-06,
+      "loss": 0.7445,
+      "step": 2187
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7732113126911941,
+      "learning_rate": 9.685891890940655e-06,
+      "loss": 0.7572,
+      "step": 2188
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7113152664974485,
+      "learning_rate": 9.685530211165459e-06,
+      "loss": 0.8348,
+      "step": 2189
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1584802544514288,
+      "learning_rate": 9.685168330042594e-06,
+      "loss": 0.7816,
+      "step": 2190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.993424253989225,
+      "learning_rate": 9.68480624758761e-06,
+      "loss": 0.89,
+      "step": 2191
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8600050484691475,
+      "learning_rate": 9.68444396381607e-06,
+      "loss": 0.887,
+      "step": 2192
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7802155520063916,
+      "learning_rate": 9.68408147874354e-06,
+      "loss": 0.8351,
+      "step": 2193
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7220804149820235,
+      "learning_rate": 9.683718792385595e-06,
+      "loss": 0.7151,
+      "step": 2194
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1673921203728985,
+      "learning_rate": 9.683355904757822e-06,
+      "loss": 0.7287,
+      "step": 2195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6791684115013985,
+      "learning_rate": 9.682992815875818e-06,
+      "loss": 0.7935,
+      "step": 2196
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.012970558968065,
+      "learning_rate": 9.682629525755182e-06,
+      "loss": 0.8804,
+      "step": 2197
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.013491321069822,
+      "learning_rate": 9.682266034411527e-06,
+      "loss": 0.8798,
+      "step": 2198
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.804455601011118,
+      "learning_rate": 9.681902341860471e-06,
+      "loss": 0.8419,
+      "step": 2199
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7355762984434042,
+      "learning_rate": 9.681538448117647e-06,
+      "loss": 0.7986,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2620252970867032,
+      "learning_rate": 9.681174353198687e-06,
+      "loss": 0.6862,
+      "step": 2201
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6716941291453344,
+      "learning_rate": 9.68081005711924e-06,
+      "loss": 0.7594,
+      "step": 2202
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6392944597407049,
+      "learning_rate": 9.68044555989496e-06,
+      "loss": 0.8608,
+      "step": 2203
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.895565493502119,
+      "learning_rate": 9.680080861541511e-06,
+      "loss": 0.7497,
+      "step": 2204
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.990370483205676,
+      "learning_rate": 9.679715962074566e-06,
+      "loss": 0.9194,
+      "step": 2205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7020509805381956,
+      "learning_rate": 9.6793508615098e-06,
+      "loss": 0.6641,
+      "step": 2206
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7213287715725158,
+      "learning_rate": 9.678985559862909e-06,
+      "loss": 0.792,
+      "step": 2207
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1043147627952763,
+      "learning_rate": 9.678620057149586e-06,
+      "loss": 0.7863,
+      "step": 2208
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5529539304052187,
+      "learning_rate": 9.678254353385538e-06,
+      "loss": 0.8392,
+      "step": 2209
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.026738997889,
+      "learning_rate": 9.677888448586485e-06,
+      "loss": 0.8287,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.17889026038274,
+      "learning_rate": 9.677522342768145e-06,
+      "loss": 0.6874,
+      "step": 2211
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5466523198575948,
+      "learning_rate": 9.677156035946253e-06,
+      "loss": 0.751,
+      "step": 2212
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.05114236460663,
+      "learning_rate": 9.67678952813655e-06,
+      "loss": 0.6539,
+      "step": 2213
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.755403315263653,
+      "learning_rate": 9.676422819354785e-06,
+      "loss": 0.7174,
+      "step": 2214
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6955386586756522,
+      "learning_rate": 9.676055909616716e-06,
+      "loss": 0.8638,
+      "step": 2215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0781734509100542,
+      "learning_rate": 9.675688798938112e-06,
+      "loss": 0.6439,
+      "step": 2216
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.914297206989689,
+      "learning_rate": 9.675321487334744e-06,
+      "loss": 0.7457,
+      "step": 2217
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.042872469635663,
+      "learning_rate": 9.674953974822402e-06,
+      "loss": 0.8269,
+      "step": 2218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5807839256257732,
+      "learning_rate": 9.674586261416874e-06,
+      "loss": 0.7868,
+      "step": 2219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.862141513893754,
+      "learning_rate": 9.674218347133965e-06,
+      "loss": 0.8635,
+      "step": 2220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0970496598946697,
+      "learning_rate": 9.673850231989484e-06,
+      "loss": 0.8651,
+      "step": 2221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5651496493738248,
+      "learning_rate": 9.673481915999249e-06,
+      "loss": 0.8252,
+      "step": 2222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.110297718725955,
+      "learning_rate": 9.673113399179085e-06,
+      "loss": 0.8795,
+      "step": 2223
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8309238270884334,
+      "learning_rate": 9.672744681544834e-06,
+      "loss": 0.7621,
+      "step": 2224
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6244991606565982,
+      "learning_rate": 9.672375763112335e-06,
+      "loss": 0.754,
+      "step": 2225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6442445431642172,
+      "learning_rate": 9.672006643897444e-06,
+      "loss": 0.6662,
+      "step": 2226
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8079349503792317,
+      "learning_rate": 9.671637323916023e-06,
+      "loss": 0.7995,
+      "step": 2227
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9746847323362462,
+      "learning_rate": 9.671267803183942e-06,
+      "loss": 0.8669,
+      "step": 2228
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.789462300290376,
+      "learning_rate": 9.670898081717079e-06,
+      "loss": 0.8217,
+      "step": 2229
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7406206352098226,
+      "learning_rate": 9.670528159531325e-06,
+      "loss": 0.8729,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8557998667222677,
+      "learning_rate": 9.670158036642572e-06,
+      "loss": 0.8015,
+      "step": 2231
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2215086841708445,
+      "learning_rate": 9.669787713066728e-06,
+      "loss": 0.7904,
+      "step": 2232
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7890949983387185,
+      "learning_rate": 9.669417188819704e-06,
+      "loss": 0.8196,
+      "step": 2233
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.766273875072158,
+      "learning_rate": 9.669046463917427e-06,
+      "loss": 0.8534,
+      "step": 2234
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2382935922010654,
+      "learning_rate": 9.668675538375823e-06,
+      "loss": 0.649,
+      "step": 2235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8336172649849738,
+      "learning_rate": 9.668304412210835e-06,
+      "loss": 0.8142,
+      "step": 2236
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7169254103629565,
+      "learning_rate": 9.667933085438407e-06,
+      "loss": 0.7961,
+      "step": 2237
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8090634387950288,
+      "learning_rate": 9.6675615580745e-06,
+      "loss": 0.8494,
+      "step": 2238
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6669141307395852,
+      "learning_rate": 9.667189830135078e-06,
+      "loss": 0.7425,
+      "step": 2239
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1837745749235358,
+      "learning_rate": 9.666817901636115e-06,
+      "loss": 0.6026,
+      "step": 2240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8237853215595428,
+      "learning_rate": 9.666445772593592e-06,
+      "loss": 0.9975,
+      "step": 2241
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6313251286878074,
+      "learning_rate": 9.666073443023502e-06,
+      "loss": 0.9934,
+      "step": 2242
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3421518591353145,
+      "learning_rate": 9.665700912941844e-06,
+      "loss": 0.7565,
+      "step": 2243
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2677323799336406,
+      "learning_rate": 9.665328182364627e-06,
+      "loss": 0.7492,
+      "step": 2244
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.116629369837278,
+      "learning_rate": 9.664955251307867e-06,
+      "loss": 0.6644,
+      "step": 2245
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6938860132104065,
+      "learning_rate": 9.66458211978759e-06,
+      "loss": 0.7751,
+      "step": 2246
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.259376822444757,
+      "learning_rate": 9.664208787819833e-06,
+      "loss": 0.7848,
+      "step": 2247
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7625911024028644,
+      "learning_rate": 9.663835255420635e-06,
+      "loss": 0.7816,
+      "step": 2248
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8713123805035425,
+      "learning_rate": 9.663461522606049e-06,
+      "loss": 0.8146,
+      "step": 2249
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7307267560729447,
+      "learning_rate": 9.663087589392135e-06,
+      "loss": 0.9138,
+      "step": 2250
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8344097512115443,
+      "learning_rate": 9.662713455794963e-06,
+      "loss": 0.7078,
+      "step": 2251
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6260554412785324,
+      "learning_rate": 9.66233912183061e-06,
+      "loss": 0.8155,
+      "step": 2252
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5133869767521866,
+      "learning_rate": 9.66196458751516e-06,
+      "loss": 0.715,
+      "step": 2253
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7038782929151224,
+      "learning_rate": 9.66158985286471e-06,
+      "loss": 0.8895,
+      "step": 2254
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6115958995576263,
+      "learning_rate": 9.661214917895361e-06,
+      "loss": 0.6775,
+      "step": 2255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5408238594157597,
+      "learning_rate": 9.660839782623226e-06,
+      "loss": 0.9237,
+      "step": 2256
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9547681470211766,
+      "learning_rate": 9.660464447064427e-06,
+      "loss": 0.7817,
+      "step": 2257
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7571828349849257,
+      "learning_rate": 9.660088911235088e-06,
+      "loss": 0.832,
+      "step": 2258
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9154885268044757,
+      "learning_rate": 9.659713175151352e-06,
+      "loss": 0.7852,
+      "step": 2259
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0452743363861796,
+      "learning_rate": 9.659337238829364e-06,
+      "loss": 0.7307,
+      "step": 2260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8324440397963786,
+      "learning_rate": 9.658961102285276e-06,
+      "loss": 0.7163,
+      "step": 2261
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1192486190719386,
+      "learning_rate": 9.658584765535254e-06,
+      "loss": 0.7604,
+      "step": 2262
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6303487923911493,
+      "learning_rate": 9.65820822859547e-06,
+      "loss": 0.869,
+      "step": 2263
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.715223891684336,
+      "learning_rate": 9.657831491482103e-06,
+      "loss": 0.9433,
+      "step": 2264
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.947103890891639,
+      "learning_rate": 9.657454554211344e-06,
+      "loss": 0.8025,
+      "step": 2265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9251627292232298,
+      "learning_rate": 9.657077416799392e-06,
+      "loss": 0.8699,
+      "step": 2266
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.872711148976995,
+      "learning_rate": 9.656700079262448e-06,
+      "loss": 0.7921,
+      "step": 2267
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8008483213577717,
+      "learning_rate": 9.656322541616734e-06,
+      "loss": 0.707,
+      "step": 2268
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6700919544113613,
+      "learning_rate": 9.655944803878467e-06,
+      "loss": 0.9165,
+      "step": 2269
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1352955837310938,
+      "learning_rate": 9.655566866063885e-06,
+      "loss": 0.7252,
+      "step": 2270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8196957420753848,
+      "learning_rate": 9.655188728189225e-06,
+      "loss": 0.7507,
+      "step": 2271
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.747067195738727,
+      "learning_rate": 9.654810390270737e-06,
+      "loss": 0.8015,
+      "step": 2272
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.903394569575593,
+      "learning_rate": 9.654431852324682e-06,
+      "loss": 0.8897,
+      "step": 2273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.806558966181426,
+      "learning_rate": 9.654053114367321e-06,
+      "loss": 0.7674,
+      "step": 2274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5515537168409075,
+      "learning_rate": 9.653674176414936e-06,
+      "loss": 0.7875,
+      "step": 2275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4305591696146136,
+      "learning_rate": 9.653295038483806e-06,
+      "loss": 0.8273,
+      "step": 2276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8727477219044626,
+      "learning_rate": 9.652915700590224e-06,
+      "loss": 0.7732,
+      "step": 2277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5703724010340852,
+      "learning_rate": 9.652536162750492e-06,
+      "loss": 0.8174,
+      "step": 2278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2629982521846905,
+      "learning_rate": 9.65215642498092e-06,
+      "loss": 0.6802,
+      "step": 2279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7909950569822901,
+      "learning_rate": 9.651776487297823e-06,
+      "loss": 0.7835,
+      "step": 2280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7793301945573459,
+      "learning_rate": 9.651396349717533e-06,
+      "loss": 0.8366,
+      "step": 2281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7599381982088405,
+      "learning_rate": 9.651016012256382e-06,
+      "loss": 0.7165,
+      "step": 2282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.544458569067566,
+      "learning_rate": 9.650635474930716e-06,
+      "loss": 0.8881,
+      "step": 2283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5371556497466332,
+      "learning_rate": 9.650254737756883e-06,
+      "loss": 0.727,
+      "step": 2284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9552939332740085,
+      "learning_rate": 9.64987380075125e-06,
+      "loss": 0.7647,
+      "step": 2285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0720541816367515,
+      "learning_rate": 9.649492663930182e-06,
+      "loss": 0.6909,
+      "step": 2286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9048536280920259,
+      "learning_rate": 9.64911132731006e-06,
+      "loss": 0.6445,
+      "step": 2287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9959666183074902,
+      "learning_rate": 9.648729790907271e-06,
+      "loss": 0.7217,
+      "step": 2288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8774118142787215,
+      "learning_rate": 9.648348054738208e-06,
+      "loss": 0.8566,
+      "step": 2289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0677267823501833,
+      "learning_rate": 9.647966118819278e-06,
+      "loss": 0.7181,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1742939718789955,
+      "learning_rate": 9.647583983166892e-06,
+      "loss": 0.8105,
+      "step": 2291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3551319789675143,
+      "learning_rate": 9.647201647797472e-06,
+      "loss": 0.7105,
+      "step": 2292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.563834554114766,
+      "learning_rate": 9.646819112727445e-06,
+      "loss": 0.8473,
+      "step": 2293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9059400433772093,
+      "learning_rate": 9.646436377973253e-06,
+      "loss": 0.8222,
+      "step": 2294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8693064045357464,
+      "learning_rate": 9.646053443551344e-06,
+      "loss": 0.8738,
+      "step": 2295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6718241489149053,
+      "learning_rate": 9.64567030947817e-06,
+      "loss": 0.7809,
+      "step": 2296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.511344012657334,
+      "learning_rate": 9.645286975770194e-06,
+      "loss": 0.823,
+      "step": 2297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6176669373682615,
+      "learning_rate": 9.644903442443894e-06,
+      "loss": 0.8009,
+      "step": 2298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0798124616318696,
+      "learning_rate": 9.644519709515746e-06,
+      "loss": 0.8604,
+      "step": 2299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2918457464432707,
+      "learning_rate": 9.644135777002244e-06,
+      "loss": 0.801,
+      "step": 2300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6807643226748445,
+      "learning_rate": 9.643751644919883e-06,
+      "loss": 0.7765,
+      "step": 2301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7583790524755978,
+      "learning_rate": 9.643367313285172e-06,
+      "loss": 0.8626,
+      "step": 2302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0364691462713624,
+      "learning_rate": 9.642982782114628e-06,
+      "loss": 0.8642,
+      "step": 2303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9660038621856513,
+      "learning_rate": 9.642598051424772e-06,
+      "loss": 0.7682,
+      "step": 2304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.176197300434573,
+      "learning_rate": 9.642213121232137e-06,
+      "loss": 0.7256,
+      "step": 2305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.928258882827929,
+      "learning_rate": 9.641827991553267e-06,
+      "loss": 0.9261,
+      "step": 2306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.197716806873791,
+      "learning_rate": 9.64144266240471e-06,
+      "loss": 0.9298,
+      "step": 2307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0890190455547533,
+      "learning_rate": 9.641057133803024e-06,
+      "loss": 0.7532,
+      "step": 2308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7815374833704993,
+      "learning_rate": 9.640671405764777e-06,
+      "loss": 0.7705,
+      "step": 2309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7076276657034497,
+      "learning_rate": 9.640285478306546e-06,
+      "loss": 0.8686,
+      "step": 2310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6851063560321875,
+      "learning_rate": 9.639899351444911e-06,
+      "loss": 0.8617,
+      "step": 2311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7664124227570341,
+      "learning_rate": 9.63951302519647e-06,
+      "loss": 0.8488,
+      "step": 2312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7036860303723003,
+      "learning_rate": 9.639126499577818e-06,
+      "loss": 0.8343,
+      "step": 2313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9614948287055263,
+      "learning_rate": 9.638739774605572e-06,
+      "loss": 0.8127,
+      "step": 2314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.631311351823667,
+      "learning_rate": 9.638352850296344e-06,
+      "loss": 0.8149,
+      "step": 2315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.663424935769926,
+      "learning_rate": 9.637965726666766e-06,
+      "loss": 0.885,
+      "step": 2316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.011091390763324,
+      "learning_rate": 9.63757840373347e-06,
+      "loss": 0.805,
+      "step": 2317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6676496791148274,
+      "learning_rate": 9.637190881513103e-06,
+      "loss": 0.8338,
+      "step": 2318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3355085265045257,
+      "learning_rate": 9.636803160022314e-06,
+      "loss": 0.8306,
+      "step": 2319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8711850594029416,
+      "learning_rate": 9.63641523927777e-06,
+      "loss": 0.7499,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6236448531014975,
+      "learning_rate": 9.636027119296133e-06,
+      "loss": 0.7836,
+      "step": 2321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.079818178393903,
+      "learning_rate": 9.63563880009409e-06,
+      "loss": 0.7952,
+      "step": 2322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8579175503712246,
+      "learning_rate": 9.63525028168832e-06,
+      "loss": 0.7566,
+      "step": 2323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.077892215056848,
+      "learning_rate": 9.634861564095525e-06,
+      "loss": 0.7212,
+      "step": 2324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2582481984810017,
+      "learning_rate": 9.634472647332405e-06,
+      "loss": 0.6799,
+      "step": 2325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.073821660257231,
+      "learning_rate": 9.634083531415672e-06,
+      "loss": 0.8757,
+      "step": 2326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9058150212020424,
+      "learning_rate": 9.633694216362051e-06,
+      "loss": 0.8423,
+      "step": 2327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7460234152151082,
+      "learning_rate": 9.633304702188271e-06,
+      "loss": 0.7301,
+      "step": 2328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6298191442692396,
+      "learning_rate": 9.632914988911066e-06,
+      "loss": 0.7164,
+      "step": 2329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.023692021222371,
+      "learning_rate": 9.632525076547187e-06,
+      "loss": 0.7543,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.287350914869732,
+      "learning_rate": 9.632134965113389e-06,
+      "loss": 0.7119,
+      "step": 2331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1246904615397475,
+      "learning_rate": 9.631744654626435e-06,
+      "loss": 0.7791,
+      "step": 2332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5173664870372996,
+      "learning_rate": 9.631354145103098e-06,
+      "loss": 0.7771,
+      "step": 2333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.207447004272542,
+      "learning_rate": 9.63096343656016e-06,
+      "loss": 0.5697,
+      "step": 2334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7229678287931405,
+      "learning_rate": 9.630572529014406e-06,
+      "loss": 0.7366,
+      "step": 2335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7603042206855004,
+      "learning_rate": 9.630181422482639e-06,
+      "loss": 0.7424,
+      "step": 2336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.816996528095039,
+      "learning_rate": 9.629790116981665e-06,
+      "loss": 0.7512,
+      "step": 2337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6564275609861805,
+      "learning_rate": 9.629398612528299e-06,
+      "loss": 0.7729,
+      "step": 2338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.849364622700429,
+      "learning_rate": 9.629006909139363e-06,
+      "loss": 0.7323,
+      "step": 2339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.547114318580256,
+      "learning_rate": 9.628615006831692e-06,
+      "loss": 0.8226,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8927951355359514,
+      "learning_rate": 9.628222905622124e-06,
+      "loss": 0.8027,
+      "step": 2341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.87517802219182,
+      "learning_rate": 9.627830605527513e-06,
+      "loss": 0.7858,
+      "step": 2342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.855971357535123,
+      "learning_rate": 9.627438106564711e-06,
+      "loss": 0.7738,
+      "step": 2343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7562064714554337,
+      "learning_rate": 9.62704540875059e-06,
+      "loss": 0.7569,
+      "step": 2344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6939795826492816,
+      "learning_rate": 9.626652512102021e-06,
+      "loss": 0.7715,
+      "step": 2345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0300281684987476,
+      "learning_rate": 9.62625941663589e-06,
+      "loss": 0.8974,
+      "step": 2346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1088915420100316,
+      "learning_rate": 9.625866122369089e-06,
+      "loss": 0.644,
+      "step": 2347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.64711899806777,
+      "learning_rate": 9.62547262931852e-06,
+      "loss": 0.8559,
+      "step": 2348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7515657252501522,
+      "learning_rate": 9.625078937501089e-06,
+      "loss": 0.8052,
+      "step": 2349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5430332497922523,
+      "learning_rate": 9.624685046933715e-06,
+      "loss": 0.7915,
+      "step": 2350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8422129599420534,
+      "learning_rate": 9.624290957633325e-06,
+      "loss": 0.8558,
+      "step": 2351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8457727570241054,
+      "learning_rate": 9.623896669616855e-06,
+      "loss": 0.8313,
+      "step": 2352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2390796232788013,
+      "learning_rate": 9.623502182901245e-06,
+      "loss": 0.8064,
+      "step": 2353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8332185573897406,
+      "learning_rate": 9.62310749750345e-06,
+      "loss": 0.8023,
+      "step": 2354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9452412908735595,
+      "learning_rate": 9.62271261344043e-06,
+      "loss": 0.8465,
+      "step": 2355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2266772013762668,
+      "learning_rate": 9.622317530729153e-06,
+      "loss": 0.6323,
+      "step": 2356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6740827469605708,
+      "learning_rate": 9.621922249386598e-06,
+      "loss": 0.8674,
+      "step": 2357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5308349489170978,
+      "learning_rate": 9.621526769429749e-06,
+      "loss": 0.7228,
+      "step": 2358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9001685524933873,
+      "learning_rate": 9.621131090875603e-06,
+      "loss": 0.7958,
+      "step": 2359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5935963550048322,
+      "learning_rate": 9.620735213741163e-06,
+      "loss": 0.8465,
+      "step": 2360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1803535574345076,
+      "learning_rate": 9.62033913804344e-06,
+      "loss": 0.8034,
+      "step": 2361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6758626655274733,
+      "learning_rate": 9.619942863799454e-06,
+      "loss": 0.7241,
+      "step": 2362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8411428985088611,
+      "learning_rate": 9.619546391026234e-06,
+      "loss": 0.8099,
+      "step": 2363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6198644543848046,
+      "learning_rate": 9.619149719740817e-06,
+      "loss": 0.7891,
+      "step": 2364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1866011980379867,
+      "learning_rate": 9.618752849960248e-06,
+      "loss": 0.6779,
+      "step": 2365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6925701517728924,
+      "learning_rate": 9.618355781701584e-06,
+      "loss": 0.8116,
+      "step": 2366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9688552569249305,
+      "learning_rate": 9.617958514981885e-06,
+      "loss": 0.8774,
+      "step": 2367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8995383644328774,
+      "learning_rate": 9.617561049818225e-06,
+      "loss": 0.7116,
+      "step": 2368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8715339652080145,
+      "learning_rate": 9.617163386227683e-06,
+      "loss": 0.8228,
+      "step": 2369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8896478228029854,
+      "learning_rate": 9.616765524227348e-06,
+      "loss": 0.7926,
+      "step": 2370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9112555642768474,
+      "learning_rate": 9.616367463834313e-06,
+      "loss": 0.8169,
+      "step": 2371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.644883890008973,
+      "learning_rate": 9.61596920506569e-06,
+      "loss": 0.8646,
+      "step": 2372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.645233094252938,
+      "learning_rate": 9.61557074793859e-06,
+      "loss": 0.7848,
+      "step": 2373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.685739767508645,
+      "learning_rate": 9.615172092470134e-06,
+      "loss": 0.7962,
+      "step": 2374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9423417167672075,
+      "learning_rate": 9.614773238677455e-06,
+      "loss": 0.8612,
+      "step": 2375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5759899689869992,
+      "learning_rate": 9.614374186577693e-06,
+      "loss": 0.6948,
+      "step": 2376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6873358010599566,
+      "learning_rate": 9.613974936187993e-06,
+      "loss": 0.6595,
+      "step": 2377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7784502959004176,
+      "learning_rate": 9.613575487525517e-06,
+      "loss": 0.877,
+      "step": 2378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7763300424743638,
+      "learning_rate": 9.613175840607428e-06,
+      "loss": 0.8474,
+      "step": 2379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7504423615874394,
+      "learning_rate": 9.612775995450896e-06,
+      "loss": 0.8558,
+      "step": 2380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7726123588844622,
+      "learning_rate": 9.612375952073108e-06,
+      "loss": 0.7897,
+      "step": 2381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8045966742674036,
+      "learning_rate": 9.611975710491252e-06,
+      "loss": 0.7163,
+      "step": 2382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1140238157882278,
+      "learning_rate": 9.611575270722529e-06,
+      "loss": 0.7485,
+      "step": 2383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.980746380230763,
+      "learning_rate": 9.611174632784147e-06,
+      "loss": 0.8193,
+      "step": 2384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6314425806668496,
+      "learning_rate": 9.61077379669332e-06,
+      "loss": 0.8226,
+      "step": 2385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9092248671359797,
+      "learning_rate": 9.610372762467275e-06,
+      "loss": 0.9913,
+      "step": 2386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8947685730294561,
+      "learning_rate": 9.609971530123243e-06,
+      "loss": 0.8085,
+      "step": 2387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.757939083464733,
+      "learning_rate": 9.60957009967847e-06,
+      "loss": 0.7969,
+      "step": 2388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.593799827428896,
+      "learning_rate": 9.609168471150202e-06,
+      "loss": 0.795,
+      "step": 2389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7940196738848773,
+      "learning_rate": 9.6087666445557e-06,
+      "loss": 0.8668,
+      "step": 2390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7731268458383636,
+      "learning_rate": 9.60836461991223e-06,
+      "loss": 0.8716,
+      "step": 2391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6304420464460738,
+      "learning_rate": 9.607962397237071e-06,
+      "loss": 0.8288,
+      "step": 2392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7302853585928475,
+      "learning_rate": 9.607559976547504e-06,
+      "loss": 0.7114,
+      "step": 2393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5076639057893897,
+      "learning_rate": 9.607157357860823e-06,
+      "loss": 0.87,
+      "step": 2394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6379603629761519,
+      "learning_rate": 9.606754541194329e-06,
+      "loss": 0.682,
+      "step": 2395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5237030808634797,
+      "learning_rate": 9.606351526565336e-06,
+      "loss": 0.7129,
+      "step": 2396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4055336522924582,
+      "learning_rate": 9.605948313991156e-06,
+      "loss": 0.6972,
+      "step": 2397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0017919260189436,
+      "learning_rate": 9.60554490348912e-06,
+      "loss": 0.896,
+      "step": 2398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.778445558540459,
+      "learning_rate": 9.605141295076561e-06,
+      "loss": 0.7767,
+      "step": 2399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2747552437642682,
+      "learning_rate": 9.604737488770826e-06,
+      "loss": 0.5617,
+      "step": 2400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6490570115950063,
+      "learning_rate": 9.604333484589266e-06,
+      "loss": 0.6429,
+      "step": 2401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.020342951783659,
+      "learning_rate": 9.60392928254924e-06,
+      "loss": 0.9382,
+      "step": 2402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.796075646649933,
+      "learning_rate": 9.60352488266812e-06,
+      "loss": 0.8201,
+      "step": 2403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6687200413096912,
+      "learning_rate": 9.603120284963284e-06,
+      "loss": 0.7848,
+      "step": 2404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0672496731734413,
+      "learning_rate": 9.602715489452117e-06,
+      "loss": 0.7688,
+      "step": 2405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6550121092838317,
+      "learning_rate": 9.602310496152014e-06,
+      "loss": 0.7568,
+      "step": 2406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1465505258822337,
+      "learning_rate": 9.60190530508038e-06,
+      "loss": 0.7065,
+      "step": 2407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.177026687036246,
+      "learning_rate": 9.601499916254626e-06,
+      "loss": 0.8306,
+      "step": 2408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7760099534639096,
+      "learning_rate": 9.601094329692173e-06,
+      "loss": 0.7199,
+      "step": 2409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6396882748801869,
+      "learning_rate": 9.600688545410449e-06,
+      "loss": 0.7393,
+      "step": 2410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1586681617270664,
+      "learning_rate": 9.60028256342689e-06,
+      "loss": 0.7699,
+      "step": 2411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6907847382258905,
+      "learning_rate": 9.599876383758946e-06,
+      "loss": 0.8336,
+      "step": 2412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4259904096654166,
+      "learning_rate": 9.599470006424069e-06,
+      "loss": 0.668,
+      "step": 2413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7284403509248591,
+      "learning_rate": 9.599063431439721e-06,
+      "loss": 0.7718,
+      "step": 2414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.995710152258815,
+      "learning_rate": 9.598656658823378e-06,
+      "loss": 0.832,
+      "step": 2415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8001619390950792,
+      "learning_rate": 9.598249688592514e-06,
+      "loss": 0.8164,
+      "step": 2416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.008111139692204,
+      "learning_rate": 9.597842520764619e-06,
+      "loss": 0.7915,
+      "step": 2417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9548491314361145,
+      "learning_rate": 9.597435155357192e-06,
+      "loss": 0.8075,
+      "step": 2418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6626209260087408,
+      "learning_rate": 9.597027592387739e-06,
+      "loss": 0.8275,
+      "step": 2419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5851235729742152,
+      "learning_rate": 9.59661983187377e-06,
+      "loss": 0.7647,
+      "step": 2420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.857273462819878,
+      "learning_rate": 9.59621187383281e-06,
+      "loss": 0.9184,
+      "step": 2421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9399521265939599,
+      "learning_rate": 9.595803718282391e-06,
+      "loss": 0.8118,
+      "step": 2422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5288494907848582,
+      "learning_rate": 9.59539536524005e-06,
+      "loss": 0.9429,
+      "step": 2423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.934243965914035,
+      "learning_rate": 9.594986814723335e-06,
+      "loss": 0.7194,
+      "step": 2424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6695435826263323,
+      "learning_rate": 9.594578066749804e-06,
+      "loss": 0.8639,
+      "step": 2425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.653562382194372,
+      "learning_rate": 9.59416912133702e-06,
+      "loss": 0.8106,
+      "step": 2426
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.662693468279136,
+      "learning_rate": 9.593759978502556e-06,
+      "loss": 0.8565,
+      "step": 2427
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6496807816112964,
+      "learning_rate": 9.593350638263998e-06,
+      "loss": 0.856,
+      "step": 2428
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6035240353703217,
+      "learning_rate": 9.59294110063893e-06,
+      "loss": 0.8776,
+      "step": 2429
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6962020103434803,
+      "learning_rate": 9.592531365644957e-06,
+      "loss": 0.7429,
+      "step": 2430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.830243412642451,
+      "learning_rate": 9.592121433299682e-06,
+      "loss": 0.8144,
+      "step": 2431
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0315993260406495,
+      "learning_rate": 9.59171130362072e-06,
+      "loss": 0.783,
+      "step": 2432
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.67957117411081,
+      "learning_rate": 9.591300976625697e-06,
+      "loss": 0.8279,
+      "step": 2433
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6484171379798924,
+      "learning_rate": 9.590890452332249e-06,
+      "loss": 0.7303,
+      "step": 2434
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5493053315354548,
+      "learning_rate": 9.59047973075801e-06,
+      "loss": 0.6983,
+      "step": 2435
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0964695036554715,
+      "learning_rate": 9.590068811920637e-06,
+      "loss": 0.9133,
+      "step": 2436
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2695780525097504,
+      "learning_rate": 9.589657695837781e-06,
+      "loss": 0.7556,
+      "step": 2437
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.03215753071385,
+      "learning_rate": 9.589246382527113e-06,
+      "loss": 0.7491,
+      "step": 2438
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.143791720927441,
+      "learning_rate": 9.588834872006308e-06,
+      "loss": 0.7612,
+      "step": 2439
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7017670093774868,
+      "learning_rate": 9.588423164293049e-06,
+      "loss": 0.7878,
+      "step": 2440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8053206217709497,
+      "learning_rate": 9.588011259405028e-06,
+      "loss": 0.6825,
+      "step": 2441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9349538958142372,
+      "learning_rate": 9.587599157359943e-06,
+      "loss": 0.7945,
+      "step": 2442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9264035451930894,
+      "learning_rate": 9.587186858175507e-06,
+      "loss": 0.8744,
+      "step": 2443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3453295264935348,
+      "learning_rate": 9.586774361869436e-06,
+      "loss": 0.7847,
+      "step": 2444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.869967795031241,
+      "learning_rate": 9.586361668459452e-06,
+      "loss": 0.9392,
+      "step": 2445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.635222223130135,
+      "learning_rate": 9.585948777963296e-06,
+      "loss": 0.7891,
+      "step": 2446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.711712854284562,
+      "learning_rate": 9.585535690398706e-06,
+      "loss": 0.8502,
+      "step": 2447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4443720446492376,
+      "learning_rate": 9.585122405783435e-06,
+      "loss": 0.6769,
+      "step": 2448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8679685455758461,
+      "learning_rate": 9.584708924135245e-06,
+      "loss": 0.7787,
+      "step": 2449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5640187445333635,
+      "learning_rate": 9.584295245471898e-06,
+      "loss": 0.7579,
+      "step": 2450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1605270557740726,
+      "learning_rate": 9.583881369811176e-06,
+      "loss": 0.9484,
+      "step": 2451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5334529350994675,
+      "learning_rate": 9.583467297170863e-06,
+      "loss": 0.7915,
+      "step": 2452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6465516190695058,
+      "learning_rate": 9.583053027568754e-06,
+      "loss": 0.8128,
+      "step": 2453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6555572343402047,
+      "learning_rate": 9.582638561022646e-06,
+      "loss": 0.7276,
+      "step": 2454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5922361742560056,
+      "learning_rate": 9.582223897550356e-06,
+      "loss": 0.7286,
+      "step": 2455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6186108492687403,
+      "learning_rate": 9.581809037169698e-06,
+      "loss": 0.7496,
+      "step": 2456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1008857063473125,
+      "learning_rate": 9.581393979898502e-06,
+      "loss": 0.9192,
+      "step": 2457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4677847563069337,
+      "learning_rate": 9.580978725754604e-06,
+      "loss": 0.6816,
+      "step": 2458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8738982635305423,
+      "learning_rate": 9.580563274755848e-06,
+      "loss": 0.8074,
+      "step": 2459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0850510622374085,
+      "learning_rate": 9.580147626920084e-06,
+      "loss": 0.7121,
+      "step": 2460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0272424968602873,
+      "learning_rate": 9.579731782265178e-06,
+      "loss": 0.9231,
+      "step": 2461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5626474418107061,
+      "learning_rate": 9.579315740808998e-06,
+      "loss": 0.8228,
+      "step": 2462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5238692608485143,
+      "learning_rate": 9.578899502569421e-06,
+      "loss": 0.6769,
+      "step": 2463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.103238213602793,
+      "learning_rate": 9.578483067564335e-06,
+      "loss": 0.6504,
+      "step": 2464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8528519370389798,
+      "learning_rate": 9.578066435811636e-06,
+      "loss": 0.8227,
+      "step": 2465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8435110420450767,
+      "learning_rate": 9.577649607329225e-06,
+      "loss": 0.817,
+      "step": 2466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2198119290735854,
+      "learning_rate": 9.577232582135014e-06,
+      "loss": 0.6527,
+      "step": 2467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9585668505697127,
+      "learning_rate": 9.576815360246926e-06,
+      "loss": 0.7791,
+      "step": 2468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6185754226168665,
+      "learning_rate": 9.576397941682891e-06,
+      "loss": 0.7772,
+      "step": 2469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2367530418051709,
+      "learning_rate": 9.575980326460841e-06,
+      "loss": 0.7971,
+      "step": 2470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7385055372344729,
+      "learning_rate": 9.575562514598727e-06,
+      "loss": 0.758,
+      "step": 2471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6314526710077495,
+      "learning_rate": 9.575144506114501e-06,
+      "loss": 0.7517,
+      "step": 2472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1722236789033553,
+      "learning_rate": 9.574726301026127e-06,
+      "loss": 0.7264,
+      "step": 2473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4660755076907477,
+      "learning_rate": 9.574307899351574e-06,
+      "loss": 0.7246,
+      "step": 2474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6228484986979488,
+      "learning_rate": 9.573889301108823e-06,
+      "loss": 0.8565,
+      "step": 2475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.147344967578374,
+      "learning_rate": 9.573470506315864e-06,
+      "loss": 0.704,
+      "step": 2476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6459949291287799,
+      "learning_rate": 9.57305151499069e-06,
+      "loss": 0.7542,
+      "step": 2477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0109077071514307,
+      "learning_rate": 9.572632327151309e-06,
+      "loss": 0.6642,
+      "step": 2478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5358263221000008,
+      "learning_rate": 9.572212942815734e-06,
+      "loss": 0.8339,
+      "step": 2479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.780347401213516,
+      "learning_rate": 9.571793362001983e-06,
+      "loss": 0.8057,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.895447158456405,
+      "learning_rate": 9.57137358472809e-06,
+      "loss": 0.7413,
+      "step": 2481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.673490430176283,
+      "learning_rate": 9.570953611012092e-06,
+      "loss": 0.8224,
+      "step": 2482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5518949004687363,
+      "learning_rate": 9.570533440872041e-06,
+      "loss": 0.7586,
+      "step": 2483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9548481384868177,
+      "learning_rate": 9.570113074325986e-06,
+      "loss": 0.7929,
+      "step": 2484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9246818018600416,
+      "learning_rate": 9.569692511391995e-06,
+      "loss": 0.7718,
+      "step": 2485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7894066514349272,
+      "learning_rate": 9.569271752088139e-06,
+      "loss": 0.7236,
+      "step": 2486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8352902555438277,
+      "learning_rate": 9.5688507964325e-06,
+      "loss": 0.7834,
+      "step": 2487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8631569074677907,
+      "learning_rate": 9.568429644443167e-06,
+      "loss": 0.8366,
+      "step": 2488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7456594186930952,
+      "learning_rate": 9.568008296138238e-06,
+      "loss": 0.7723,
+      "step": 2489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7305136207458183,
+      "learning_rate": 9.567586751535818e-06,
+      "loss": 0.9378,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8457057066570728,
+      "learning_rate": 9.567165010654022e-06,
+      "loss": 0.7923,
+      "step": 2491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6347235160342877,
+      "learning_rate": 9.566743073510976e-06,
+      "loss": 0.8543,
+      "step": 2492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9610597865043533,
+      "learning_rate": 9.566320940124809e-06,
+      "loss": 0.7206,
+      "step": 2493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2926912735112874,
+      "learning_rate": 9.565898610513661e-06,
+      "loss": 0.7957,
+      "step": 2494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6596003057439184,
+      "learning_rate": 9.565476084695681e-06,
+      "loss": 0.8264,
+      "step": 2495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8868814037340649,
+      "learning_rate": 9.565053362689026e-06,
+      "loss": 0.7908,
+      "step": 2496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6921152162719146,
+      "learning_rate": 9.56463044451186e-06,
+      "loss": 0.8161,
+      "step": 2497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8531497018441265,
+      "learning_rate": 9.564207330182359e-06,
+      "loss": 0.7456,
+      "step": 2498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.178923279410791,
+      "learning_rate": 9.563784019718704e-06,
+      "loss": 0.6762,
+      "step": 2499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7042371210095975,
+      "learning_rate": 9.563360513139085e-06,
+      "loss": 0.7918,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8154172985891261,
+      "learning_rate": 9.562936810461701e-06,
+      "loss": 0.8317,
+      "step": 2501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2447760864454,
+      "learning_rate": 9.562512911704761e-06,
+      "loss": 0.8858,
+      "step": 2502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3474521115059086,
+      "learning_rate": 9.56208881688648e-06,
+      "loss": 0.6541,
+      "step": 2503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8781584874622346,
+      "learning_rate": 9.561664526025082e-06,
+      "loss": 0.849,
+      "step": 2504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5064293889711493,
+      "learning_rate": 9.5612400391388e-06,
+      "loss": 0.7606,
+      "step": 2505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.66372577639925,
+      "learning_rate": 9.560815356245875e-06,
+      "loss": 0.804,
+      "step": 2506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8067220657838896,
+      "learning_rate": 9.560390477364555e-06,
+      "loss": 0.8546,
+      "step": 2507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6371130237492422,
+      "learning_rate": 9.559965402513103e-06,
+      "loss": 0.7164,
+      "step": 2508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7011569000234703,
+      "learning_rate": 9.55954013170978e-06,
+      "loss": 0.776,
+      "step": 2509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.207546327483192,
+      "learning_rate": 9.559114664972864e-06,
+      "loss": 0.9293,
+      "step": 2510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9846973139977433,
+      "learning_rate": 9.558689002320635e-06,
+      "loss": 0.9341,
+      "step": 2511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.954604037149035,
+      "learning_rate": 9.55826314377139e-06,
+      "loss": 0.9012,
+      "step": 2512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8624858154926096,
+      "learning_rate": 9.557837089343424e-06,
+      "loss": 0.9228,
+      "step": 2513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.91004293587829,
+      "learning_rate": 9.557410839055047e-06,
+      "loss": 0.7775,
+      "step": 2514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.327605737136217,
+      "learning_rate": 9.556984392924577e-06,
+      "loss": 0.5377,
+      "step": 2515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.701341096030783,
+      "learning_rate": 9.55655775097034e-06,
+      "loss": 0.7997,
+      "step": 2516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7592375343448001,
+      "learning_rate": 9.556130913210666e-06,
+      "loss": 0.8415,
+      "step": 2517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.943095687012894,
+      "learning_rate": 9.5557038796639e-06,
+      "loss": 0.7507,
+      "step": 2518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2694774474170263,
+      "learning_rate": 9.555276650348393e-06,
+      "loss": 0.8365,
+      "step": 2519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.135294562172385,
+      "learning_rate": 9.554849225282503e-06,
+      "loss": 0.79,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7557380917599865,
+      "learning_rate": 9.554421604484599e-06,
+      "loss": 0.9534,
+      "step": 2521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8242805842660543,
+      "learning_rate": 9.553993787973053e-06,
+      "loss": 0.8138,
+      "step": 2522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.518029844808416,
+      "learning_rate": 9.553565775766253e-06,
+      "loss": 0.779,
+      "step": 2523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5680417230401769,
+      "learning_rate": 9.55313756788259e-06,
+      "loss": 0.804,
+      "step": 2524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.368697922582615,
+      "learning_rate": 9.552709164340466e-06,
+      "loss": 0.8824,
+      "step": 2525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6173965277113695,
+      "learning_rate": 9.552280565158289e-06,
+      "loss": 0.767,
+      "step": 2526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1421592540816774,
+      "learning_rate": 9.551851770354477e-06,
+      "loss": 0.8048,
+      "step": 2527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9236595363698887,
+      "learning_rate": 9.551422779947458e-06,
+      "loss": 0.7391,
+      "step": 2528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8758459105008405,
+      "learning_rate": 9.550993593955665e-06,
+      "loss": 0.77,
+      "step": 2529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8179295767738248,
+      "learning_rate": 9.550564212397541e-06,
+      "loss": 0.8357,
+      "step": 2530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.666905998121751,
+      "learning_rate": 9.55013463529154e-06,
+      "loss": 0.8179,
+      "step": 2531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3224608366867183,
+      "learning_rate": 9.54970486265612e-06,
+      "loss": 0.799,
+      "step": 2532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0167665348324597,
+      "learning_rate": 9.549274894509748e-06,
+      "loss": 0.7832,
+      "step": 2533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.848805397041797,
+      "learning_rate": 9.548844730870903e-06,
+      "loss": 0.8178,
+      "step": 2534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3741313686533134,
+      "learning_rate": 9.548414371758068e-06,
+      "loss": 0.75,
+      "step": 2535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6634043823992042,
+      "learning_rate": 9.547983817189738e-06,
+      "loss": 0.8009,
+      "step": 2536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2548979062287753,
+      "learning_rate": 9.547553067184413e-06,
+      "loss": 0.6732,
+      "step": 2537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.099078077264673,
+      "learning_rate": 9.547122121760607e-06,
+      "loss": 0.7838,
+      "step": 2538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6844285717705405,
+      "learning_rate": 9.546690980936836e-06,
+      "loss": 0.7903,
+      "step": 2539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7645444481208523,
+      "learning_rate": 9.546259644731628e-06,
+      "loss": 0.8341,
+      "step": 2540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.977200249510224,
+      "learning_rate": 9.545828113163516e-06,
+      "loss": 0.9607,
+      "step": 2541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8915785836121595,
+      "learning_rate": 9.545396386251048e-06,
+      "loss": 0.8115,
+      "step": 2542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7164887004792508,
+      "learning_rate": 9.544964464012774e-06,
+      "loss": 0.8192,
+      "step": 2543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7685298894631443,
+      "learning_rate": 9.544532346467254e-06,
+      "loss": 0.861,
+      "step": 2544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.372201178838868,
+      "learning_rate": 9.544100033633057e-06,
+      "loss": 0.7656,
+      "step": 2545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.746056643794335,
+      "learning_rate": 9.543667525528762e-06,
+      "loss": 0.9052,
+      "step": 2546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3298552699025223,
+      "learning_rate": 9.543234822172956e-06,
+      "loss": 0.7127,
+      "step": 2547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9022448626429787,
+      "learning_rate": 9.542801923584228e-06,
+      "loss": 0.7694,
+      "step": 2548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8037725872184363,
+      "learning_rate": 9.542368829781186e-06,
+      "loss": 0.7474,
+      "step": 2549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.103750441530473,
+      "learning_rate": 9.541935540782439e-06,
+      "loss": 0.7602,
+      "step": 2550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7043982053033295,
+      "learning_rate": 9.541502056606606e-06,
+      "loss": 0.7165,
+      "step": 2551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8616947773457697,
+      "learning_rate": 9.541068377272315e-06,
+      "loss": 0.8999,
+      "step": 2552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.642525404691411,
+      "learning_rate": 9.540634502798203e-06,
+      "loss": 0.7929,
+      "step": 2553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9218925978255175,
+      "learning_rate": 9.540200433202913e-06,
+      "loss": 0.7801,
+      "step": 2554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7138641448491245,
+      "learning_rate": 9.5397661685051e-06,
+      "loss": 0.7847,
+      "step": 2555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8102502005195356,
+      "learning_rate": 9.539331708723423e-06,
+      "loss": 0.8428,
+      "step": 2556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8309734653667094,
+      "learning_rate": 9.538897053876553e-06,
+      "loss": 0.7496,
+      "step": 2557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8995464386944236,
+      "learning_rate": 9.538462203983169e-06,
+      "loss": 0.7703,
+      "step": 2558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0166455998329353,
+      "learning_rate": 9.538027159061955e-06,
+      "loss": 0.5939,
+      "step": 2559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.674916350308103,
+      "learning_rate": 9.537591919131608e-06,
+      "loss": 0.7519,
+      "step": 2560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.03554377064545,
+      "learning_rate": 9.53715648421083e-06,
+      "loss": 0.8894,
+      "step": 2561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.228884688844856,
+      "learning_rate": 9.536720854318333e-06,
+      "loss": 0.7421,
+      "step": 2562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7423718306195977,
+      "learning_rate": 9.536285029472838e-06,
+      "loss": 0.9346,
+      "step": 2563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.704174397702617,
+      "learning_rate": 9.535849009693072e-06,
+      "loss": 0.7597,
+      "step": 2564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7415799902935254,
+      "learning_rate": 9.535412794997772e-06,
+      "loss": 0.7193,
+      "step": 2565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5400560685136073,
+      "learning_rate": 9.534976385405683e-06,
+      "loss": 0.8051,
+      "step": 2566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.105578878635876,
+      "learning_rate": 9.53453978093556e-06,
+      "loss": 0.6655,
+      "step": 2567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4748619772088838,
+      "learning_rate": 9.534102981606162e-06,
+      "loss": 0.6763,
+      "step": 2568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.419610785395681,
+      "learning_rate": 9.533665987436262e-06,
+      "loss": 0.8161,
+      "step": 2569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0057674521961126,
+      "learning_rate": 9.533228798444637e-06,
+      "loss": 0.7336,
+      "step": 2570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.250227290410125,
+      "learning_rate": 9.532791414650075e-06,
+      "loss": 0.6399,
+      "step": 2571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4960381843473405,
+      "learning_rate": 9.53235383607137e-06,
+      "loss": 0.769,
+      "step": 2572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5612006761544,
+      "learning_rate": 9.531916062727328e-06,
+      "loss": 0.7058,
+      "step": 2573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3222604745113555,
+      "learning_rate": 9.531478094636758e-06,
+      "loss": 0.782,
+      "step": 2574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7357516533275419,
+      "learning_rate": 9.531039931818485e-06,
+      "loss": 0.8356,
+      "step": 2575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7672814332353053,
+      "learning_rate": 9.530601574291331e-06,
+      "loss": 0.8673,
+      "step": 2576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1163900511110378,
+      "learning_rate": 9.530163022074139e-06,
+      "loss": 0.7314,
+      "step": 2577
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.378345426770908,
+      "learning_rate": 9.529724275185753e-06,
+      "loss": 0.903,
+      "step": 2578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3206568960418887,
+      "learning_rate": 9.529285333645027e-06,
+      "loss": 0.6769,
+      "step": 2579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9740504181162248,
+      "learning_rate": 9.528846197470822e-06,
+      "loss": 0.8028,
+      "step": 2580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.753102022205658,
+      "learning_rate": 9.528406866682009e-06,
+      "loss": 0.7958,
+      "step": 2581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6673841287437607,
+      "learning_rate": 9.527967341297471e-06,
+      "loss": 0.7301,
+      "step": 2582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5864592310348915,
+      "learning_rate": 9.527527621336087e-06,
+      "loss": 0.7475,
+      "step": 2583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1504331547310067,
+      "learning_rate": 9.527087706816762e-06,
+      "loss": 0.5943,
+      "step": 2584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.015104134132117,
+      "learning_rate": 9.526647597758394e-06,
+      "loss": 0.8582,
+      "step": 2585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5787365100669626,
+      "learning_rate": 9.526207294179897e-06,
+      "loss": 0.7128,
+      "step": 2586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.92281701810294,
+      "learning_rate": 9.525766796100194e-06,
+      "loss": 0.8753,
+      "step": 2587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7810528953604612,
+      "learning_rate": 9.52532610353821e-06,
+      "loss": 0.7387,
+      "step": 2588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6077461888437272,
+      "learning_rate": 9.524885216512887e-06,
+      "loss": 0.7882,
+      "step": 2589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.78778677886561,
+      "learning_rate": 9.524444135043168e-06,
+      "loss": 0.8289,
+      "step": 2590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.048045557401015,
+      "learning_rate": 9.524002859148006e-06,
+      "loss": 0.8099,
+      "step": 2591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.638177634647703,
+      "learning_rate": 9.523561388846368e-06,
+      "loss": 0.7894,
+      "step": 2592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9146089984562171,
+      "learning_rate": 9.523119724157222e-06,
+      "loss": 0.6446,
+      "step": 2593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6966347776958788,
+      "learning_rate": 9.522677865099548e-06,
+      "loss": 0.8466,
+      "step": 2594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7463912534503219,
+      "learning_rate": 9.522235811692335e-06,
+      "loss": 0.6922,
+      "step": 2595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7982359142677302,
+      "learning_rate": 9.521793563954575e-06,
+      "loss": 0.7247,
+      "step": 2596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2391962232756033,
+      "learning_rate": 9.521351121905278e-06,
+      "loss": 0.6805,
+      "step": 2597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6634756155278876,
+      "learning_rate": 9.520908485563452e-06,
+      "loss": 0.7697,
+      "step": 2598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0480352084097486,
+      "learning_rate": 9.520465654948119e-06,
+      "loss": 0.8267,
+      "step": 2599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7120839660653795,
+      "learning_rate": 9.52002263007831e-06,
+      "loss": 0.8842,
+      "step": 2600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6978214147568078,
+      "learning_rate": 9.519579410973061e-06,
+      "loss": 0.7083,
+      "step": 2601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8764893025157536,
+      "learning_rate": 9.51913599765142e-06,
+      "loss": 0.9489,
+      "step": 2602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.876482231834845,
+      "learning_rate": 9.518692390132441e-06,
+      "loss": 0.7738,
+      "step": 2603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.600915676263327,
+      "learning_rate": 9.518248588435185e-06,
+      "loss": 0.7753,
+      "step": 2604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2363036318904792,
+      "learning_rate": 9.517804592578725e-06,
+      "loss": 0.6558,
+      "step": 2605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6901089969801817,
+      "learning_rate": 9.517360402582141e-06,
+      "loss": 0.747,
+      "step": 2606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9772417671754969,
+      "learning_rate": 9.51691601846452e-06,
+      "loss": 0.6925,
+      "step": 2607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.059475170638417,
+      "learning_rate": 9.516471440244956e-06,
+      "loss": 0.8326,
+      "step": 2608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8283790722750808,
+      "learning_rate": 9.516026667942557e-06,
+      "loss": 0.9565,
+      "step": 2609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7200863922390448,
+      "learning_rate": 9.515581701576433e-06,
+      "loss": 0.7489,
+      "step": 2610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.668349860857229,
+      "learning_rate": 9.515136541165708e-06,
+      "loss": 0.8681,
+      "step": 2611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7744542971173867,
+      "learning_rate": 9.514691186729509e-06,
+      "loss": 0.7488,
+      "step": 2612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7032672948336782,
+      "learning_rate": 9.514245638286976e-06,
+      "loss": 0.8941,
+      "step": 2613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6688225016140386,
+      "learning_rate": 9.513799895857252e-06,
+      "loss": 0.6904,
+      "step": 2614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6694590432964604,
+      "learning_rate": 9.513353959459496e-06,
+      "loss": 0.8094,
+      "step": 2615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6453807124370237,
+      "learning_rate": 9.512907829112867e-06,
+      "loss": 0.598,
+      "step": 2616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.120809877864745,
+      "learning_rate": 9.512461504836538e-06,
+      "loss": 0.8126,
+      "step": 2617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.770151414847364,
+      "learning_rate": 9.512014986649691e-06,
+      "loss": 0.7158,
+      "step": 2618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2658868956584448,
+      "learning_rate": 9.511568274571508e-06,
+      "loss": 0.6749,
+      "step": 2619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7631882586149068,
+      "learning_rate": 9.51112136862119e-06,
+      "loss": 0.79,
+      "step": 2620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7595939578104098,
+      "learning_rate": 9.51067426881794e-06,
+      "loss": 0.8768,
+      "step": 2621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0754924646128186,
+      "learning_rate": 9.510226975180969e-06,
+      "loss": 0.8195,
+      "step": 2622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7209140595738377,
+      "learning_rate": 9.509779487729502e-06,
+      "loss": 0.7691,
+      "step": 2623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9358107336834174,
+      "learning_rate": 9.509331806482767e-06,
+      "loss": 0.7698,
+      "step": 2624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9274231529627017,
+      "learning_rate": 9.50888393146e-06,
+      "loss": 0.8502,
+      "step": 2625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7994053995610004,
+      "learning_rate": 9.508435862680449e-06,
+      "loss": 0.7956,
+      "step": 2626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.837519208264157,
+      "learning_rate": 9.507987600163369e-06,
+      "loss": 0.6969,
+      "step": 2627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7285331101656467,
+      "learning_rate": 9.507539143928021e-06,
+      "loss": 0.8267,
+      "step": 2628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7427807320774287,
+      "learning_rate": 9.507090493993677e-06,
+      "loss": 0.7206,
+      "step": 2629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0168564312726516,
+      "learning_rate": 9.506641650379617e-06,
+      "loss": 0.5875,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0163333282325646,
+      "learning_rate": 9.50619261310513e-06,
+      "loss": 0.7615,
+      "step": 2631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1028775367536836,
+      "learning_rate": 9.50574338218951e-06,
+      "loss": 0.7033,
+      "step": 2632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8526807595534263,
+      "learning_rate": 9.50529395765206e-06,
+      "loss": 0.8828,
+      "step": 2633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2464252106990907,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.6067,
+      "step": 2634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8683601328996922,
+      "learning_rate": 9.504394527788939e-06,
+      "loss": 0.7785,
+      "step": 2635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8063135012628044,
+      "learning_rate": 9.503944522501915e-06,
+      "loss": 0.8158,
+      "step": 2636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9376882784739629,
+      "learning_rate": 9.503494323670366e-06,
+      "loss": 0.9078,
+      "step": 2637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1257307302871524,
+      "learning_rate": 9.503043931313637e-06,
+      "loss": 0.681,
+      "step": 2638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.731153154477238,
+      "learning_rate": 9.502593345451078e-06,
+      "loss": 0.7587,
+      "step": 2639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.242887598211898,
+      "learning_rate": 9.502142566102058e-06,
+      "loss": 0.8171,
+      "step": 2640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6829786989613211,
+      "learning_rate": 9.501691593285943e-06,
+      "loss": 0.8884,
+      "step": 2641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7193543543231375,
+      "learning_rate": 9.501240427022115e-06,
+      "loss": 0.88,
+      "step": 2642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9685243897454896,
+      "learning_rate": 9.500789067329963e-06,
+      "loss": 0.7742,
+      "step": 2643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6191251354961425,
+      "learning_rate": 9.500337514228878e-06,
+      "loss": 0.6451,
+      "step": 2644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6475498271295272,
+      "learning_rate": 9.49988576773827e-06,
+      "loss": 0.8107,
+      "step": 2645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7212555029166803,
+      "learning_rate": 9.499433827877547e-06,
+      "loss": 0.6633,
+      "step": 2646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6009207715270808,
+      "learning_rate": 9.498981694666133e-06,
+      "loss": 0.7075,
+      "step": 2647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7550222909457258,
+      "learning_rate": 9.498529368123455e-06,
+      "loss": 0.7816,
+      "step": 2648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8122186942215612,
+      "learning_rate": 9.49807684826895e-06,
+      "loss": 0.7909,
+      "step": 2649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8867683361971377,
+      "learning_rate": 9.497624135122068e-06,
+      "loss": 0.7699,
+      "step": 2650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9275123678810653,
+      "learning_rate": 9.497171228702257e-06,
+      "loss": 0.9369,
+      "step": 2651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5154936525952563,
+      "learning_rate": 9.496718129028986e-06,
+      "loss": 0.7736,
+      "step": 2652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7643955605806776,
+      "learning_rate": 9.49626483612172e-06,
+      "loss": 0.8677,
+      "step": 2653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9997037881657533,
+      "learning_rate": 9.495811349999941e-06,
+      "loss": 0.7847,
+      "step": 2654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9657126970905128,
+      "learning_rate": 9.495357670683135e-06,
+      "loss": 0.7603,
+      "step": 2655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6241045731711012,
+      "learning_rate": 9.494903798190798e-06,
+      "loss": 0.7095,
+      "step": 2656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.711374791690865,
+      "learning_rate": 9.494449732542435e-06,
+      "loss": 0.9741,
+      "step": 2657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6008332387473665,
+      "learning_rate": 9.493995473757558e-06,
+      "loss": 0.7479,
+      "step": 2658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9608653323980374,
+      "learning_rate": 9.493541021855685e-06,
+      "loss": 0.8966,
+      "step": 2659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0373761265026147,
+      "learning_rate": 9.493086376856346e-06,
+      "loss": 0.9113,
+      "step": 2660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.637558484394247,
+      "learning_rate": 9.492631538779082e-06,
+      "loss": 0.7152,
+      "step": 2661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6977286482609748,
+      "learning_rate": 9.492176507643433e-06,
+      "loss": 0.7822,
+      "step": 2662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7849867363702214,
+      "learning_rate": 9.491721283468954e-06,
+      "loss": 0.8186,
+      "step": 2663
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1260467179097187,
+      "learning_rate": 9.49126586627521e-06,
+      "loss": 0.6743,
+      "step": 2664
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.90753493114231,
+      "learning_rate": 9.490810256081766e-06,
+      "loss": 0.8407,
+      "step": 2665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6273531748866732,
+      "learning_rate": 9.490354452908208e-06,
+      "loss": 0.8445,
+      "step": 2666
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9186060105626206,
+      "learning_rate": 9.489898456774116e-06,
+      "loss": 0.7377,
+      "step": 2667
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9458628255794532,
+      "learning_rate": 9.489442267699087e-06,
+      "loss": 0.8273,
+      "step": 2668
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1146158829918527,
+      "learning_rate": 9.488985885702728e-06,
+      "loss": 0.8875,
+      "step": 2669
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.563122578695036,
+      "learning_rate": 9.488529310804645e-06,
+      "loss": 0.849,
+      "step": 2670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7735492283595795,
+      "learning_rate": 9.488072543024462e-06,
+      "loss": 0.7727,
+      "step": 2671
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0134195696872363,
+      "learning_rate": 9.487615582381809e-06,
+      "loss": 0.82,
+      "step": 2672
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7248303747023277,
+      "learning_rate": 9.487158428896317e-06,
+      "loss": 0.7798,
+      "step": 2673
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.779843688565786,
+      "learning_rate": 9.486701082587635e-06,
+      "loss": 0.7578,
+      "step": 2674
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7060186447592918,
+      "learning_rate": 9.486243543475414e-06,
+      "loss": 0.7123,
+      "step": 2675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.147833145197602,
+      "learning_rate": 9.485785811579318e-06,
+      "loss": 0.666,
+      "step": 2676
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9965977712486849,
+      "learning_rate": 9.485327886919013e-06,
+      "loss": 0.7542,
+      "step": 2677
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3001074968881896,
+      "learning_rate": 9.484869769514182e-06,
+      "loss": 0.6953,
+      "step": 2678
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8997273380031414,
+      "learning_rate": 9.484411459384508e-06,
+      "loss": 0.7626,
+      "step": 2679
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8690792718190017,
+      "learning_rate": 9.483952956549684e-06,
+      "loss": 0.7232,
+      "step": 2680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7017885634714083,
+      "learning_rate": 9.483494261029418e-06,
+      "loss": 0.8052,
+      "step": 2681
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.765053962832087,
+      "learning_rate": 9.483035372843417e-06,
+      "loss": 0.8595,
+      "step": 2682
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7618686651761004,
+      "learning_rate": 9.482576292011402e-06,
+      "loss": 0.8268,
+      "step": 2683
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6230483193993135,
+      "learning_rate": 9.482117018553101e-06,
+      "loss": 0.8522,
+      "step": 2684
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9633165783382176,
+      "learning_rate": 9.481657552488248e-06,
+      "loss": 0.7539,
+      "step": 2685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.708236125810896,
+      "learning_rate": 9.481197893836591e-06,
+      "loss": 0.8093,
+      "step": 2686
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.994931782954903,
+      "learning_rate": 9.480738042617878e-06,
+      "loss": 0.7887,
+      "step": 2687
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2550255019277663,
+      "learning_rate": 9.480277998851875e-06,
+      "loss": 0.6785,
+      "step": 2688
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.12909507071156,
+      "learning_rate": 9.479817762558345e-06,
+      "loss": 0.7744,
+      "step": 2689
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4892598711692941,
+      "learning_rate": 9.479357333757072e-06,
+      "loss": 0.7769,
+      "step": 2690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9499498289763462,
+      "learning_rate": 9.478896712467836e-06,
+      "loss": 0.6952,
+      "step": 2691
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.889169586336507,
+      "learning_rate": 9.478435898710434e-06,
+      "loss": 0.6887,
+      "step": 2692
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.284756477392592,
+      "learning_rate": 9.477974892504666e-06,
+      "loss": 0.8829,
+      "step": 2693
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.910461549120787,
+      "learning_rate": 9.477513693870347e-06,
+      "loss": 0.9231,
+      "step": 2694
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6784229729948683,
+      "learning_rate": 9.47705230282729e-06,
+      "loss": 0.7188,
+      "step": 2695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7636560382536977,
+      "learning_rate": 9.476590719395324e-06,
+      "loss": 0.7197,
+      "step": 2696
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6012341360411781,
+      "learning_rate": 9.476128943594287e-06,
+      "loss": 0.796,
+      "step": 2697
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9743383700717432,
+      "learning_rate": 9.47566697544402e-06,
+      "loss": 0.841,
+      "step": 2698
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.752155621032904,
+      "learning_rate": 9.475204814964374e-06,
+      "loss": 0.777,
+      "step": 2699
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7859747611463628,
+      "learning_rate": 9.474742462175212e-06,
+      "loss": 0.8454,
+      "step": 2700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6579002252379205,
+      "learning_rate": 9.4742799170964e-06,
+      "loss": 0.823,
+      "step": 2701
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9533388824566915,
+      "learning_rate": 9.473817179747815e-06,
+      "loss": 0.7602,
+      "step": 2702
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.866751178802544,
+      "learning_rate": 9.473354250149342e-06,
+      "loss": 0.8589,
+      "step": 2703
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.804410633599083,
+      "learning_rate": 9.472891128320874e-06,
+      "loss": 0.7868,
+      "step": 2704
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5894227286075142,
+      "learning_rate": 9.472427814282311e-06,
+      "loss": 0.878,
+      "step": 2705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1471957094647305,
+      "learning_rate": 9.471964308053567e-06,
+      "loss": 0.7234,
+      "step": 2706
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5629862635753673,
+      "learning_rate": 9.471500609654556e-06,
+      "loss": 0.7642,
+      "step": 2707
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9962985826047214,
+      "learning_rate": 9.471036719105205e-06,
+      "loss": 0.6648,
+      "step": 2708
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2508541789819763,
+      "learning_rate": 9.470572636425451e-06,
+      "loss": 0.6927,
+      "step": 2709
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7110201516183594,
+      "learning_rate": 9.470108361635231e-06,
+      "loss": 0.7791,
+      "step": 2710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7666530826952231,
+      "learning_rate": 9.469643894754502e-06,
+      "loss": 0.7267,
+      "step": 2711
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7945678950965898,
+      "learning_rate": 9.469179235803217e-06,
+      "loss": 0.6672,
+      "step": 2712
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9717249934763674,
+      "learning_rate": 9.46871438480135e-06,
+      "loss": 0.7479,
+      "step": 2713
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1376719598477278,
+      "learning_rate": 9.46824934176887e-06,
+      "loss": 0.6442,
+      "step": 2714
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.904951849423754,
+      "learning_rate": 9.467784106725767e-06,
+      "loss": 0.7897,
+      "step": 2715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9273924339389825,
+      "learning_rate": 9.467318679692031e-06,
+      "loss": 0.794,
+      "step": 2716
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.137983544506982,
+      "learning_rate": 9.466853060687662e-06,
+      "loss": 0.6706,
+      "step": 2717
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.850380727217803,
+      "learning_rate": 9.466387249732668e-06,
+      "loss": 0.8154,
+      "step": 2718
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8643688002954242,
+      "learning_rate": 9.465921246847067e-06,
+      "loss": 0.7843,
+      "step": 2719
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2113156648204038,
+      "learning_rate": 9.465455052050884e-06,
+      "loss": 0.6721,
+      "step": 2720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9077374781375795,
+      "learning_rate": 9.464988665364152e-06,
+      "loss": 0.8185,
+      "step": 2721
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.614646864675138,
+      "learning_rate": 9.464522086806914e-06,
+      "loss": 0.8978,
+      "step": 2722
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8683212990632247,
+      "learning_rate": 9.464055316399217e-06,
+      "loss": 0.8165,
+      "step": 2723
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9485808064329635,
+      "learning_rate": 9.463588354161122e-06,
+      "loss": 0.9163,
+      "step": 2724
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8088751027464187,
+      "learning_rate": 9.463121200112694e-06,
+      "loss": 0.7259,
+      "step": 2725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7058429023641442,
+      "learning_rate": 9.462653854274008e-06,
+      "loss": 0.8429,
+      "step": 2726
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6381108522464671,
+      "learning_rate": 9.462186316665148e-06,
+      "loss": 0.7014,
+      "step": 2727
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.678578582959986,
+      "learning_rate": 9.461718587306205e-06,
+      "loss": 0.8193,
+      "step": 2728
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9265598389165752,
+      "learning_rate": 9.461250666217277e-06,
+      "loss": 0.8017,
+      "step": 2729
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.565195531791346,
+      "learning_rate": 9.46078255341847e-06,
+      "loss": 0.8437,
+      "step": 2730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1723222073747221,
+      "learning_rate": 9.460314248929903e-06,
+      "loss": 0.6783,
+      "step": 2731
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.074581306090162,
+      "learning_rate": 9.4598457527717e-06,
+      "loss": 0.6315,
+      "step": 2732
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7376415791491504,
+      "learning_rate": 9.459377064963993e-06,
+      "loss": 0.8501,
+      "step": 2733
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8952427297770613,
+      "learning_rate": 9.458908185526921e-06,
+      "loss": 0.87,
+      "step": 2734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1124911719689443,
+      "learning_rate": 9.458439114480634e-06,
+      "loss": 0.7847,
+      "step": 2735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7092861703583857,
+      "learning_rate": 9.457969851845287e-06,
+      "loss": 0.7329,
+      "step": 2736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.935517304377744,
+      "learning_rate": 9.457500397641049e-06,
+      "loss": 0.7297,
+      "step": 2737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.583302668179417,
+      "learning_rate": 9.45703075188809e-06,
+      "loss": 0.6518,
+      "step": 2738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6987350640546826,
+      "learning_rate": 9.456560914606594e-06,
+      "loss": 0.7011,
+      "step": 2739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6807241667463542,
+      "learning_rate": 9.456090885816751e-06,
+      "loss": 0.799,
+      "step": 2740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.222730740082295,
+      "learning_rate": 9.455620665538758e-06,
+      "loss": 0.7125,
+      "step": 2741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6172580045748841,
+      "learning_rate": 9.455150253792822e-06,
+      "loss": 0.8292,
+      "step": 2742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7713530159604876,
+      "learning_rate": 9.454679650599156e-06,
+      "loss": 0.8324,
+      "step": 2743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.743930172590722,
+      "learning_rate": 9.454208855977986e-06,
+      "loss": 0.8241,
+      "step": 2744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7738558994445386,
+      "learning_rate": 9.45373786994954e-06,
+      "loss": 0.7977,
+      "step": 2745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8084478897047234,
+      "learning_rate": 9.45326669253406e-06,
+      "loss": 0.7058,
+      "step": 2746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.072036897925759,
+      "learning_rate": 9.452795323751793e-06,
+      "loss": 0.7725,
+      "step": 2747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6734680332636391,
+      "learning_rate": 9.452323763622993e-06,
+      "loss": 0.8728,
+      "step": 2748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2863072389360592,
+      "learning_rate": 9.451852012167924e-06,
+      "loss": 0.5707,
+      "step": 2749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7514184472358096,
+      "learning_rate": 9.45138006940686e-06,
+      "loss": 0.7541,
+      "step": 2750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6676106605061065,
+      "learning_rate": 9.450907935360081e-06,
+      "loss": 0.7546,
+      "step": 2751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7818893476233149,
+      "learning_rate": 9.450435610047875e-06,
+      "loss": 0.8059,
+      "step": 2752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.654356403214228,
+      "learning_rate": 9.44996309349054e-06,
+      "loss": 0.8335,
+      "step": 2753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9004755639277342,
+      "learning_rate": 9.449490385708378e-06,
+      "loss": 0.8835,
+      "step": 2754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.883322647916259,
+      "learning_rate": 9.449017486721705e-06,
+      "loss": 0.7149,
+      "step": 2755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7309640189131255,
+      "learning_rate": 9.448544396550846e-06,
+      "loss": 0.7008,
+      "step": 2756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0225417978616465,
+      "learning_rate": 9.448071115216123e-06,
+      "loss": 0.7286,
+      "step": 2757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1474213361156678,
+      "learning_rate": 9.447597642737878e-06,
+      "loss": 0.6624,
+      "step": 2758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9576257064367475,
+      "learning_rate": 9.447123979136457e-06,
+      "loss": 0.8111,
+      "step": 2759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.802647593179244,
+      "learning_rate": 9.446650124432215e-06,
+      "loss": 0.81,
+      "step": 2760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8474009776880331,
+      "learning_rate": 9.446176078645515e-06,
+      "loss": 0.7337,
+      "step": 2761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4926023159596953,
+      "learning_rate": 9.445701841796726e-06,
+      "loss": 0.7755,
+      "step": 2762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7457090548694933,
+      "learning_rate": 9.445227413906227e-06,
+      "loss": 0.7574,
+      "step": 2763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.083977093305125,
+      "learning_rate": 9.444752794994408e-06,
+      "loss": 0.6654,
+      "step": 2764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6981350316557182,
+      "learning_rate": 9.44427798508166e-06,
+      "loss": 0.7994,
+      "step": 2765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.88606020467518,
+      "learning_rate": 9.44380298418839e-06,
+      "loss": 0.8193,
+      "step": 2766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9017010721712346,
+      "learning_rate": 9.443327792335009e-06,
+      "loss": 0.8523,
+      "step": 2767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2362127377866239,
+      "learning_rate": 9.442852409541935e-06,
+      "loss": 0.6623,
+      "step": 2768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6172935482998059,
+      "learning_rate": 9.4423768358296e-06,
+      "loss": 0.8056,
+      "step": 2769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7667162326761312,
+      "learning_rate": 9.44190107121844e-06,
+      "loss": 0.7011,
+      "step": 2770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.696405123581682,
+      "learning_rate": 9.441425115728898e-06,
+      "loss": 0.9128,
+      "step": 2771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.756144472361467,
+      "learning_rate": 9.440948969381425e-06,
+      "loss": 0.7969,
+      "step": 2772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0816870616582133,
+      "learning_rate": 9.440472632196485e-06,
+      "loss": 0.7592,
+      "step": 2773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.0512118303988895,
+      "learning_rate": 9.439996104194546e-06,
+      "loss": 0.7947,
+      "step": 2774
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7562954781468973,
+      "learning_rate": 9.439519385396088e-06,
+      "loss": 0.7967,
+      "step": 2775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6755204307504723,
+      "learning_rate": 9.439042475821594e-06,
+      "loss": 0.7788,
+      "step": 2776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6167530037603195,
+      "learning_rate": 9.438565375491558e-06,
+      "loss": 0.7948,
+      "step": 2777
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7827103754243652,
+      "learning_rate": 9.438088084426482e-06,
+      "loss": 0.7146,
+      "step": 2778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8052175827547305,
+      "learning_rate": 9.437610602646878e-06,
+      "loss": 0.8063,
+      "step": 2779
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5627717057878672,
+      "learning_rate": 9.437132930173264e-06,
+      "loss": 0.8133,
+      "step": 2780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8722218972171425,
+      "learning_rate": 9.436655067026164e-06,
+      "loss": 0.6812,
+      "step": 2781
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.757128817401547,
+      "learning_rate": 9.436177013226116e-06,
+      "loss": 0.6988,
+      "step": 2782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8266330585145045,
+      "learning_rate": 9.43569876879366e-06,
+      "loss": 0.763,
+      "step": 2783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8808517454314724,
+      "learning_rate": 9.43522033374935e-06,
+      "loss": 0.8365,
+      "step": 2784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5470257715231586,
+      "learning_rate": 9.434741708113745e-06,
+      "loss": 0.7769,
+      "step": 2785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7549985094910288,
+      "learning_rate": 9.434262891907413e-06,
+      "loss": 0.8126,
+      "step": 2786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7454875568959844,
+      "learning_rate": 9.433783885150929e-06,
+      "loss": 0.8325,
+      "step": 2787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1608884361155813,
+      "learning_rate": 9.433304687864876e-06,
+      "loss": 0.75,
+      "step": 2788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.597426359576469,
+      "learning_rate": 9.432825300069848e-06,
+      "loss": 0.7417,
+      "step": 2789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3504622347444943,
+      "learning_rate": 9.432345721786442e-06,
+      "loss": 0.7367,
+      "step": 2790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.859505907179868,
+      "learning_rate": 9.431865953035271e-06,
+      "loss": 0.7705,
+      "step": 2791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6070576428652767,
+      "learning_rate": 9.431385993836949e-06,
+      "loss": 0.7192,
+      "step": 2792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.623837065250548,
+      "learning_rate": 9.430905844212102e-06,
+      "loss": 0.6976,
+      "step": 2793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9396356541163053,
+      "learning_rate": 9.430425504181361e-06,
+      "loss": 0.8238,
+      "step": 2794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9300953653563588,
+      "learning_rate": 9.429944973765373e-06,
+      "loss": 0.9462,
+      "step": 2795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8459992722839837,
+      "learning_rate": 9.429464252984779e-06,
+      "loss": 0.7385,
+      "step": 2796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.739151315216294,
+      "learning_rate": 9.428983341860244e-06,
+      "loss": 0.7846,
+      "step": 2797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4229200746772004,
+      "learning_rate": 9.42850224041243e-06,
+      "loss": 0.7393,
+      "step": 2798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.119335485559721,
+      "learning_rate": 9.428020948662012e-06,
+      "loss": 0.6741,
+      "step": 2799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6457014663318903,
+      "learning_rate": 9.427539466629672e-06,
+      "loss": 0.9435,
+      "step": 2800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9515724599120068,
+      "learning_rate": 9.427057794336101e-06,
+      "loss": 0.8513,
+      "step": 2801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8600245473904287,
+      "learning_rate": 9.426575931801996e-06,
+      "loss": 0.7424,
+      "step": 2802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8404386087045297,
+      "learning_rate": 9.426093879048065e-06,
+      "loss": 0.8567,
+      "step": 2803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.14115953034429,
+      "learning_rate": 9.425611636095023e-06,
+      "loss": 0.9691,
+      "step": 2804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5602174592644709,
+      "learning_rate": 9.425129202963592e-06,
+      "loss": 0.6658,
+      "step": 2805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7028925661747054,
+      "learning_rate": 9.424646579674504e-06,
+      "loss": 0.7882,
+      "step": 2806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7190162064086252,
+      "learning_rate": 9.424163766248499e-06,
+      "loss": 0.7754,
+      "step": 2807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.491323134253921,
+      "learning_rate": 9.423680762706323e-06,
+      "loss": 0.6992,
+      "step": 2808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.522201464489184,
+      "learning_rate": 9.423197569068733e-06,
+      "loss": 0.8785,
+      "step": 2809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6913603401089368,
+      "learning_rate": 9.422714185356491e-06,
+      "loss": 0.8279,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.548952052359859,
+      "learning_rate": 9.422230611590373e-06,
+      "loss": 0.7622,
+      "step": 2811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6634119472298632,
+      "learning_rate": 9.421746847791156e-06,
+      "loss": 0.8348,
+      "step": 2812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.953379804199383,
+      "learning_rate": 9.421262893979628e-06,
+      "loss": 0.8376,
+      "step": 2813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8660848533981984,
+      "learning_rate": 9.420778750176588e-06,
+      "loss": 0.8079,
+      "step": 2814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3543931241431446,
+      "learning_rate": 9.420294416402839e-06,
+      "loss": 0.7536,
+      "step": 2815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0393000571328397,
+      "learning_rate": 9.419809892679194e-06,
+      "loss": 0.755,
+      "step": 2816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0143763023650774,
+      "learning_rate": 9.419325179026475e-06,
+      "loss": 0.8596,
+      "step": 2817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7038870293077353,
+      "learning_rate": 9.41884027546551e-06,
+      "loss": 0.7638,
+      "step": 2818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6645582443713516,
+      "learning_rate": 9.418355182017138e-06,
+      "loss": 0.8367,
+      "step": 2819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.655031206552679,
+      "learning_rate": 9.417869898702205e-06,
+      "loss": 0.6976,
+      "step": 2820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7873976060939676,
+      "learning_rate": 9.41738442554156e-06,
+      "loss": 0.767,
+      "step": 2821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7282883830807936,
+      "learning_rate": 9.41689876255607e-06,
+      "loss": 0.7133,
+      "step": 2822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4477124459847466,
+      "learning_rate": 9.416412909766604e-06,
+      "loss": 0.7191,
+      "step": 2823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.297436143857127,
+      "learning_rate": 9.41592686719404e-06,
+      "loss": 0.7605,
+      "step": 2824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5668233714917785,
+      "learning_rate": 9.41544063485926e-06,
+      "loss": 0.8681,
+      "step": 2825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9701594105981421,
+      "learning_rate": 9.414954212783165e-06,
+      "loss": 0.5491,
+      "step": 2826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7449745535507735,
+      "learning_rate": 9.414467600986655e-06,
+      "loss": 0.9542,
+      "step": 2827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5676773093173118,
+      "learning_rate": 9.41398079949064e-06,
+      "loss": 0.7783,
+      "step": 2828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5737705416021537,
+      "learning_rate": 9.413493808316038e-06,
+      "loss": 0.791,
+      "step": 2829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8268925849505593,
+      "learning_rate": 9.41300662748378e-06,
+      "loss": 0.7718,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8223989528117852,
+      "learning_rate": 9.412519257014798e-06,
+      "loss": 0.8127,
+      "step": 2831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.110573517144879,
+      "learning_rate": 9.412031696930036e-06,
+      "loss": 0.6534,
+      "step": 2832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2302274224089096,
+      "learning_rate": 9.411543947250447e-06,
+      "loss": 0.6746,
+      "step": 2833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9881731673918406,
+      "learning_rate": 9.411056007996989e-06,
+      "loss": 0.8885,
+      "step": 2834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5694878164561579,
+      "learning_rate": 9.41056787919063e-06,
+      "loss": 0.7271,
+      "step": 2835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4912007685433157,
+      "learning_rate": 9.410079560852347e-06,
+      "loss": 0.7029,
+      "step": 2836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8424549344916257,
+      "learning_rate": 9.409591053003121e-06,
+      "loss": 0.7632,
+      "step": 2837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8179913705833683,
+      "learning_rate": 9.40910235566395e-06,
+      "loss": 0.7481,
+      "step": 2838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0891646412365805,
+      "learning_rate": 9.408613468855829e-06,
+      "loss": 0.9596,
+      "step": 2839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7295342804063558,
+      "learning_rate": 9.40812439259977e-06,
+      "loss": 0.8094,
+      "step": 2840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7043573212006402,
+      "learning_rate": 9.40763512691679e-06,
+      "loss": 0.7581,
+      "step": 2841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0753007257977223,
+      "learning_rate": 9.407145671827909e-06,
+      "loss": 0.6746,
+      "step": 2842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4029128595555738,
+      "learning_rate": 9.406656027354166e-06,
+      "loss": 0.6796,
+      "step": 2843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3751695304259324,
+      "learning_rate": 9.406166193516596e-06,
+      "loss": 0.748,
+      "step": 2844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8348236224584964,
+      "learning_rate": 9.405676170336255e-06,
+      "loss": 0.7827,
+      "step": 2845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9194721363896008,
+      "learning_rate": 9.405185957834196e-06,
+      "loss": 0.787,
+      "step": 2846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9702426175222312,
+      "learning_rate": 9.404695556031485e-06,
+      "loss": 0.7229,
+      "step": 2847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8780495504667156,
+      "learning_rate": 9.404204964949196e-06,
+      "loss": 0.6865,
+      "step": 2848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.48465723675729,
+      "learning_rate": 9.403714184608411e-06,
+      "loss": 0.8861,
+      "step": 2849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9187298099380357,
+      "learning_rate": 9.403223215030221e-06,
+      "loss": 0.7912,
+      "step": 2850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0898988809134087,
+      "learning_rate": 9.402732056235722e-06,
+      "loss": 0.81,
+      "step": 2851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6745278193818869,
+      "learning_rate": 9.402240708246022e-06,
+      "loss": 0.8104,
+      "step": 2852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8242451596840719,
+      "learning_rate": 9.401749171082233e-06,
+      "loss": 0.8213,
+      "step": 2853
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.670645059282653,
+      "learning_rate": 9.40125744476548e-06,
+      "loss": 0.7927,
+      "step": 2854
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1845057155843133,
+      "learning_rate": 9.400765529316892e-06,
+      "loss": 0.6431,
+      "step": 2855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.096208309904523,
+      "learning_rate": 9.400273424757607e-06,
+      "loss": 0.9128,
+      "step": 2856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9597896570460374,
+      "learning_rate": 9.399781131108775e-06,
+      "loss": 0.7774,
+      "step": 2857
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.978221106173137,
+      "learning_rate": 9.399288648391548e-06,
+      "loss": 0.7972,
+      "step": 2858
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6447753033846841,
+      "learning_rate": 9.398795976627091e-06,
+      "loss": 0.6681,
+      "step": 2859
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.710139841585123,
+      "learning_rate": 9.398303115836573e-06,
+      "loss": 0.7513,
+      "step": 2860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.908697070898909,
+      "learning_rate": 9.397810066041176e-06,
+      "loss": 0.8723,
+      "step": 2861
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7799891533987726,
+      "learning_rate": 9.397316827262085e-06,
+      "loss": 0.8958,
+      "step": 2862
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8538072780311892,
+      "learning_rate": 9.396823399520495e-06,
+      "loss": 0.7461,
+      "step": 2863
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9822658460834623,
+      "learning_rate": 9.396329782837614e-06,
+      "loss": 0.7706,
+      "step": 2864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.782076773009734,
+      "learning_rate": 9.39583597723465e-06,
+      "loss": 0.8659,
+      "step": 2865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.538251477208933,
+      "learning_rate": 9.395341982732825e-06,
+      "loss": 0.7991,
+      "step": 2866
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6245755108339293,
+      "learning_rate": 9.394847799353363e-06,
+      "loss": 0.7898,
+      "step": 2867
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5834862682297832,
+      "learning_rate": 9.394353427117506e-06,
+      "loss": 0.6289,
+      "step": 2868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5421784605996276,
+      "learning_rate": 9.393858866046494e-06,
+      "loss": 0.7822,
+      "step": 2869
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.74393128164791,
+      "learning_rate": 9.393364116161582e-06,
+      "loss": 0.6541,
+      "step": 2870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1285563584436884,
+      "learning_rate": 9.392869177484029e-06,
+      "loss": 0.6842,
+      "step": 2871
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2530870111547163,
+      "learning_rate": 9.392374050035104e-06,
+      "loss": 0.7179,
+      "step": 2872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1836700547138506,
+      "learning_rate": 9.391878733836085e-06,
+      "loss": 0.8027,
+      "step": 2873
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.78363911311905,
+      "learning_rate": 9.391383228908253e-06,
+      "loss": 0.8897,
+      "step": 2874
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.457004737958957,
+      "learning_rate": 9.390887535272906e-06,
+      "loss": 0.8656,
+      "step": 2875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1860503015341464,
+      "learning_rate": 9.39039165295134e-06,
+      "loss": 0.8138,
+      "step": 2876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5776918129228652,
+      "learning_rate": 9.38989558196487e-06,
+      "loss": 0.7063,
+      "step": 2877
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.694274666395654,
+      "learning_rate": 9.389399322334806e-06,
+      "loss": 0.7874,
+      "step": 2878
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7003197517234863,
+      "learning_rate": 9.388902874082482e-06,
+      "loss": 0.8358,
+      "step": 2879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0333836986240033,
+      "learning_rate": 9.388406237229223e-06,
+      "loss": 0.7146,
+      "step": 2880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9675163116820382,
+      "learning_rate": 9.387909411796377e-06,
+      "loss": 0.8445,
+      "step": 2881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5237165257666558,
+      "learning_rate": 9.387412397805289e-06,
+      "loss": 0.7394,
+      "step": 2882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.555429325214831,
+      "learning_rate": 9.386915195277321e-06,
+      "loss": 0.7795,
+      "step": 2883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8444577533394653,
+      "learning_rate": 9.386417804233836e-06,
+      "loss": 0.7767,
+      "step": 2884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9593727077586796,
+      "learning_rate": 9.385920224696208e-06,
+      "loss": 0.799,
+      "step": 2885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1127742610096356,
+      "learning_rate": 9.385422456685822e-06,
+      "loss": 0.9116,
+      "step": 2886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7363698421162999,
+      "learning_rate": 9.384924500224065e-06,
+      "loss": 0.7635,
+      "step": 2887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.815636786246473,
+      "learning_rate": 9.384426355332336e-06,
+      "loss": 0.8055,
+      "step": 2888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.77172783130554,
+      "learning_rate": 9.383928022032044e-06,
+      "loss": 0.8266,
+      "step": 2889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7649059443469526,
+      "learning_rate": 9.383429500344598e-06,
+      "loss": 0.7865,
+      "step": 2890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6358678082931724,
+      "learning_rate": 9.382930790291426e-06,
+      "loss": 0.781,
+      "step": 2891
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7266319023248466,
+      "learning_rate": 9.382431891893957e-06,
+      "loss": 0.8564,
+      "step": 2892
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9450509846677029,
+      "learning_rate": 9.381932805173628e-06,
+      "loss": 0.8912,
+      "step": 2893
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8743004032178496,
+      "learning_rate": 9.381433530151887e-06,
+      "loss": 0.8391,
+      "step": 2894
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.580771293045074,
+      "learning_rate": 9.380934066850191e-06,
+      "loss": 0.9764,
+      "step": 2895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4363974555627352,
+      "learning_rate": 9.38043441529e-06,
+      "loss": 0.6768,
+      "step": 2896
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5831143436033102,
+      "learning_rate": 9.379934575492788e-06,
+      "loss": 0.7697,
+      "step": 2897
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.307861011980168,
+      "learning_rate": 9.37943454748003e-06,
+      "loss": 0.8547,
+      "step": 2898
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6386110943906902,
+      "learning_rate": 9.37893433127322e-06,
+      "loss": 0.8231,
+      "step": 2899
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.765931029689286,
+      "learning_rate": 9.378433926893845e-06,
+      "loss": 0.8369,
+      "step": 2900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7164955721673147,
+      "learning_rate": 9.377933334363416e-06,
+      "loss": 0.7755,
+      "step": 2901
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0387034073718335,
+      "learning_rate": 9.377432553703439e-06,
+      "loss": 0.8432,
+      "step": 2902
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.526493423989316,
+      "learning_rate": 9.376931584935439e-06,
+      "loss": 0.7994,
+      "step": 2903
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9538694552770433,
+      "learning_rate": 9.376430428080939e-06,
+      "loss": 0.8349,
+      "step": 2904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7680847811258018,
+      "learning_rate": 9.375929083161475e-06,
+      "loss": 0.6578,
+      "step": 2905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.845876114887219,
+      "learning_rate": 9.375427550198595e-06,
+      "loss": 0.9615,
+      "step": 2906
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0720469559879142,
+      "learning_rate": 9.374925829213848e-06,
+      "loss": 0.5422,
+      "step": 2907
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5578492939759758,
+      "learning_rate": 9.374423920228795e-06,
+      "loss": 0.819,
+      "step": 2908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4127038153494325,
+      "learning_rate": 9.373921823265004e-06,
+      "loss": 0.7236,
+      "step": 2909
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5977795750933939,
+      "learning_rate": 9.373419538344049e-06,
+      "loss": 0.7893,
+      "step": 2910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9450744095473003,
+      "learning_rate": 9.372917065487518e-06,
+      "loss": 0.9317,
+      "step": 2911
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.641997326941341,
+      "learning_rate": 9.372414404717001e-06,
+      "loss": 0.7183,
+      "step": 2912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.473261294674531,
+      "learning_rate": 9.3719115560541e-06,
+      "loss": 0.7811,
+      "step": 2913
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4406634515489016,
+      "learning_rate": 9.371408519520421e-06,
+      "loss": 0.7468,
+      "step": 2914
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7631299624084138,
+      "learning_rate": 9.370905295137584e-06,
+      "loss": 0.839,
+      "step": 2915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.705668298772368,
+      "learning_rate": 9.370401882927212e-06,
+      "loss": 0.7872,
+      "step": 2916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.860813244041042,
+      "learning_rate": 9.369898282910937e-06,
+      "loss": 0.8425,
+      "step": 2917
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.005104042978183,
+      "learning_rate": 9.369394495110401e-06,
+      "loss": 0.7272,
+      "step": 2918
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5850477247455779,
+      "learning_rate": 9.36889051954725e-06,
+      "loss": 0.7402,
+      "step": 2919
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6381965806071745,
+      "learning_rate": 9.368386356243148e-06,
+      "loss": 0.9826,
+      "step": 2920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4129950344296467,
+      "learning_rate": 9.367882005219752e-06,
+      "loss": 0.9113,
+      "step": 2921
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8358274778200845,
+      "learning_rate": 9.367377466498739e-06,
+      "loss": 0.8388,
+      "step": 2922
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5833620143356257,
+      "learning_rate": 9.36687274010179e-06,
+      "loss": 0.8218,
+      "step": 2923
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7737484937680836,
+      "learning_rate": 9.366367826050593e-06,
+      "loss": 0.8063,
+      "step": 2924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.765922792780213,
+      "learning_rate": 9.365862724366848e-06,
+      "loss": 0.7717,
+      "step": 2925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8429557057652903,
+      "learning_rate": 9.365357435072255e-06,
+      "loss": 0.7888,
+      "step": 2926
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.114440875006172,
+      "learning_rate": 9.364851958188534e-06,
+      "loss": 0.8415,
+      "step": 2927
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7649911673132077,
+      "learning_rate": 9.364346293737403e-06,
+      "loss": 0.8026,
+      "step": 2928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7264108177075805,
+      "learning_rate": 9.36384044174059e-06,
+      "loss": 0.746,
+      "step": 2929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.91117287944512,
+      "learning_rate": 9.363334402219836e-06,
+      "loss": 0.8899,
+      "step": 2930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7984389034776116,
+      "learning_rate": 9.362828175196884e-06,
+      "loss": 0.7437,
+      "step": 2931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.910891223820443,
+      "learning_rate": 9.362321760693489e-06,
+      "loss": 0.7049,
+      "step": 2932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2140492435732604,
+      "learning_rate": 9.361815158731413e-06,
+      "loss": 0.8049,
+      "step": 2933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6362478448709754,
+      "learning_rate": 9.361308369332426e-06,
+      "loss": 0.8066,
+      "step": 2934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6001481621862348,
+      "learning_rate": 9.360801392518303e-06,
+      "loss": 0.7198,
+      "step": 2935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1791802250143832,
+      "learning_rate": 9.360294228310834e-06,
+      "loss": 0.7158,
+      "step": 2936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.550174626580674,
+      "learning_rate": 9.359786876731811e-06,
+      "loss": 0.7556,
+      "step": 2937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6990809919176286,
+      "learning_rate": 9.359279337803036e-06,
+      "loss": 0.8138,
+      "step": 2938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.858356611501386,
+      "learning_rate": 9.358771611546319e-06,
+      "loss": 0.7296,
+      "step": 2939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7333902147367315,
+      "learning_rate": 9.358263697983479e-06,
+      "loss": 0.8325,
+      "step": 2940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6289793910930856,
+      "learning_rate": 9.35775559713634e-06,
+      "loss": 0.754,
+      "step": 2941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.789000445844293,
+      "learning_rate": 9.357247309026738e-06,
+      "loss": 0.8266,
+      "step": 2942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1185026627135375,
+      "learning_rate": 9.356738833676517e-06,
+      "loss": 0.6349,
+      "step": 2943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2668467030307666,
+      "learning_rate": 9.356230171107524e-06,
+      "loss": 0.6703,
+      "step": 2944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.994962994063417,
+      "learning_rate": 9.355721321341618e-06,
+      "loss": 0.8509,
+      "step": 2945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6400899097349981,
+      "learning_rate": 9.355212284400668e-06,
+      "loss": 0.8288,
+      "step": 2946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8195792928709147,
+      "learning_rate": 9.354703060306546e-06,
+      "loss": 0.7415,
+      "step": 2947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6149516158337642,
+      "learning_rate": 9.354193649081134e-06,
+      "loss": 0.9412,
+      "step": 2948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6012545206322981,
+      "learning_rate": 9.353684050746323e-06,
+      "loss": 0.8182,
+      "step": 2949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6162384458838106,
+      "learning_rate": 9.353174265324014e-06,
+      "loss": 0.799,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4132464368471924,
+      "learning_rate": 9.35266429283611e-06,
+      "loss": 0.6432,
+      "step": 2951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8422680520748693,
+      "learning_rate": 9.352154133304528e-06,
+      "loss": 0.7369,
+      "step": 2952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6486902439959974,
+      "learning_rate": 9.351643786751191e-06,
+      "loss": 0.7738,
+      "step": 2953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.243042606813264,
+      "learning_rate": 9.351133253198027e-06,
+      "loss": 0.8426,
+      "step": 2954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.51745779706298,
+      "learning_rate": 9.350622532666979e-06,
+      "loss": 0.7686,
+      "step": 2955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1755362729170176,
+      "learning_rate": 9.350111625179991e-06,
+      "loss": 0.9402,
+      "step": 2956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1576450698396,
+      "learning_rate": 9.349600530759018e-06,
+      "loss": 0.6309,
+      "step": 2957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5056757386406785,
+      "learning_rate": 9.349089249426022e-06,
+      "loss": 0.7974,
+      "step": 2958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2401182824632677,
+      "learning_rate": 9.348577781202976e-06,
+      "loss": 0.8387,
+      "step": 2959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4286215474879314,
+      "learning_rate": 9.348066126111857e-06,
+      "loss": 0.6174,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6979423158194005,
+      "learning_rate": 9.347554284174654e-06,
+      "loss": 0.6663,
+      "step": 2961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.067882874879238,
+      "learning_rate": 9.347042255413362e-06,
+      "loss": 0.7622,
+      "step": 2962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6181971014789367,
+      "learning_rate": 9.34653003984998e-06,
+      "loss": 0.79,
+      "step": 2963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6225768321975895,
+      "learning_rate": 9.346017637506523e-06,
+      "loss": 0.7957,
+      "step": 2964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.907517451154521,
+      "learning_rate": 9.34550504840501e-06,
+      "loss": 0.932,
+      "step": 2965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9953073809862527,
+      "learning_rate": 9.344992272567467e-06,
+      "loss": 0.795,
+      "step": 2966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8066289502513413,
+      "learning_rate": 9.344479310015928e-06,
+      "loss": 0.8294,
+      "step": 2967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7311433077740694,
+      "learning_rate": 9.343966160772438e-06,
+      "loss": 0.8223,
+      "step": 2968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8882608216008971,
+      "learning_rate": 9.343452824859048e-06,
+      "loss": 0.788,
+      "step": 2969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.439574592430231,
+      "learning_rate": 9.342939302297816e-06,
+      "loss": 0.7284,
+      "step": 2970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8540073305183935,
+      "learning_rate": 9.34242559311081e-06,
+      "loss": 0.7416,
+      "step": 2971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6614781287121902,
+      "learning_rate": 9.341911697320105e-06,
+      "loss": 0.721,
+      "step": 2972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6024367938810986,
+      "learning_rate": 9.341397614947785e-06,
+      "loss": 0.8559,
+      "step": 2973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7676424150329535,
+      "learning_rate": 9.340883346015941e-06,
+      "loss": 0.7502,
+      "step": 2974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1747494381908299,
+      "learning_rate": 9.340368890546672e-06,
+      "loss": 0.7996,
+      "step": 2975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6466784348897374,
+      "learning_rate": 9.339854248562084e-06,
+      "loss": 0.7984,
+      "step": 2976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7732606902498482,
+      "learning_rate": 9.339339420084295e-06,
+      "loss": 0.8297,
+      "step": 2977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.292057072620092,
+      "learning_rate": 9.338824405135425e-06,
+      "loss": 0.6276,
+      "step": 2978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.739119700720906,
+      "learning_rate": 9.338309203737609e-06,
+      "loss": 0.7488,
+      "step": 2979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7404960533527365,
+      "learning_rate": 9.337793815912984e-06,
+      "loss": 0.7706,
+      "step": 2980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8697474018352798,
+      "learning_rate": 9.337278241683698e-06,
+      "loss": 0.745,
+      "step": 2981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9461107002707922,
+      "learning_rate": 9.336762481071906e-06,
+      "loss": 0.8139,
+      "step": 2982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7676947917793566,
+      "learning_rate": 9.336246534099772e-06,
+      "loss": 0.8588,
+      "step": 2983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8063002688079097,
+      "learning_rate": 9.335730400789466e-06,
+      "loss": 0.8012,
+      "step": 2984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.968569536197607,
+      "learning_rate": 9.335214081163171e-06,
+      "loss": 0.9154,
+      "step": 2985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8090887696048672,
+      "learning_rate": 9.33469757524307e-06,
+      "loss": 0.7795,
+      "step": 2986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.802578736576689,
+      "learning_rate": 9.33418088305136e-06,
+      "loss": 0.9023,
+      "step": 2987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8918849609985895,
+      "learning_rate": 9.333664004610246e-06,
+      "loss": 0.78,
+      "step": 2988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.761031272222049,
+      "learning_rate": 9.333146939941938e-06,
+      "loss": 0.7405,
+      "step": 2989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.671980131804438,
+      "learning_rate": 9.332629689068655e-06,
+      "loss": 0.7581,
+      "step": 2990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6854721065656473,
+      "learning_rate": 9.332112252012625e-06,
+      "loss": 0.8412,
+      "step": 2991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7372885799399382,
+      "learning_rate": 9.331594628796083e-06,
+      "loss": 0.8237,
+      "step": 2992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5165699573367495,
+      "learning_rate": 9.331076819441273e-06,
+      "loss": 0.7307,
+      "step": 2993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7052558213173534,
+      "learning_rate": 9.330558823970448e-06,
+      "loss": 0.7713,
+      "step": 2994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4088536989841836,
+      "learning_rate": 9.330040642405864e-06,
+      "loss": 0.8669,
+      "step": 2995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2134826068439586,
+      "learning_rate": 9.329522274769791e-06,
+      "loss": 0.6094,
+      "step": 2996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7054454489719846,
+      "learning_rate": 9.329003721084501e-06,
+      "loss": 0.8072,
+      "step": 2997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9492416236922965,
+      "learning_rate": 9.328484981372283e-06,
+      "loss": 0.7907,
+      "step": 2998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7462172145923516,
+      "learning_rate": 9.327966055655424e-06,
+      "loss": 0.7737,
+      "step": 2999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4439827226747468,
+      "learning_rate": 9.327446943956224e-06,
+      "loss": 0.8311,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5436395067737232,
+      "learning_rate": 9.326927646296992e-06,
+      "loss": 0.7882,
+      "step": 3001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.832300898433666,
+      "learning_rate": 9.326408162700043e-06,
+      "loss": 0.8574,
+      "step": 3002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.094015611300055,
+      "learning_rate": 9.325888493187699e-06,
+      "loss": 0.7427,
+      "step": 3003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.495871961022774,
+      "learning_rate": 9.325368637782292e-06,
+      "loss": 0.6676,
+      "step": 3004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.714528172522491,
+      "learning_rate": 9.324848596506162e-06,
+      "loss": 0.7973,
+      "step": 3005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7048794989256946,
+      "learning_rate": 9.324328369381654e-06,
+      "loss": 0.873,
+      "step": 3006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.646408289681697,
+      "learning_rate": 9.323807956431128e-06,
+      "loss": 0.8762,
+      "step": 3007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8208402815239517,
+      "learning_rate": 9.323287357676944e-06,
+      "loss": 0.8088,
+      "step": 3008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8465637783010715,
+      "learning_rate": 9.322766573141473e-06,
+      "loss": 0.7329,
+      "step": 3009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6346595074582857,
+      "learning_rate": 9.322245602847094e-06,
+      "loss": 0.8239,
+      "step": 3010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.71797580509366,
+      "learning_rate": 9.321724446816197e-06,
+      "loss": 0.8618,
+      "step": 3011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.04947376213863,
+      "learning_rate": 9.321203105071176e-06,
+      "loss": 0.8366,
+      "step": 3012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6689938661843369,
+      "learning_rate": 9.320681577634432e-06,
+      "loss": 0.7503,
+      "step": 3013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7360813331958884,
+      "learning_rate": 9.320159864528378e-06,
+      "loss": 0.7235,
+      "step": 3014
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.00228240129201,
+      "learning_rate": 9.319637965775432e-06,
+      "loss": 0.9291,
+      "step": 3015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8818819159159292,
+      "learning_rate": 9.319115881398024e-06,
+      "loss": 0.9881,
+      "step": 3016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.78237078374086,
+      "learning_rate": 9.31859361141859e-06,
+      "loss": 0.7789,
+      "step": 3017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7216609856484275,
+      "learning_rate": 9.318071155859567e-06,
+      "loss": 0.782,
+      "step": 3018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2422126741227602,
+      "learning_rate": 9.31754851474341e-06,
+      "loss": 0.7155,
+      "step": 3019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9350910768665932,
+      "learning_rate": 9.317025688092577e-06,
+      "loss": 0.7025,
+      "step": 3020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8603155022893823,
+      "learning_rate": 9.316502675929537e-06,
+      "loss": 0.8341,
+      "step": 3021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6459391556589815,
+      "learning_rate": 9.315979478276763e-06,
+      "loss": 0.8365,
+      "step": 3022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0288363808770062,
+      "learning_rate": 9.315456095156739e-06,
+      "loss": 0.6875,
+      "step": 3023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5716786236189426,
+      "learning_rate": 9.314932526591956e-06,
+      "loss": 0.8402,
+      "step": 3024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.061364157791136,
+      "learning_rate": 9.314408772604911e-06,
+      "loss": 0.8577,
+      "step": 3025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8726095509140679,
+      "learning_rate": 9.313884833218115e-06,
+      "loss": 0.8113,
+      "step": 3026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5951445160035966,
+      "learning_rate": 9.313360708454078e-06,
+      "loss": 0.8178,
+      "step": 3027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7261129271271811,
+      "learning_rate": 9.312836398335326e-06,
+      "loss": 0.8427,
+      "step": 3028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0691305817187065,
+      "learning_rate": 9.312311902884388e-06,
+      "loss": 0.8233,
+      "step": 3029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6754693570536539,
+      "learning_rate": 9.311787222123803e-06,
+      "loss": 0.8245,
+      "step": 3030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7920006606026064,
+      "learning_rate": 9.311262356076118e-06,
+      "loss": 0.739,
+      "step": 3031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0060676432454914,
+      "learning_rate": 9.31073730476389e-06,
+      "loss": 0.9251,
+      "step": 3032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.319816759112963,
+      "learning_rate": 9.310212068209678e-06,
+      "loss": 0.8725,
+      "step": 3033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8824931312283972,
+      "learning_rate": 9.309686646436053e-06,
+      "loss": 0.8343,
+      "step": 3034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7449567518892606,
+      "learning_rate": 9.309161039465597e-06,
+      "loss": 0.737,
+      "step": 3035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9142106933125436,
+      "learning_rate": 9.308635247320893e-06,
+      "loss": 0.8104,
+      "step": 3036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.955595715494467,
+      "learning_rate": 9.308109270024537e-06,
+      "loss": 0.9022,
+      "step": 3037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6906863949341429,
+      "learning_rate": 9.30758310759913e-06,
+      "loss": 0.8113,
+      "step": 3038
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7641636490560662,
+      "learning_rate": 9.307056760067284e-06,
+      "loss": 0.77,
+      "step": 3039
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2794759406368799,
+      "learning_rate": 9.306530227451616e-06,
+      "loss": 0.6219,
+      "step": 3040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6682346118128888,
+      "learning_rate": 9.306003509774755e-06,
+      "loss": 0.7441,
+      "step": 3041
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8777992931904646,
+      "learning_rate": 9.305476607059332e-06,
+      "loss": 0.7936,
+      "step": 3042
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9001582365176177,
+      "learning_rate": 9.30494951932799e-06,
+      "loss": 0.8528,
+      "step": 3043
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.188667361430987,
+      "learning_rate": 9.30442224660338e-06,
+      "loss": 0.7726,
+      "step": 3044
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6864038612893062,
+      "learning_rate": 9.303894788908158e-06,
+      "loss": 0.7288,
+      "step": 3045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8887924787151842,
+      "learning_rate": 9.303367146264995e-06,
+      "loss": 0.779,
+      "step": 3046
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7463294408922443,
+      "learning_rate": 9.302839318696558e-06,
+      "loss": 0.7084,
+      "step": 3047
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9460168477821533,
+      "learning_rate": 9.302311306225534e-06,
+      "loss": 0.8263,
+      "step": 3048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9645427505969821,
+      "learning_rate": 9.301783108874611e-06,
+      "loss": 0.7613,
+      "step": 3049
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6145087353486933,
+      "learning_rate": 9.301254726666488e-06,
+      "loss": 0.7167,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6972687546577032,
+      "learning_rate": 9.300726159623869e-06,
+      "loss": 0.7635,
+      "step": 3051
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5835891551396146,
+      "learning_rate": 9.300197407769472e-06,
+      "loss": 0.7558,
+      "step": 3052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6282075380029137,
+      "learning_rate": 9.299668471126011e-06,
+      "loss": 0.778,
+      "step": 3053
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6438150948488546,
+      "learning_rate": 9.299139349716221e-06,
+      "loss": 0.909,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.755601727839862,
+      "learning_rate": 9.29861004356284e-06,
+      "loss": 0.7811,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1185456127696285,
+      "learning_rate": 9.29808055268861e-06,
+      "loss": 0.6047,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.181003197803534,
+      "learning_rate": 9.297550877116288e-06,
+      "loss": 0.8854,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0848541986808078,
+      "learning_rate": 9.297021016868634e-06,
+      "loss": 0.6753,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.730612667466387,
+      "learning_rate": 9.296490971968416e-06,
+      "loss": 0.7131,
+      "step": 3059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7304343228868926,
+      "learning_rate": 9.295960742438413e-06,
+      "loss": 0.7791,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.910821473065423,
+      "learning_rate": 9.295430328301406e-06,
+      "loss": 0.7555,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8377759846168646,
+      "learning_rate": 9.294899729580196e-06,
+      "loss": 0.7283,
+      "step": 3062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5619999975248229,
+      "learning_rate": 9.294368946297576e-06,
+      "loss": 0.6304,
+      "step": 3063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9379646310078318,
+      "learning_rate": 9.293837978476359e-06,
+      "loss": 0.8801,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2733757826775938,
+      "learning_rate": 9.293306826139361e-06,
+      "loss": 0.7043,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.08757208844996,
+      "learning_rate": 9.292775489309409e-06,
+      "loss": 0.8892,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.704937087433748,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 0.923,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.020368618265744,
+      "learning_rate": 9.291712262261972e-06,
+      "loss": 0.8436,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7099640551251316,
+      "learning_rate": 9.291180372090178e-06,
+      "loss": 0.7247,
+      "step": 3069
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.418247460666252,
+      "learning_rate": 9.290648297516806e-06,
+      "loss": 0.8243,
+      "step": 3070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8021568672855364,
+      "learning_rate": 9.290116038564722e-06,
+      "loss": 0.887,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.855781630132528,
+      "learning_rate": 9.2895835952568e-06,
+      "loss": 0.9339,
+      "step": 3072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5241079070265748,
+      "learning_rate": 9.289050967615914e-06,
+      "loss": 0.7749,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3564358321130023,
+      "learning_rate": 9.288518155664956e-06,
+      "loss": 0.9957,
+      "step": 3074
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7336512640755692,
+      "learning_rate": 9.28798515942682e-06,
+      "loss": 0.6793,
+      "step": 3075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8949179342517264,
+      "learning_rate": 9.287451978924415e-06,
+      "loss": 0.874,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6090304220194593,
+      "learning_rate": 9.28691861418065e-06,
+      "loss": 0.7424,
+      "step": 3077
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.110972887667611,
+      "learning_rate": 9.286385065218442e-06,
+      "loss": 0.6807,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8881916008029744,
+      "learning_rate": 9.285851332060722e-06,
+      "loss": 0.8964,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6170468537469547,
+      "learning_rate": 9.285317414730427e-06,
+      "loss": 0.7967,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7049927224298551,
+      "learning_rate": 9.284783313250499e-06,
+      "loss": 0.7914,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9987755603716417,
+      "learning_rate": 9.284249027643887e-06,
+      "loss": 0.8741,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.636234899162729,
+      "learning_rate": 9.283714557933554e-06,
+      "loss": 0.7435,
+      "step": 3083
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1723842479930942,
+      "learning_rate": 9.283179904142465e-06,
+      "loss": 0.6459,
+      "step": 3084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7013597147449604,
+      "learning_rate": 9.282645066293597e-06,
+      "loss": 0.7663,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9539488322071359,
+      "learning_rate": 9.282110044409933e-06,
+      "loss": 0.8362,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.437448405480939,
+      "learning_rate": 9.281574838514464e-06,
+      "loss": 0.8353,
+      "step": 3087
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.269439054447562,
+      "learning_rate": 9.281039448630187e-06,
+      "loss": 0.7382,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2460577883382813,
+      "learning_rate": 9.280503874780112e-06,
+      "loss": 0.6678,
+      "step": 3089
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8371824875811347,
+      "learning_rate": 9.279968116987253e-06,
+      "loss": 0.8011,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8751359498746454,
+      "learning_rate": 9.27943217527463e-06,
+      "loss": 0.782,
+      "step": 3091
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9807189536897207,
+      "learning_rate": 9.278896049665276e-06,
+      "loss": 0.6973,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.155102508326652,
+      "learning_rate": 9.27835974018223e-06,
+      "loss": 0.7891,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.567152786163554,
+      "learning_rate": 9.277823246848537e-06,
+      "loss": 0.8843,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1655930731977402,
+      "learning_rate": 9.277286569687252e-06,
+      "loss": 0.6723,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7096477865215123,
+      "learning_rate": 9.276749708721438e-06,
+      "loss": 0.7505,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8582741863326036,
+      "learning_rate": 9.276212663974163e-06,
+      "loss": 0.8841,
+      "step": 3097
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8309859836324336,
+      "learning_rate": 9.275675435468505e-06,
+      "loss": 0.6811,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7464544270026987,
+      "learning_rate": 9.275138023227555e-06,
+      "loss": 0.7944,
+      "step": 3099
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.65120640161193,
+      "learning_rate": 9.274600427274399e-06,
+      "loss": 0.7274,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4949317565053912,
+      "learning_rate": 9.274062647632144e-06,
+      "loss": 0.745,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5825909814418524,
+      "learning_rate": 9.273524684323897e-06,
+      "loss": 0.6703,
+      "step": 3102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6494527064858753,
+      "learning_rate": 9.27298653737278e-06,
+      "loss": 0.8308,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9095915606616332,
+      "learning_rate": 9.272448206801912e-06,
+      "loss": 0.7261,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.178274804796751,
+      "learning_rate": 9.271909692634431e-06,
+      "loss": 0.695,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8103240826931113,
+      "learning_rate": 9.271370994893475e-06,
+      "loss": 0.7911,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6136355395409039,
+      "learning_rate": 9.270832113602196e-06,
+      "loss": 0.7395,
+      "step": 3107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7771685001320168,
+      "learning_rate": 9.270293048783747e-06,
+      "loss": 0.709,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.070651920542647,
+      "learning_rate": 9.269753800461299e-06,
+      "loss": 0.8512,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7586508347345906,
+      "learning_rate": 9.269214368658016e-06,
+      "loss": 0.7854,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5136289647679135,
+      "learning_rate": 9.268674753397087e-06,
+      "loss": 0.7641,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3471493733803683,
+      "learning_rate": 9.268134954701697e-06,
+      "loss": 0.6391,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.720861703918287,
+      "learning_rate": 9.267594972595042e-06,
+      "loss": 0.9709,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.188774803759399,
+      "learning_rate": 9.267054807100327e-06,
+      "loss": 0.8264,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.117945762058685,
+      "learning_rate": 9.266514458240762e-06,
+      "loss": 0.6751,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5127921192480211,
+      "learning_rate": 9.26597392603957e-06,
+      "loss": 0.7954,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1243267576589195,
+      "learning_rate": 9.265433210519976e-06,
+      "loss": 0.7282,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.156981304982138,
+      "learning_rate": 9.26489231170522e-06,
+      "loss": 0.664,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.611665335457097,
+      "learning_rate": 9.264351229618541e-06,
+      "loss": 0.7753,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.66838484486203,
+      "learning_rate": 9.263809964283192e-06,
+      "loss": 0.6996,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8584624312256968,
+      "learning_rate": 9.263268515722435e-06,
+      "loss": 0.8994,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6439699230016036,
+      "learning_rate": 9.262726883959535e-06,
+      "loss": 0.7375,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5720817939257126,
+      "learning_rate": 9.262185069017764e-06,
+      "loss": 0.8332,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.60004802166346,
+      "learning_rate": 9.261643070920409e-06,
+      "loss": 0.8205,
+      "step": 3124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2588482334221476,
+      "learning_rate": 9.261100889690763e-06,
+      "loss": 0.6822,
+      "step": 3125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.584485485846514,
+      "learning_rate": 9.26055852535212e-06,
+      "loss": 0.8555,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9836186570596037,
+      "learning_rate": 9.260015977927788e-06,
+      "loss": 0.7596,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1723663899119567,
+      "learning_rate": 9.259473247441081e-06,
+      "loss": 0.6669,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.758313019166394,
+      "learning_rate": 9.258930333915325e-06,
+      "loss": 0.764,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6869591488691775,
+      "learning_rate": 9.258387237373845e-06,
+      "loss": 0.7496,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9660732784272077,
+      "learning_rate": 9.257843957839982e-06,
+      "loss": 0.8246,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.701590357548018,
+      "learning_rate": 9.257300495337082e-06,
+      "loss": 0.7354,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.545164480720529,
+      "learning_rate": 9.256756849888499e-06,
+      "loss": 0.7834,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7136775108109876,
+      "learning_rate": 9.256213021517593e-06,
+      "loss": 0.8573,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.920099861300204,
+      "learning_rate": 9.255669010247734e-06,
+      "loss": 0.7471,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8555149998487075,
+      "learning_rate": 9.2551248161023e-06,
+      "loss": 0.879,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8336185163257879,
+      "learning_rate": 9.254580439104677e-06,
+      "loss": 0.857,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.286956297496952,
+      "learning_rate": 9.254035879278255e-06,
+      "loss": 0.7489,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8607229407616688,
+      "learning_rate": 9.253491136646437e-06,
+      "loss": 0.8477,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1905305302536946,
+      "learning_rate": 9.252946211232634e-06,
+      "loss": 0.8865,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1007214342556142,
+      "learning_rate": 9.25240110306026e-06,
+      "loss": 0.6363,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8655668496766773,
+      "learning_rate": 9.25185581215274e-06,
+      "loss": 0.8519,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1201324512871487,
+      "learning_rate": 9.251310338533504e-06,
+      "loss": 0.6285,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5609201277042635,
+      "learning_rate": 9.250764682225997e-06,
+      "loss": 0.8685,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8374504843925916,
+      "learning_rate": 9.250218843253663e-06,
+      "loss": 0.816,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8825966794955502,
+      "learning_rate": 9.24967282163996e-06,
+      "loss": 0.7825,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5299556195221697,
+      "learning_rate": 9.249126617408353e-06,
+      "loss": 0.701,
+      "step": 3147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.302428946384016,
+      "learning_rate": 9.248580230582311e-06,
+      "loss": 0.8803,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6516632614522218,
+      "learning_rate": 9.248033661185313e-06,
+      "loss": 0.7449,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3579606680375735,
+      "learning_rate": 9.247486909240849e-06,
+      "loss": 0.6074,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2219856171073333,
+      "learning_rate": 9.246939974772413e-06,
+      "loss": 0.6251,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.306194856235809,
+      "learning_rate": 9.246392857803508e-06,
+      "loss": 0.7878,
+      "step": 3152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.761620897699683,
+      "learning_rate": 9.245845558357643e-06,
+      "loss": 0.77,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8943766292210422,
+      "learning_rate": 9.24529807645834e-06,
+      "loss": 0.7436,
+      "step": 3154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5263935064903218,
+      "learning_rate": 9.244750412129123e-06,
+      "loss": 0.7842,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1727424443156464,
+      "learning_rate": 9.244202565393528e-06,
+      "loss": 0.8556,
+      "step": 3156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5848036753280672,
+      "learning_rate": 9.243654536275095e-06,
+      "loss": 0.8704,
+      "step": 3157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9708393186480766,
+      "learning_rate": 9.243106324797377e-06,
+      "loss": 0.8004,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6948705058117761,
+      "learning_rate": 9.24255793098393e-06,
+      "loss": 0.8735,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6733467805262587,
+      "learning_rate": 9.24200935485832e-06,
+      "loss": 0.8522,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7145567070577727,
+      "learning_rate": 9.24146059644412e-06,
+      "loss": 0.7968,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.143947246149112,
+      "learning_rate": 9.240911655764912e-06,
+      "loss": 0.5243,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6923265535384748,
+      "learning_rate": 9.240362532844287e-06,
+      "loss": 0.8018,
+      "step": 3163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1982379690759177,
+      "learning_rate": 9.23981322770584e-06,
+      "loss": 0.6368,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8019059054756903,
+      "learning_rate": 9.239263740373173e-06,
+      "loss": 0.7301,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5644370280955906,
+      "learning_rate": 9.238714070869903e-06,
+      "loss": 0.6799,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6836135207405607,
+      "learning_rate": 9.238164219219649e-06,
+      "loss": 0.877,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4766063051131875,
+      "learning_rate": 9.23761418544604e-06,
+      "loss": 0.8914,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.46007914986976,
+      "learning_rate": 9.237063969572713e-06,
+      "loss": 0.6587,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6515002939856065,
+      "learning_rate": 9.23651357162331e-06,
+      "loss": 0.815,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8073321157632294,
+      "learning_rate": 9.235962991621484e-06,
+      "loss": 0.6975,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5364098356718534,
+      "learning_rate": 9.235412229590893e-06,
+      "loss": 0.8088,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2811405736578403,
+      "learning_rate": 9.234861285555207e-06,
+      "loss": 0.6106,
+      "step": 3173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7641921548954644,
+      "learning_rate": 9.2343101595381e-06,
+      "loss": 0.8615,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9197361480550483,
+      "learning_rate": 9.233758851563254e-06,
+      "loss": 0.808,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9042655784429443,
+      "learning_rate": 9.233207361654362e-06,
+      "loss": 0.804,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5053505804083922,
+      "learning_rate": 9.23265568983512e-06,
+      "loss": 0.6022,
+      "step": 3177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.090875436955105,
+      "learning_rate": 9.232103836129239e-06,
+      "loss": 0.7443,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6865122687831373,
+      "learning_rate": 9.23155180056043e-06,
+      "loss": 0.8402,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.097017130496829,
+      "learning_rate": 9.230999583152413e-06,
+      "loss": 0.7664,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2957307791138077,
+      "learning_rate": 9.230447183928926e-06,
+      "loss": 0.6437,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5190608246330912,
+      "learning_rate": 9.229894602913698e-06,
+      "loss": 0.8222,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.690003668019586,
+      "learning_rate": 9.22934184013048e-06,
+      "loss": 0.8227,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9259791071889358,
+      "learning_rate": 9.228788895603024e-06,
+      "loss": 0.8397,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3636781962201467,
+      "learning_rate": 9.22823576935509e-06,
+      "loss": 0.7126,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.089548894932977,
+      "learning_rate": 9.22768246141045e-06,
+      "loss": 0.7658,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1715411701103537,
+      "learning_rate": 9.227128971792877e-06,
+      "loss": 0.6953,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5828974954490358,
+      "learning_rate": 9.226575300526159e-06,
+      "loss": 0.7365,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8482718709297967,
+      "learning_rate": 9.226021447634085e-06,
+      "loss": 0.7829,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2256662064306205,
+      "learning_rate": 9.22546741314046e-06,
+      "loss": 0.8139,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.608120880696179,
+      "learning_rate": 9.22491319706909e-06,
+      "loss": 0.7593,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.193873415684634,
+      "learning_rate": 9.224358799443791e-06,
+      "loss": 0.6249,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7208328597064042,
+      "learning_rate": 9.223804220288384e-06,
+      "loss": 0.8097,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0215446479617576,
+      "learning_rate": 9.223249459626704e-06,
+      "loss": 0.8484,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.621784775330957,
+      "learning_rate": 9.22269451748259e-06,
+      "loss": 0.8222,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7980080787586514,
+      "learning_rate": 9.222139393879885e-06,
+      "loss": 0.878,
+      "step": 3196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7896175146037723,
+      "learning_rate": 9.22158408884245e-06,
+      "loss": 0.8185,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7689185200235849,
+      "learning_rate": 9.221028602394144e-06,
+      "loss": 0.8546,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8813439678211237,
+      "learning_rate": 9.220472934558838e-06,
+      "loss": 0.7682,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7081217141588227,
+      "learning_rate": 9.21991708536041e-06,
+      "loss": 0.7502,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.747745839968018,
+      "learning_rate": 9.219361054822746e-06,
+      "loss": 0.7647,
+      "step": 3201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4878593117192507,
+      "learning_rate": 9.218804842969742e-06,
+      "loss": 0.7433,
+      "step": 3202
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9031466016741718,
+      "learning_rate": 9.218248449825297e-06,
+      "loss": 0.8759,
+      "step": 3203
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6981885381368154,
+      "learning_rate": 9.217691875413323e-06,
+      "loss": 0.8336,
+      "step": 3204
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8000236371388065,
+      "learning_rate": 9.217135119757734e-06,
+      "loss": 0.8302,
+      "step": 3205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.035942084903227,
+      "learning_rate": 9.216578182882459e-06,
+      "loss": 0.5793,
+      "step": 3206
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1713543858912077,
+      "learning_rate": 9.216021064811428e-06,
+      "loss": 0.767,
+      "step": 3207
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0801178171518906,
+      "learning_rate": 9.21546376556858e-06,
+      "loss": 0.7148,
+      "step": 3208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0194869838593674,
+      "learning_rate": 9.214906285177867e-06,
+      "loss": 0.772,
+      "step": 3209
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.891623611180714,
+      "learning_rate": 9.214348623663244e-06,
+      "loss": 0.8247,
+      "step": 3210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7658297972412564,
+      "learning_rate": 9.213790781048677e-06,
+      "loss": 0.7,
+      "step": 3211
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.890175370229312,
+      "learning_rate": 9.213232757358133e-06,
+      "loss": 0.6926,
+      "step": 3212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.76818038270597,
+      "learning_rate": 9.212674552615594e-06,
+      "loss": 0.8687,
+      "step": 3213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1254124488617243,
+      "learning_rate": 9.212116166845048e-06,
+      "loss": 0.6797,
+      "step": 3214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.772569362330218,
+      "learning_rate": 9.211557600070489e-06,
+      "loss": 0.7167,
+      "step": 3215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.714477452638042,
+      "learning_rate": 9.21099885231592e-06,
+      "loss": 0.7493,
+      "step": 3216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8350486021176602,
+      "learning_rate": 9.210439923605352e-06,
+      "loss": 0.7587,
+      "step": 3217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5963428719479693,
+      "learning_rate": 9.209880813962806e-06,
+      "loss": 0.8001,
+      "step": 3218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6880554504700336,
+      "learning_rate": 9.209321523412303e-06,
+      "loss": 0.8333,
+      "step": 3219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0774632238814412,
+      "learning_rate": 9.208762051977879e-06,
+      "loss": 0.643,
+      "step": 3220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6677662738949643,
+      "learning_rate": 9.208202399683577e-06,
+      "loss": 0.7606,
+      "step": 3221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8578749149767595,
+      "learning_rate": 9.207642566553445e-06,
+      "loss": 0.7453,
+      "step": 3222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8075723064625757,
+      "learning_rate": 9.207082552611541e-06,
+      "loss": 0.781,
+      "step": 3223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2766072306871599,
+      "learning_rate": 9.206522357881931e-06,
+      "loss": 0.6531,
+      "step": 3224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7233382957530976,
+      "learning_rate": 9.205961982388686e-06,
+      "loss": 0.7721,
+      "step": 3225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7711259641368575,
+      "learning_rate": 9.205401426155889e-06,
+      "loss": 0.8092,
+      "step": 3226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8558608804852623,
+      "learning_rate": 9.204840689207626e-06,
+      "loss": 0.7742,
+      "step": 3227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5928570220268377,
+      "learning_rate": 9.204279771567994e-06,
+      "loss": 0.6941,
+      "step": 3228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6911952013494271,
+      "learning_rate": 9.203718673261098e-06,
+      "loss": 0.7947,
+      "step": 3229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5257472323347996,
+      "learning_rate": 9.203157394311045e-06,
+      "loss": 0.6976,
+      "step": 3230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.651452493022971,
+      "learning_rate": 9.20259593474196e-06,
+      "loss": 0.8712,
+      "step": 3231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7095632237683116,
+      "learning_rate": 9.202034294577968e-06,
+      "loss": 0.8809,
+      "step": 3232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.230040051397984,
+      "learning_rate": 9.201472473843204e-06,
+      "loss": 0.5939,
+      "step": 3233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2205616947583593,
+      "learning_rate": 9.20091047256181e-06,
+      "loss": 0.7304,
+      "step": 3234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7855800899793761,
+      "learning_rate": 9.200348290757937e-06,
+      "loss": 0.7964,
+      "step": 3235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1215570600576537,
+      "learning_rate": 9.199785928455745e-06,
+      "loss": 0.6432,
+      "step": 3236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8474493519628274,
+      "learning_rate": 9.199223385679396e-06,
+      "loss": 0.8081,
+      "step": 3237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4907961482488035,
+      "learning_rate": 9.198660662453066e-06,
+      "loss": 0.7372,
+      "step": 3238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.501817479692298,
+      "learning_rate": 9.198097758800938e-06,
+      "loss": 0.8119,
+      "step": 3239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6448276467726124,
+      "learning_rate": 9.197534674747199e-06,
+      "loss": 0.7718,
+      "step": 3240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6800255315504717,
+      "learning_rate": 9.196971410316047e-06,
+      "loss": 0.907,
+      "step": 3241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8975343162236482,
+      "learning_rate": 9.196407965531686e-06,
+      "loss": 0.773,
+      "step": 3242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.845776825279862,
+      "learning_rate": 9.195844340418328e-06,
+      "loss": 0.8013,
+      "step": 3243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.908284281463583,
+      "learning_rate": 9.195280535000196e-06,
+      "loss": 0.717,
+      "step": 3244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7002312036639669,
+      "learning_rate": 9.194716549301514e-06,
+      "loss": 0.8413,
+      "step": 3245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7630676554651146,
+      "learning_rate": 9.19415238334652e-06,
+      "loss": 0.7706,
+      "step": 3246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6655195974377817,
+      "learning_rate": 9.193588037159457e-06,
+      "loss": 0.7241,
+      "step": 3247
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1585725809910394,
+      "learning_rate": 9.193023510764578e-06,
+      "loss": 0.7005,
+      "step": 3248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7071949445169012,
+      "learning_rate": 9.19245880418614e-06,
+      "loss": 0.8061,
+      "step": 3249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9979540912409579,
+      "learning_rate": 9.191893917448409e-06,
+      "loss": 0.634,
+      "step": 3250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6299589812050093,
+      "learning_rate": 9.19132885057566e-06,
+      "loss": 0.7637,
+      "step": 3251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.55566435404803,
+      "learning_rate": 9.190763603592177e-06,
+      "loss": 0.7785,
+      "step": 3252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8902637587301947,
+      "learning_rate": 9.190198176522249e-06,
+      "loss": 0.7061,
+      "step": 3253
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7518187691640994,
+      "learning_rate": 9.189632569390172e-06,
+      "loss": 0.939,
+      "step": 3254
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.702064883885696,
+      "learning_rate": 9.189066782220253e-06,
+      "loss": 0.7795,
+      "step": 3255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7000701242442537,
+      "learning_rate": 9.188500815036806e-06,
+      "loss": 0.8124,
+      "step": 3256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6321613935144992,
+      "learning_rate": 9.18793466786415e-06,
+      "loss": 0.698,
+      "step": 3257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0986285255257293,
+      "learning_rate": 9.187368340726615e-06,
+      "loss": 0.7997,
+      "step": 3258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9804591495849613,
+      "learning_rate": 9.186801833648535e-06,
+      "loss": 0.6794,
+      "step": 3259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6045547694291447,
+      "learning_rate": 9.186235146654257e-06,
+      "loss": 0.9086,
+      "step": 3260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.649770528492737,
+      "learning_rate": 9.18566827976813e-06,
+      "loss": 0.7422,
+      "step": 3261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.304316710584783,
+      "learning_rate": 9.185101233014516e-06,
+      "loss": 0.6537,
+      "step": 3262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7080649821840623,
+      "learning_rate": 9.184534006417782e-06,
+      "loss": 0.7127,
+      "step": 3263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6066439406555917,
+      "learning_rate": 9.183966600002301e-06,
+      "loss": 0.6909,
+      "step": 3264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6952388974506827,
+      "learning_rate": 9.183399013792457e-06,
+      "loss": 0.8601,
+      "step": 3265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.293415018476136,
+      "learning_rate": 9.18283124781264e-06,
+      "loss": 0.7857,
+      "step": 3266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7714670738222098,
+      "learning_rate": 9.18226330208725e-06,
+      "loss": 0.811,
+      "step": 3267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0782428367091796,
+      "learning_rate": 9.181695176640693e-06,
+      "loss": 0.6744,
+      "step": 3268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6245638336407549,
+      "learning_rate": 9.181126871497378e-06,
+      "loss": 0.8585,
+      "step": 3269
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5659617400230663,
+      "learning_rate": 9.180558386681731e-06,
+      "loss": 0.7879,
+      "step": 3270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.681073213106985,
+      "learning_rate": 9.179989722218179e-06,
+      "loss": 0.7928,
+      "step": 3271
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.527407725128882,
+      "learning_rate": 9.17942087813116e-06,
+      "loss": 0.6983,
+      "step": 3272
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6107387314090735,
+      "learning_rate": 9.178851854445118e-06,
+      "loss": 0.8307,
+      "step": 3273
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7165408447469677,
+      "learning_rate": 9.178282651184506e-06,
+      "loss": 0.7474,
+      "step": 3274
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6585385754978337,
+      "learning_rate": 9.177713268373781e-06,
+      "loss": 0.709,
+      "step": 3275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.774717312406015,
+      "learning_rate": 9.177143706037411e-06,
+      "loss": 0.7985,
+      "step": 3276
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8995827183914527,
+      "learning_rate": 9.176573964199875e-06,
+      "loss": 0.8446,
+      "step": 3277
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8875516748668673,
+      "learning_rate": 9.176004042885653e-06,
+      "loss": 0.898,
+      "step": 3278
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5847604820944459,
+      "learning_rate": 9.175433942119238e-06,
+      "loss": 0.7854,
+      "step": 3279
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.289955356411128,
+      "learning_rate": 9.174863661925128e-06,
+      "loss": 0.7864,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.662388602337828,
+      "learning_rate": 9.174293202327828e-06,
+      "loss": 0.792,
+      "step": 3281
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7354551557391624,
+      "learning_rate": 9.173722563351852e-06,
+      "loss": 0.998,
+      "step": 3282
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7412424136346867,
+      "learning_rate": 9.173151745021722e-06,
+      "loss": 0.8028,
+      "step": 3283
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.687326381374173,
+      "learning_rate": 9.172580747361968e-06,
+      "loss": 0.8065,
+      "step": 3284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8335906409556524,
+      "learning_rate": 9.172009570397124e-06,
+      "loss": 0.8596,
+      "step": 3285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6707965566695315,
+      "learning_rate": 9.17143821415174e-06,
+      "loss": 0.731,
+      "step": 3286
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.008479848157415,
+      "learning_rate": 9.170866678650366e-06,
+      "loss": 0.824,
+      "step": 3287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8030290206160884,
+      "learning_rate": 9.17029496391756e-06,
+      "loss": 0.8755,
+      "step": 3288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7321639454873115,
+      "learning_rate": 9.169723069977892e-06,
+      "loss": 0.6867,
+      "step": 3289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7492181695952813,
+      "learning_rate": 9.169150996855939e-06,
+      "loss": 0.7927,
+      "step": 3290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8918206252888885,
+      "learning_rate": 9.16857874457628e-06,
+      "loss": 0.79,
+      "step": 3291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.962085205305974,
+      "learning_rate": 9.16800631316351e-06,
+      "loss": 0.7706,
+      "step": 3292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.725544579080786,
+      "learning_rate": 9.167433702642224e-06,
+      "loss": 0.8046,
+      "step": 3293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6359033535258096,
+      "learning_rate": 9.166860913037032e-06,
+      "loss": 0.8255,
+      "step": 3294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7240160611763675,
+      "learning_rate": 9.166287944372546e-06,
+      "loss": 0.7454,
+      "step": 3295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8438607791449755,
+      "learning_rate": 9.165714796673389e-06,
+      "loss": 0.7724,
+      "step": 3296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5614514605285914,
+      "learning_rate": 9.16514146996419e-06,
+      "loss": 0.7625,
+      "step": 3297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8406957959909558,
+      "learning_rate": 9.164567964269583e-06,
+      "loss": 0.7947,
+      "step": 3298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.506936896696113,
+      "learning_rate": 9.163994279614218e-06,
+      "loss": 0.7966,
+      "step": 3299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8946860074151206,
+      "learning_rate": 9.163420416022745e-06,
+      "loss": 0.7801,
+      "step": 3300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7269821129958893,
+      "learning_rate": 9.162846373519824e-06,
+      "loss": 0.863,
+      "step": 3301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5606110549268009,
+      "learning_rate": 9.162272152130123e-06,
+      "loss": 0.734,
+      "step": 3302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6773678999855088,
+      "learning_rate": 9.161697751878317e-06,
+      "loss": 0.861,
+      "step": 3303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1421375204201483,
+      "learning_rate": 9.161123172789091e-06,
+      "loss": 0.6969,
+      "step": 3304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7055782079760968,
+      "learning_rate": 9.160548414887135e-06,
+      "loss": 0.7697,
+      "step": 3305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2460402148740903,
+      "learning_rate": 9.159973478197148e-06,
+      "loss": 0.7008,
+      "step": 3306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.424031227224806,
+      "learning_rate": 9.159398362743834e-06,
+      "loss": 0.7202,
+      "step": 3307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.726604666215194,
+      "learning_rate": 9.15882306855191e-06,
+      "loss": 0.894,
+      "step": 3308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8176604430405576,
+      "learning_rate": 9.158247595646098e-06,
+      "loss": 0.7674,
+      "step": 3309
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4869435251887404,
+      "learning_rate": 9.157671944051125e-06,
+      "loss": 0.7558,
+      "step": 3310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6995158668033652,
+      "learning_rate": 9.157096113791727e-06,
+      "loss": 0.816,
+      "step": 3311
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0429033681766329,
+      "learning_rate": 9.156520104892653e-06,
+      "loss": 0.6684,
+      "step": 3312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4233949159655888,
+      "learning_rate": 9.155943917378652e-06,
+      "loss": 0.6995,
+      "step": 3313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7105533623441207,
+      "learning_rate": 9.155367551274485e-06,
+      "loss": 0.8214,
+      "step": 3314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3193943600672677,
+      "learning_rate": 9.15479100660492e-06,
+      "loss": 0.6788,
+      "step": 3315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8160704280033093,
+      "learning_rate": 9.154214283394733e-06,
+      "loss": 0.8341,
+      "step": 3316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8083072953944264,
+      "learning_rate": 9.153637381668705e-06,
+      "loss": 0.7638,
+      "step": 3317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9346498853593048,
+      "learning_rate": 9.153060301451629e-06,
+      "loss": 0.6494,
+      "step": 3318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7418405007971434,
+      "learning_rate": 9.152483042768302e-06,
+      "loss": 0.7827,
+      "step": 3319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.038861872425429,
+      "learning_rate": 9.151905605643531e-06,
+      "loss": 0.8335,
+      "step": 3320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5864750126960196,
+      "learning_rate": 9.151327990102129e-06,
+      "loss": 0.6702,
+      "step": 3321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6741375292171832,
+      "learning_rate": 9.150750196168918e-06,
+      "loss": 0.7102,
+      "step": 3322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5479092049462577,
+      "learning_rate": 9.150172223868727e-06,
+      "loss": 0.8071,
+      "step": 3323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6011190763258054,
+      "learning_rate": 9.149594073226391e-06,
+      "loss": 0.779,
+      "step": 3324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.750080153576488,
+      "learning_rate": 9.149015744266759e-06,
+      "loss": 0.8465,
+      "step": 3325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.679277193821173,
+      "learning_rate": 9.148437237014677e-06,
+      "loss": 0.7806,
+      "step": 3326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.915581258530193,
+      "learning_rate": 9.14785855149501e-06,
+      "loss": 0.8645,
+      "step": 3327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7626131223694024,
+      "learning_rate": 9.147279687732622e-06,
+      "loss": 0.8663,
+      "step": 3328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4941632924034292,
+      "learning_rate": 9.14670064575239e-06,
+      "loss": 0.819,
+      "step": 3329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7908775519131857,
+      "learning_rate": 9.146121425579197e-06,
+      "loss": 0.923,
+      "step": 3330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6084027968220744,
+      "learning_rate": 9.14554202723793e-06,
+      "loss": 0.838,
+      "step": 3331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2371964291825506,
+      "learning_rate": 9.144962450753491e-06,
+      "loss": 0.6906,
+      "step": 3332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0358111237333345,
+      "learning_rate": 9.144382696150785e-06,
+      "loss": 0.6847,
+      "step": 3333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6279121007816806,
+      "learning_rate": 9.143802763454723e-06,
+      "loss": 0.7864,
+      "step": 3334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7213812828321997,
+      "learning_rate": 9.143222652690227e-06,
+      "loss": 0.7361,
+      "step": 3335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6997956254685782,
+      "learning_rate": 9.142642363882228e-06,
+      "loss": 0.7612,
+      "step": 3336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.894787732826655,
+      "learning_rate": 9.14206189705566e-06,
+      "loss": 0.8477,
+      "step": 3337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.897124470975648,
+      "learning_rate": 9.141481252235467e-06,
+      "loss": 0.7358,
+      "step": 3338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.086115347580932,
+      "learning_rate": 9.140900429446601e-06,
+      "loss": 0.6723,
+      "step": 3339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.991838455367142,
+      "learning_rate": 9.140319428714022e-06,
+      "loss": 0.8974,
+      "step": 3340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7244265368457905,
+      "learning_rate": 9.139738250062695e-06,
+      "loss": 0.8599,
+      "step": 3341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.302633007057662,
+      "learning_rate": 9.139156893517595e-06,
+      "loss": 0.6658,
+      "step": 3342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.585447044086551,
+      "learning_rate": 9.138575359103707e-06,
+      "loss": 0.7407,
+      "step": 3343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6677900680518534,
+      "learning_rate": 9.137993646846018e-06,
+      "loss": 0.7696,
+      "step": 3344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.626205733954862,
+      "learning_rate": 9.137411756769525e-06,
+      "loss": 0.7882,
+      "step": 3345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7996953095696435,
+      "learning_rate": 9.136829688899236e-06,
+      "loss": 0.7516,
+      "step": 3346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.3427207829672465,
+      "learning_rate": 9.136247443260162e-06,
+      "loss": 0.7881,
+      "step": 3347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.933316688125392,
+      "learning_rate": 9.135665019877323e-06,
+      "loss": 0.6606,
+      "step": 3348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6471319215660014,
+      "learning_rate": 9.135082418775746e-06,
+      "loss": 0.7243,
+      "step": 3349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5101310822316818,
+      "learning_rate": 9.134499639980469e-06,
+      "loss": 0.7601,
+      "step": 3350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0157641011353666,
+      "learning_rate": 9.133916683516536e-06,
+      "loss": 0.8273,
+      "step": 3351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.110005843330155,
+      "learning_rate": 9.133333549408997e-06,
+      "loss": 0.7618,
+      "step": 3352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7621042000994276,
+      "learning_rate": 9.132750237682907e-06,
+      "loss": 0.6953,
+      "step": 3353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8861180844176924,
+      "learning_rate": 9.132166748363335e-06,
+      "loss": 0.7946,
+      "step": 3354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7106114456154586,
+      "learning_rate": 9.131583081475356e-06,
+      "loss": 0.7063,
+      "step": 3355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6652022109003177,
+      "learning_rate": 9.130999237044052e-06,
+      "loss": 0.799,
+      "step": 3356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6441069321972177,
+      "learning_rate": 9.13041521509451e-06,
+      "loss": 0.7819,
+      "step": 3357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.950954513681022,
+      "learning_rate": 9.129831015651827e-06,
+      "loss": 0.915,
+      "step": 3358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7539715996960537,
+      "learning_rate": 9.129246638741108e-06,
+      "loss": 0.7546,
+      "step": 3359
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1236439024726947,
+      "learning_rate": 9.128662084387462e-06,
+      "loss": 0.5215,
+      "step": 3360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5573037180810936,
+      "learning_rate": 9.128077352616013e-06,
+      "loss": 0.6431,
+      "step": 3361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6744285194722794,
+      "learning_rate": 9.127492443451887e-06,
+      "loss": 0.8964,
+      "step": 3362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6518047208746782,
+      "learning_rate": 9.126907356920219e-06,
+      "loss": 0.7303,
+      "step": 3363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.613111614090439,
+      "learning_rate": 9.126322093046149e-06,
+      "loss": 0.9352,
+      "step": 3364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6876559664159925,
+      "learning_rate": 9.12573665185483e-06,
+      "loss": 0.7208,
+      "step": 3365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5385931539341784,
+      "learning_rate": 9.125151033371417e-06,
+      "loss": 0.6801,
+      "step": 3366
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7845590632310615,
+      "learning_rate": 9.12456523762108e-06,
+      "loss": 0.7589,
+      "step": 3367
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5238789019972592,
+      "learning_rate": 9.123979264628985e-06,
+      "loss": 0.7986,
+      "step": 3368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9873634695542508,
+      "learning_rate": 9.123393114420318e-06,
+      "loss": 0.7992,
+      "step": 3369
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6632164203117739,
+      "learning_rate": 9.122806787020266e-06,
+      "loss": 0.8834,
+      "step": 3370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7672236963011296,
+      "learning_rate": 9.122220282454024e-06,
+      "loss": 0.8025,
+      "step": 3371
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6253180328243793,
+      "learning_rate": 9.121633600746796e-06,
+      "loss": 0.7904,
+      "step": 3372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6963599580578248,
+      "learning_rate": 9.121046741923792e-06,
+      "loss": 0.7913,
+      "step": 3373
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5216412836033497,
+      "learning_rate": 9.120459706010233e-06,
+      "loss": 0.7558,
+      "step": 3374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7544901532698922,
+      "learning_rate": 9.119872493031343e-06,
+      "loss": 0.8656,
+      "step": 3375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.710990812709986,
+      "learning_rate": 9.119285103012356e-06,
+      "loss": 0.9108,
+      "step": 3376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5881570226271826,
+      "learning_rate": 9.118697535978513e-06,
+      "loss": 0.7742,
+      "step": 3377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.165040288192903,
+      "learning_rate": 9.118109791955067e-06,
+      "loss": 0.7088,
+      "step": 3378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.070011842532096,
+      "learning_rate": 9.11752187096727e-06,
+      "loss": 0.8262,
+      "step": 3379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6904471559168344,
+      "learning_rate": 9.116933773040389e-06,
+      "loss": 0.7745,
+      "step": 3380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9609819445020809,
+      "learning_rate": 9.116345498199693e-06,
+      "loss": 0.7816,
+      "step": 3381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0857949924199242,
+      "learning_rate": 9.115757046470464e-06,
+      "loss": 0.6959,
+      "step": 3382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.656889341293985,
+      "learning_rate": 9.115168417877988e-06,
+      "loss": 0.7816,
+      "step": 3383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7238926233568042,
+      "learning_rate": 9.114579612447562e-06,
+      "loss": 0.8704,
+      "step": 3384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7736808064330478,
+      "learning_rate": 9.113990630204485e-06,
+      "loss": 0.8155,
+      "step": 3385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.658782627081086,
+      "learning_rate": 9.113401471174068e-06,
+      "loss": 0.7515,
+      "step": 3386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6968827832343503,
+      "learning_rate": 9.11281213538163e-06,
+      "loss": 0.845,
+      "step": 3387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8421584164124873,
+      "learning_rate": 9.112222622852494e-06,
+      "loss": 0.7747,
+      "step": 3388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9171291609802514,
+      "learning_rate": 9.111632933611993e-06,
+      "loss": 0.8245,
+      "step": 3389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7778165526922205,
+      "learning_rate": 9.111043067685468e-06,
+      "loss": 0.7543,
+      "step": 3390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.834446349218806,
+      "learning_rate": 9.110453025098267e-06,
+      "loss": 0.7887,
+      "step": 3391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9620307755567831,
+      "learning_rate": 9.109862805875745e-06,
+      "loss": 0.6618,
+      "step": 3392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7914210575833538,
+      "learning_rate": 9.109272410043265e-06,
+      "loss": 0.7632,
+      "step": 3393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6661578487851068,
+      "learning_rate": 9.108681837626199e-06,
+      "loss": 0.8343,
+      "step": 3394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.703765389691244,
+      "learning_rate": 9.108091088649922e-06,
+      "loss": 0.896,
+      "step": 3395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6715041133549307,
+      "learning_rate": 9.107500163139822e-06,
+      "loss": 0.7322,
+      "step": 3396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.846410454780997,
+      "learning_rate": 9.106909061121291e-06,
+      "loss": 0.7695,
+      "step": 3397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1886314596924832,
+      "learning_rate": 9.106317782619733e-06,
+      "loss": 0.6891,
+      "step": 3398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.651226096239,
+      "learning_rate": 9.105726327660556e-06,
+      "loss": 0.8551,
+      "step": 3399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5837585627759174,
+      "learning_rate": 9.105134696269172e-06,
+      "loss": 0.7499,
+      "step": 3400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.26301197670665,
+      "learning_rate": 9.10454288847101e-06,
+      "loss": 0.7767,
+      "step": 3401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8254392799377759,
+      "learning_rate": 9.103950904291496e-06,
+      "loss": 0.7143,
+      "step": 3402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9145346550963225,
+      "learning_rate": 9.103358743756072e-06,
+      "loss": 0.7589,
+      "step": 3403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2977138492179805,
+      "learning_rate": 9.102766406890185e-06,
+      "loss": 0.6576,
+      "step": 3404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8923672403976894,
+      "learning_rate": 9.102173893719288e-06,
+      "loss": 0.806,
+      "step": 3405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5983063235539705,
+      "learning_rate": 9.101581204268843e-06,
+      "loss": 0.8114,
+      "step": 3406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.022726903791898,
+      "learning_rate": 9.100988338564318e-06,
+      "loss": 0.5474,
+      "step": 3407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.29562849636415,
+      "learning_rate": 9.100395296631192e-06,
+      "loss": 0.8467,
+      "step": 3408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.488206677413028,
+      "learning_rate": 9.099802078494947e-06,
+      "loss": 0.7077,
+      "step": 3409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.891352688317962,
+      "learning_rate": 9.099208684181078e-06,
+      "loss": 0.8322,
+      "step": 3410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3068494149582481,
+      "learning_rate": 9.098615113715078e-06,
+      "loss": 0.7388,
+      "step": 3411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6747355183181485,
+      "learning_rate": 9.098021367122462e-06,
+      "loss": 0.7725,
+      "step": 3412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8769428708377183,
+      "learning_rate": 9.097427444428742e-06,
+      "loss": 0.7697,
+      "step": 3413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0923194676460477,
+      "learning_rate": 9.096833345659437e-06,
+      "loss": 0.657,
+      "step": 3414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0542327312091,
+      "learning_rate": 9.096239070840079e-06,
+      "loss": 0.717,
+      "step": 3415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7524190615024262,
+      "learning_rate": 9.095644619996206e-06,
+      "loss": 0.6592,
+      "step": 3416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8547822337457203,
+      "learning_rate": 9.095049993153362e-06,
+      "loss": 0.7872,
+      "step": 3417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8834094101063918,
+      "learning_rate": 9.0944551903371e-06,
+      "loss": 0.7708,
+      "step": 3418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2577418286154403,
+      "learning_rate": 9.09386021157298e-06,
+      "loss": 0.7801,
+      "step": 3419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7406222730652396,
+      "learning_rate": 9.09326505688657e-06,
+      "loss": 0.877,
+      "step": 3420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5815174787044712,
+      "learning_rate": 9.092669726303441e-06,
+      "loss": 0.8939,
+      "step": 3421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.618383884368719,
+      "learning_rate": 9.092074219849181e-06,
+      "loss": 0.7849,
+      "step": 3422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2902492219610764,
+      "learning_rate": 9.09147853754938e-06,
+      "loss": 0.7561,
+      "step": 3423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.135797373821015,
+      "learning_rate": 9.09088267942963e-06,
+      "loss": 0.608,
+      "step": 3424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1488182394970563,
+      "learning_rate": 9.090286645515542e-06,
+      "loss": 0.6811,
+      "step": 3425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6029279452499205,
+      "learning_rate": 9.089690435832729e-06,
+      "loss": 0.7435,
+      "step": 3426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7503254901012668,
+      "learning_rate": 9.089094050406808e-06,
+      "loss": 0.763,
+      "step": 3427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6546158805045381,
+      "learning_rate": 9.088497489263411e-06,
+      "loss": 0.8031,
+      "step": 3428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6355855983503371,
+      "learning_rate": 9.087900752428168e-06,
+      "loss": 0.7318,
+      "step": 3429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6033061044297559,
+      "learning_rate": 9.087303839926727e-06,
+      "loss": 0.6994,
+      "step": 3430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8021286563500571,
+      "learning_rate": 9.086706751784738e-06,
+      "loss": 0.8026,
+      "step": 3431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.557526469242445,
+      "learning_rate": 9.086109488027857e-06,
+      "loss": 0.751,
+      "step": 3432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7317844020221178,
+      "learning_rate": 9.08551204868175e-06,
+      "loss": 0.8756,
+      "step": 3433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0220451729487694,
+      "learning_rate": 9.084914433772094e-06,
+      "loss": 0.6675,
+      "step": 3434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7932301460263225,
+      "learning_rate": 9.084316643324567e-06,
+      "loss": 0.686,
+      "step": 3435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6323031131396433,
+      "learning_rate": 9.083718677364855e-06,
+      "loss": 0.8239,
+      "step": 3436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6738432454674572,
+      "learning_rate": 9.08312053591866e-06,
+      "loss": 0.8006,
+      "step": 3437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6885552196999085,
+      "learning_rate": 9.082522219011678e-06,
+      "loss": 0.8045,
+      "step": 3438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8579840630590474,
+      "learning_rate": 9.081923726669626e-06,
+      "loss": 0.8566,
+      "step": 3439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7916656864823044,
+      "learning_rate": 9.081325058918222e-06,
+      "loss": 0.8083,
+      "step": 3440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7457626259168588,
+      "learning_rate": 9.080726215783189e-06,
+      "loss": 0.7413,
+      "step": 3441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6735712585861555,
+      "learning_rate": 9.080127197290262e-06,
+      "loss": 0.7842,
+      "step": 3442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0953792539685065,
+      "learning_rate": 9.079528003465183e-06,
+      "loss": 0.6322,
+      "step": 3443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.037742057158698,
+      "learning_rate": 9.0789286343337e-06,
+      "loss": 0.7968,
+      "step": 3444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1910721251056922,
+      "learning_rate": 9.07832908992157e-06,
+      "loss": 0.7017,
+      "step": 3445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0628498094640055,
+      "learning_rate": 9.077729370254555e-06,
+      "loss": 0.5165,
+      "step": 3446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7384652670291738,
+      "learning_rate": 9.077129475358426e-06,
+      "loss": 0.8382,
+      "step": 3447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5650572756158003,
+      "learning_rate": 9.076529405258966e-06,
+      "loss": 0.7547,
+      "step": 3448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.882401562517371,
+      "learning_rate": 9.075929159981957e-06,
+      "loss": 0.9093,
+      "step": 3449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4947244366032475,
+      "learning_rate": 9.075328739553195e-06,
+      "loss": 0.6593,
+      "step": 3450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.285535669366235,
+      "learning_rate": 9.07472814399848e-06,
+      "loss": 0.8309,
+      "step": 3451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.471929566248201,
+      "learning_rate": 9.074127373343623e-06,
+      "loss": 0.7041,
+      "step": 3452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5141049565509286,
+      "learning_rate": 9.07352642761444e-06,
+      "loss": 0.7105,
+      "step": 3453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.820083129386083,
+      "learning_rate": 9.072925306836751e-06,
+      "loss": 0.843,
+      "step": 3454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7806665880802588,
+      "learning_rate": 9.072324011036392e-06,
+      "loss": 0.765,
+      "step": 3455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7392109037580952,
+      "learning_rate": 9.071722540239203e-06,
+      "loss": 0.8791,
+      "step": 3456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4584091843132323,
+      "learning_rate": 9.071120894471026e-06,
+      "loss": 0.6175,
+      "step": 3457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1509162144106246,
+      "learning_rate": 9.070519073757717e-06,
+      "loss": 0.6057,
+      "step": 3458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9137144427182091,
+      "learning_rate": 9.06991707812514e-06,
+      "loss": 0.8988,
+      "step": 3459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9856706899440422,
+      "learning_rate": 9.069314907599159e-06,
+      "loss": 0.7177,
+      "step": 3460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.749954564709225,
+      "learning_rate": 9.068712562205655e-06,
+      "loss": 0.787,
+      "step": 3461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7950683136585657,
+      "learning_rate": 9.06811004197051e-06,
+      "loss": 0.8391,
+      "step": 3462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6494570478919492,
+      "learning_rate": 9.067507346919618e-06,
+      "loss": 0.7365,
+      "step": 3463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6254370953178086,
+      "learning_rate": 9.066904477078875e-06,
+      "loss": 0.8663,
+      "step": 3464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9201983563895473,
+      "learning_rate": 9.06630143247419e-06,
+      "loss": 0.7037,
+      "step": 3465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9659507037494068,
+      "learning_rate": 9.065698213131477e-06,
+      "loss": 0.809,
+      "step": 3466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7102270813948346,
+      "learning_rate": 9.065094819076655e-06,
+      "loss": 0.6905,
+      "step": 3467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9067937149681882,
+      "learning_rate": 9.064491250335656e-06,
+      "loss": 0.837,
+      "step": 3468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8713627291152573,
+      "learning_rate": 9.063887506934417e-06,
+      "loss": 0.7647,
+      "step": 3469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.550457680968236,
+      "learning_rate": 9.06328358889888e-06,
+      "loss": 0.7006,
+      "step": 3470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8135142324176423,
+      "learning_rate": 9.062679496254997e-06,
+      "loss": 0.6891,
+      "step": 3471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.652891143794443,
+      "learning_rate": 9.062075229028728e-06,
+      "loss": 0.6842,
+      "step": 3472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5393605696388664,
+      "learning_rate": 9.061470787246042e-06,
+      "loss": 0.7119,
+      "step": 3473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3202190318062048,
+      "learning_rate": 9.06086617093291e-06,
+      "loss": 0.788,
+      "step": 3474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7632970130247048,
+      "learning_rate": 9.060261380115314e-06,
+      "loss": 0.7586,
+      "step": 3475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7802220062697893,
+      "learning_rate": 9.059656414819243e-06,
+      "loss": 0.7326,
+      "step": 3476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6047268526914122,
+      "learning_rate": 9.059051275070696e-06,
+      "loss": 0.8099,
+      "step": 3477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5510438285080366,
+      "learning_rate": 9.058445960895674e-06,
+      "loss": 0.6992,
+      "step": 3478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7524084572976752,
+      "learning_rate": 9.057840472320192e-06,
+      "loss": 0.7813,
+      "step": 3479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.682113856892597,
+      "learning_rate": 9.057234809370266e-06,
+      "loss": 0.6953,
+      "step": 3480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8124516768548655,
+      "learning_rate": 9.056628972071926e-06,
+      "loss": 0.7097,
+      "step": 3481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7065764676428128,
+      "learning_rate": 9.056022960451204e-06,
+      "loss": 0.8735,
+      "step": 3482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7323782473398488,
+      "learning_rate": 9.055416774534142e-06,
+      "loss": 0.7062,
+      "step": 3483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8136918272881346,
+      "learning_rate": 9.054810414346789e-06,
+      "loss": 0.8353,
+      "step": 3484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0457259536904755,
+      "learning_rate": 9.054203879915202e-06,
+      "loss": 0.8121,
+      "step": 3485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.601229162789039,
+      "learning_rate": 9.053597171265447e-06,
+      "loss": 0.961,
+      "step": 3486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.251275773424158,
+      "learning_rate": 9.052990288423591e-06,
+      "loss": 0.7279,
+      "step": 3487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6255856058123195,
+      "learning_rate": 9.052383231415717e-06,
+      "loss": 0.831,
+      "step": 3488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6501828536463046,
+      "learning_rate": 9.05177600026791e-06,
+      "loss": 0.8296,
+      "step": 3489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8066409136871437,
+      "learning_rate": 9.051168595006264e-06,
+      "loss": 0.7568,
+      "step": 3490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7562798392490673,
+      "learning_rate": 9.050561015656882e-06,
+      "loss": 0.8629,
+      "step": 3491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.16877412113732,
+      "learning_rate": 9.049953262245872e-06,
+      "loss": 0.7463,
+      "step": 3492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.627454821111083,
+      "learning_rate": 9.04934533479935e-06,
+      "loss": 0.8191,
+      "step": 3493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5601436214426037,
+      "learning_rate": 9.048737233343442e-06,
+      "loss": 0.8789,
+      "step": 3494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.498598494768459,
+      "learning_rate": 9.048128957904277e-06,
+      "loss": 0.8263,
+      "step": 3495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9914549608491698,
+      "learning_rate": 9.047520508507994e-06,
+      "loss": 0.7479,
+      "step": 3496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3510861975782915,
+      "learning_rate": 9.046911885180743e-06,
+      "loss": 0.7328,
+      "step": 3497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.358919832659965,
+      "learning_rate": 9.046303087948674e-06,
+      "loss": 0.7335,
+      "step": 3498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5798160012192495,
+      "learning_rate": 9.045694116837948e-06,
+      "loss": 0.8075,
+      "step": 3499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.811657966213238,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.7989,
+      "step": 3500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5876673428651107,
+      "learning_rate": 9.044475653085218e-06,
+      "loss": 0.8336,
+      "step": 3501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2832737857232592,
+      "learning_rate": 9.04386616049557e-06,
+      "loss": 0.7267,
+      "step": 3502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8237369051301187,
+      "learning_rate": 9.043256494131987e-06,
+      "loss": 0.7584,
+      "step": 3503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9070568385121736,
+      "learning_rate": 9.042646654020667e-06,
+      "loss": 0.7822,
+      "step": 3504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6419592938683194,
+      "learning_rate": 9.04203664018782e-06,
+      "loss": 0.7767,
+      "step": 3505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6762042941958415,
+      "learning_rate": 9.041426452659654e-06,
+      "loss": 0.7899,
+      "step": 3506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6533559664681847,
+      "learning_rate": 9.040816091462393e-06,
+      "loss": 0.7597,
+      "step": 3507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8445030124485926,
+      "learning_rate": 9.040205556622266e-06,
+      "loss": 0.8434,
+      "step": 3508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8592119680300265,
+      "learning_rate": 9.039594848165507e-06,
+      "loss": 0.8889,
+      "step": 3509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0703551204612125,
+      "learning_rate": 9.038983966118359e-06,
+      "loss": 0.6378,
+      "step": 3510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0831784324661748,
+      "learning_rate": 9.038372910507079e-06,
+      "loss": 0.7412,
+      "step": 3511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8493164993594555,
+      "learning_rate": 9.03776168135792e-06,
+      "loss": 0.7625,
+      "step": 3512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.136710854343452,
+      "learning_rate": 9.037150278697147e-06,
+      "loss": 0.9171,
+      "step": 3513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9917710083639981,
+      "learning_rate": 9.036538702551037e-06,
+      "loss": 0.6895,
+      "step": 3514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9617974826867532,
+      "learning_rate": 9.035926952945868e-06,
+      "loss": 0.9223,
+      "step": 3515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6514077846456707,
+      "learning_rate": 9.035315029907928e-06,
+      "loss": 0.8336,
+      "step": 3516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7227157834261275,
+      "learning_rate": 9.034702933463516e-06,
+      "loss": 0.8452,
+      "step": 3517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.418058369817902,
+      "learning_rate": 9.034090663638933e-06,
+      "loss": 0.7646,
+      "step": 3518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1405293419966593,
+      "learning_rate": 9.03347822046049e-06,
+      "loss": 0.7797,
+      "step": 3519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7662538038769773,
+      "learning_rate": 9.032865603954504e-06,
+      "loss": 0.8963,
+      "step": 3520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.557818468157723,
+      "learning_rate": 9.032252814147302e-06,
+      "loss": 0.7245,
+      "step": 3521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7119346620954485,
+      "learning_rate": 9.031639851065217e-06,
+      "loss": 0.653,
+      "step": 3522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7669165853677637,
+      "learning_rate": 9.031026714734588e-06,
+      "loss": 0.8884,
+      "step": 3523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6619879621800866,
+      "learning_rate": 9.03041340518176e-06,
+      "loss": 0.7465,
+      "step": 3524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.084051031232644,
+      "learning_rate": 9.029799922433098e-06,
+      "loss": 0.6067,
+      "step": 3525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8255616705162585,
+      "learning_rate": 9.029186266514956e-06,
+      "loss": 0.7362,
+      "step": 3526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2130796544017768,
+      "learning_rate": 9.028572437453705e-06,
+      "loss": 0.6795,
+      "step": 3527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.11292534364054,
+      "learning_rate": 9.027958435275726e-06,
+      "loss": 0.8111,
+      "step": 3528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6067014034974811,
+      "learning_rate": 9.027344260007401e-06,
+      "loss": 0.7768,
+      "step": 3529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8454905610146064,
+      "learning_rate": 9.026729911675124e-06,
+      "loss": 0.7995,
+      "step": 3530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.018588919111351,
+      "learning_rate": 9.026115390305295e-06,
+      "loss": 0.9198,
+      "step": 3531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6636938457885388,
+      "learning_rate": 9.025500695924321e-06,
+      "loss": 0.8415,
+      "step": 3532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0701328730553041,
+      "learning_rate": 9.024885828558618e-06,
+      "loss": 0.6711,
+      "step": 3533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4427591163706144,
+      "learning_rate": 9.024270788234606e-06,
+      "loss": 0.7781,
+      "step": 3534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6274520335463813,
+      "learning_rate": 9.023655574978716e-06,
+      "loss": 0.6694,
+      "step": 3535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.936033433339851,
+      "learning_rate": 9.023040188817385e-06,
+      "loss": 0.7261,
+      "step": 3536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0413526758978842,
+      "learning_rate": 9.022424629777056e-06,
+      "loss": 0.5993,
+      "step": 3537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7059354137001954,
+      "learning_rate": 9.021808897884183e-06,
+      "loss": 0.7426,
+      "step": 3538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6362854422854267,
+      "learning_rate": 9.021192993165224e-06,
+      "loss": 0.7081,
+      "step": 3539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8769823830626022,
+      "learning_rate": 9.020576915646649e-06,
+      "loss": 0.9509,
+      "step": 3540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6915789348844712,
+      "learning_rate": 9.019960665354926e-06,
+      "loss": 0.7787,
+      "step": 3541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7657275678665096,
+      "learning_rate": 9.019344242316542e-06,
+      "loss": 0.7898,
+      "step": 3542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5885056324048648,
+      "learning_rate": 9.018727646557985e-06,
+      "loss": 0.7797,
+      "step": 3543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0873666494699175,
+      "learning_rate": 9.01811087810575e-06,
+      "loss": 0.6098,
+      "step": 3544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.731836126653387,
+      "learning_rate": 9.01749393698634e-06,
+      "loss": 0.8722,
+      "step": 3545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0227833045968886,
+      "learning_rate": 9.01687682322627e-06,
+      "loss": 0.7673,
+      "step": 3546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.648181228875407,
+      "learning_rate": 9.016259536852059e-06,
+      "loss": 0.7985,
+      "step": 3547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6464610452727102,
+      "learning_rate": 9.015642077890228e-06,
+      "loss": 0.7306,
+      "step": 3548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0894367900813795,
+      "learning_rate": 9.015024446367315e-06,
+      "loss": 0.8208,
+      "step": 3549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.612780049321006,
+      "learning_rate": 9.014406642309859e-06,
+      "loss": 0.7194,
+      "step": 3550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6468566984028596,
+      "learning_rate": 9.013788665744409e-06,
+      "loss": 0.7923,
+      "step": 3551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.879254703829632,
+      "learning_rate": 9.01317051669752e-06,
+      "loss": 0.8133,
+      "step": 3552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7047961346927485,
+      "learning_rate": 9.012552195195759e-06,
+      "loss": 0.8414,
+      "step": 3553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6175487856530415,
+      "learning_rate": 9.01193370126569e-06,
+      "loss": 0.6367,
+      "step": 3554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.825180719976429,
+      "learning_rate": 9.011315034933896e-06,
+      "loss": 0.6886,
+      "step": 3555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7716312105458087,
+      "learning_rate": 9.010696196226963e-06,
+      "loss": 0.6836,
+      "step": 3556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6671555889775864,
+      "learning_rate": 9.010077185171481e-06,
+      "loss": 0.8471,
+      "step": 3557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7550077281998449,
+      "learning_rate": 9.009458001794052e-06,
+      "loss": 0.7234,
+      "step": 3558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7332411916565682,
+      "learning_rate": 9.008838646121282e-06,
+      "loss": 0.8266,
+      "step": 3559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.572301436637751,
+      "learning_rate": 9.008219118179787e-06,
+      "loss": 0.7522,
+      "step": 3560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.627079095012505,
+      "learning_rate": 9.007599417996191e-06,
+      "loss": 0.8139,
+      "step": 3561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9317729623362585,
+      "learning_rate": 9.006979545597123e-06,
+      "loss": 0.7305,
+      "step": 3562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7508294757884713,
+      "learning_rate": 9.00635950100922e-06,
+      "loss": 0.7826,
+      "step": 3563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.541847745341615,
+      "learning_rate": 9.005739284259123e-06,
+      "loss": 0.7991,
+      "step": 3564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1490297323028125,
+      "learning_rate": 9.00511889537349e-06,
+      "loss": 0.7336,
+      "step": 3565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6651998444472864,
+      "learning_rate": 9.00449833437898e-06,
+      "loss": 0.8174,
+      "step": 3566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7829188985744633,
+      "learning_rate": 9.003877601302256e-06,
+      "loss": 0.7857,
+      "step": 3567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2195129767132618,
+      "learning_rate": 9.003256696169995e-06,
+      "loss": 0.6761,
+      "step": 3568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6526720051525183,
+      "learning_rate": 9.002635619008877e-06,
+      "loss": 0.779,
+      "step": 3569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1696079117180438,
+      "learning_rate": 9.002014369845592e-06,
+      "loss": 0.7081,
+      "step": 3570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0691297797083164,
+      "learning_rate": 9.00139294870684e-06,
+      "loss": 0.7889,
+      "step": 3571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.157980849339808,
+      "learning_rate": 9.000771355619318e-06,
+      "loss": 0.7406,
+      "step": 3572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6149423921714356,
+      "learning_rate": 9.000149590609741e-06,
+      "loss": 0.7625,
+      "step": 3573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7715364714810418,
+      "learning_rate": 8.999527653704829e-06,
+      "loss": 0.7647,
+      "step": 3574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6200197267193956,
+      "learning_rate": 8.998905544931304e-06,
+      "loss": 0.8866,
+      "step": 3575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.953466900449746,
+      "learning_rate": 8.998283264315903e-06,
+      "loss": 0.8154,
+      "step": 3576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7434233515790922,
+      "learning_rate": 8.997660811885367e-06,
+      "loss": 0.9197,
+      "step": 3577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.925932117011957,
+      "learning_rate": 8.99703818766644e-06,
+      "loss": 0.8867,
+      "step": 3578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7427458745057858,
+      "learning_rate": 8.996415391685882e-06,
+      "loss": 0.8668,
+      "step": 3579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8325667963200338,
+      "learning_rate": 8.995792423970454e-06,
+      "loss": 0.8564,
+      "step": 3580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7604897407552145,
+      "learning_rate": 8.995169284546926e-06,
+      "loss": 0.7622,
+      "step": 3581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7370048818094546,
+      "learning_rate": 8.994545973442076e-06,
+      "loss": 0.8247,
+      "step": 3582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6251178145424068,
+      "learning_rate": 8.993922490682687e-06,
+      "loss": 0.816,
+      "step": 3583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6670496391702676,
+      "learning_rate": 8.993298836295556e-06,
+      "loss": 0.8192,
+      "step": 3584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6129124081976243,
+      "learning_rate": 8.99267501030748e-06,
+      "loss": 0.7654,
+      "step": 3585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5838406213573823,
+      "learning_rate": 8.992051012745268e-06,
+      "loss": 0.7692,
+      "step": 3586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5217081383270938,
+      "learning_rate": 8.991426843635731e-06,
+      "loss": 0.6626,
+      "step": 3587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5369819084868483,
+      "learning_rate": 8.990802503005695e-06,
+      "loss": 0.7255,
+      "step": 3588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.86661740554068,
+      "learning_rate": 8.990177990881986e-06,
+      "loss": 0.7119,
+      "step": 3589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.878654149144602,
+      "learning_rate": 8.989553307291443e-06,
+      "loss": 0.7034,
+      "step": 3590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5198374883773986,
+      "learning_rate": 8.988928452260909e-06,
+      "loss": 0.7641,
+      "step": 3591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.651900736970262,
+      "learning_rate": 8.988303425817235e-06,
+      "loss": 0.8153,
+      "step": 3592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.70192867471851,
+      "learning_rate": 8.987678227987282e-06,
+      "loss": 0.7123,
+      "step": 3593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0767704637006752,
+      "learning_rate": 8.987052858797914e-06,
+      "loss": 0.6657,
+      "step": 3594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6039687083422067,
+      "learning_rate": 8.986427318276003e-06,
+      "loss": 0.8727,
+      "step": 3595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7476250416268755,
+      "learning_rate": 8.985801606448435e-06,
+      "loss": 0.8499,
+      "step": 3596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7571322205278603,
+      "learning_rate": 8.985175723342093e-06,
+      "loss": 0.817,
+      "step": 3597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.200933538202953,
+      "learning_rate": 8.984549668983875e-06,
+      "loss": 0.7959,
+      "step": 3598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8771026333971859,
+      "learning_rate": 8.983923443400682e-06,
+      "loss": 0.7397,
+      "step": 3599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5450514165068365,
+      "learning_rate": 8.983297046619427e-06,
+      "loss": 0.7856,
+      "step": 3600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.208040129061402,
+      "learning_rate": 8.982670478667029e-06,
+      "loss": 0.7345,
+      "step": 3601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7668743143050196,
+      "learning_rate": 8.98204373957041e-06,
+      "loss": 0.7296,
+      "step": 3602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.181626136460768,
+      "learning_rate": 8.981416829356502e-06,
+      "loss": 0.7365,
+      "step": 3603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.735323532701778,
+      "learning_rate": 8.980789748052245e-06,
+      "loss": 0.8406,
+      "step": 3604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7752077208812462,
+      "learning_rate": 8.980162495684587e-06,
+      "loss": 0.7676,
+      "step": 3605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6551708280831352,
+      "learning_rate": 8.979535072280483e-06,
+      "loss": 0.7192,
+      "step": 3606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9566601962461054,
+      "learning_rate": 8.978907477866894e-06,
+      "loss": 0.8291,
+      "step": 3607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9043436104904317,
+      "learning_rate": 8.978279712470789e-06,
+      "loss": 0.8323,
+      "step": 3608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6182890924645206,
+      "learning_rate": 8.977651776119145e-06,
+      "loss": 0.8818,
+      "step": 3609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1225881551475005,
+      "learning_rate": 8.977023668838945e-06,
+      "loss": 0.6928,
+      "step": 3610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7713963496974452,
+      "learning_rate": 8.976395390657181e-06,
+      "loss": 0.781,
+      "step": 3611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5159119088448127,
+      "learning_rate": 8.975766941600852e-06,
+      "loss": 0.8578,
+      "step": 3612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9495922002918076,
+      "learning_rate": 8.975138321696961e-06,
+      "loss": 0.8104,
+      "step": 3613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6589411841358412,
+      "learning_rate": 8.974509530972523e-06,
+      "loss": 0.8522,
+      "step": 3614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.846399465857466,
+      "learning_rate": 8.973880569454561e-06,
+      "loss": 0.8747,
+      "step": 3615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9421864391720898,
+      "learning_rate": 8.9732514371701e-06,
+      "loss": 0.7304,
+      "step": 3616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.53830028970098,
+      "learning_rate": 8.972622134146174e-06,
+      "loss": 0.7638,
+      "step": 3617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.249844911232969,
+      "learning_rate": 8.97199266040983e-06,
+      "loss": 0.8205,
+      "step": 3618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.524119838372441,
+      "learning_rate": 8.971363015988115e-06,
+      "loss": 0.7234,
+      "step": 3619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9467313069262795,
+      "learning_rate": 8.970733200908084e-06,
+      "loss": 0.6158,
+      "step": 3620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.653466274019399,
+      "learning_rate": 8.970103215196807e-06,
+      "loss": 0.8807,
+      "step": 3621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.635678685363191,
+      "learning_rate": 8.96947305888135e-06,
+      "loss": 0.7607,
+      "step": 3622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.124209338260535,
+      "learning_rate": 8.968842731988798e-06,
+      "loss": 0.8492,
+      "step": 3623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9896362452081013,
+      "learning_rate": 8.968212234546235e-06,
+      "loss": 0.8451,
+      "step": 3624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4846593201426512,
+      "learning_rate": 8.967581566580752e-06,
+      "loss": 0.906,
+      "step": 3625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7244103778939683,
+      "learning_rate": 8.966950728119453e-06,
+      "loss": 0.8656,
+      "step": 3626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7069218492870915,
+      "learning_rate": 8.966319719189447e-06,
+      "loss": 0.8424,
+      "step": 3627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.154043948692091,
+      "learning_rate": 8.96568853981785e-06,
+      "loss": 0.8611,
+      "step": 3628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7027768279394369,
+      "learning_rate": 8.965057190031785e-06,
+      "loss": 0.8103,
+      "step": 3629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0700527511492117,
+      "learning_rate": 8.96442566985838e-06,
+      "loss": 0.8954,
+      "step": 3630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6798382399143492,
+      "learning_rate": 8.963793979324775e-06,
+      "loss": 0.7811,
+      "step": 3631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6051810483874096,
+      "learning_rate": 8.963162118458115e-06,
+      "loss": 0.7434,
+      "step": 3632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3150222700180885,
+      "learning_rate": 8.962530087285552e-06,
+      "loss": 0.6912,
+      "step": 3633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.797344784664565,
+      "learning_rate": 8.961897885834247e-06,
+      "loss": 0.6969,
+      "step": 3634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5688116401300662,
+      "learning_rate": 8.961265514131368e-06,
+      "loss": 0.765,
+      "step": 3635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.117261938918543,
+      "learning_rate": 8.960632972204084e-06,
+      "loss": 0.8049,
+      "step": 3636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2841168109732086,
+      "learning_rate": 8.960000260079582e-06,
+      "loss": 0.6614,
+      "step": 3637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1397307578486968,
+      "learning_rate": 8.95936737778505e-06,
+      "loss": 0.6845,
+      "step": 3638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7990419551461727,
+      "learning_rate": 8.958734325347684e-06,
+      "loss": 0.7941,
+      "step": 3639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8542314579449923,
+      "learning_rate": 8.958101102794686e-06,
+      "loss": 0.8547,
+      "step": 3640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.157752256095621,
+      "learning_rate": 8.95746771015327e-06,
+      "loss": 0.6566,
+      "step": 3641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6215191379962621,
+      "learning_rate": 8.95683414745065e-06,
+      "loss": 0.6786,
+      "step": 3642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6849438105147616,
+      "learning_rate": 8.956200414714057e-06,
+      "loss": 0.7554,
+      "step": 3643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0978658123313292,
+      "learning_rate": 8.955566511970721e-06,
+      "loss": 0.6273,
+      "step": 3644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1134116133842595,
+      "learning_rate": 8.954932439247882e-06,
+      "loss": 0.6802,
+      "step": 3645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7035964063123958,
+      "learning_rate": 8.954298196572789e-06,
+      "loss": 0.7637,
+      "step": 3646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7430217177380667,
+      "learning_rate": 8.953663783972692e-06,
+      "loss": 0.6962,
+      "step": 3647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7852938062559462,
+      "learning_rate": 8.95302920147486e-06,
+      "loss": 0.7781,
+      "step": 3648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4048321971446134,
+      "learning_rate": 8.95239444910656e-06,
+      "loss": 0.6498,
+      "step": 3649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8339292710683892,
+      "learning_rate": 8.951759526895067e-06,
+      "loss": 0.7772,
+      "step": 3650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6074358799578925,
+      "learning_rate": 8.951124434867666e-06,
+      "loss": 0.7798,
+      "step": 3651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.841225479913545,
+      "learning_rate": 8.950489173051648e-06,
+      "loss": 0.8001,
+      "step": 3652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7839405423837578,
+      "learning_rate": 8.949853741474312e-06,
+      "loss": 0.8203,
+      "step": 3653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6857028742626592,
+      "learning_rate": 8.949218140162965e-06,
+      "loss": 0.8191,
+      "step": 3654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4785796176620685,
+      "learning_rate": 8.948582369144919e-06,
+      "loss": 0.8397,
+      "step": 3655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1666422195072732,
+      "learning_rate": 8.947946428447495e-06,
+      "loss": 0.6285,
+      "step": 3656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6775625171582143,
+      "learning_rate": 8.94731031809802e-06,
+      "loss": 0.8221,
+      "step": 3657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0380484179943497,
+      "learning_rate": 8.946674038123829e-06,
+      "loss": 0.7589,
+      "step": 3658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1068133553350161,
+      "learning_rate": 8.946037588552266e-06,
+      "loss": 0.6727,
+      "step": 3659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.570313600216091,
+      "learning_rate": 8.945400969410678e-06,
+      "loss": 0.7152,
+      "step": 3660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.193030118426774,
+      "learning_rate": 8.944764180726423e-06,
+      "loss": 0.9174,
+      "step": 3661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0756567576224512,
+      "learning_rate": 8.944127222526867e-06,
+      "loss": 0.6747,
+      "step": 3662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1835121352322133,
+      "learning_rate": 8.94349009483938e-06,
+      "loss": 0.6891,
+      "step": 3663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5881692816246158,
+      "learning_rate": 8.94285279769134e-06,
+      "loss": 0.6898,
+      "step": 3664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5170125289971736,
+      "learning_rate": 8.942215331110136e-06,
+      "loss": 0.7491,
+      "step": 3665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9389562504547542,
+      "learning_rate": 8.941577695123158e-06,
+      "loss": 0.7989,
+      "step": 3666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8780482986153855,
+      "learning_rate": 8.940939889757809e-06,
+      "loss": 0.8442,
+      "step": 3667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.041434702071168,
+      "learning_rate": 8.940301915041496e-06,
+      "loss": 0.7893,
+      "step": 3668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5891455782365143,
+      "learning_rate": 8.939663771001632e-06,
+      "loss": 0.8048,
+      "step": 3669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.337191251551206,
+      "learning_rate": 8.939025457665644e-06,
+      "loss": 0.5469,
+      "step": 3670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8209503567487777,
+      "learning_rate": 8.93838697506096e-06,
+      "loss": 0.8781,
+      "step": 3671
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5959740949509333,
+      "learning_rate": 8.937748323215013e-06,
+      "loss": 0.7801,
+      "step": 3672
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6029986949715511,
+      "learning_rate": 8.937109502155254e-06,
+      "loss": 0.7235,
+      "step": 3673
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4597266607955905,
+      "learning_rate": 8.93647051190913e-06,
+      "loss": 0.8285,
+      "step": 3674
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6988182433548353,
+      "learning_rate": 8.935831352504103e-06,
+      "loss": 0.7287,
+      "step": 3675
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7501296366281767,
+      "learning_rate": 8.935192023967634e-06,
+      "loss": 0.7013,
+      "step": 3676
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.228135286720274,
+      "learning_rate": 8.934552526327201e-06,
+      "loss": 0.8224,
+      "step": 3677
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0915384313452245,
+      "learning_rate": 8.933912859610284e-06,
+      "loss": 0.646,
+      "step": 3678
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7268313857858815,
+      "learning_rate": 8.93327302384437e-06,
+      "loss": 0.738,
+      "step": 3679
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5919613981382408,
+      "learning_rate": 8.932633019056954e-06,
+      "loss": 0.7074,
+      "step": 3680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7099953923728726,
+      "learning_rate": 8.931992845275539e-06,
+      "loss": 0.7874,
+      "step": 3681
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6981233798859068,
+      "learning_rate": 8.931352502527633e-06,
+      "loss": 0.8967,
+      "step": 3682
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1382581178006923,
+      "learning_rate": 8.930711990840755e-06,
+      "loss": 0.7965,
+      "step": 3683
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8068158503902285,
+      "learning_rate": 8.930071310242429e-06,
+      "loss": 0.7768,
+      "step": 3684
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6544771871693078,
+      "learning_rate": 8.929430460760187e-06,
+      "loss": 0.7183,
+      "step": 3685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0430024920673273,
+      "learning_rate": 8.928789442421564e-06,
+      "loss": 0.6458,
+      "step": 3686
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4945232564628053,
+      "learning_rate": 8.928148255254112e-06,
+      "loss": 0.7853,
+      "step": 3687
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6398171055915174,
+      "learning_rate": 8.927506899285378e-06,
+      "loss": 0.7052,
+      "step": 3688
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.804206739195225,
+      "learning_rate": 8.926865374542928e-06,
+      "loss": 0.8211,
+      "step": 3689
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8854666308078452,
+      "learning_rate": 8.926223681054326e-06,
+      "loss": 0.7205,
+      "step": 3690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7558839793204155,
+      "learning_rate": 8.92558181884715e-06,
+      "loss": 0.7489,
+      "step": 3691
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2504772747759612,
+      "learning_rate": 8.924939787948977e-06,
+      "loss": 0.6566,
+      "step": 3692
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9996936189906787,
+      "learning_rate": 8.924297588387404e-06,
+      "loss": 0.8591,
+      "step": 3693
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9473092801122918,
+      "learning_rate": 8.92365522019002e-06,
+      "loss": 0.7452,
+      "step": 3694
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8422915285421488,
+      "learning_rate": 8.923012683384436e-06,
+      "loss": 0.8581,
+      "step": 3695
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6842197947734303,
+      "learning_rate": 8.922369977998257e-06,
+      "loss": 0.9025,
+      "step": 3696
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6625195622694835,
+      "learning_rate": 8.921727104059107e-06,
+      "loss": 0.876,
+      "step": 3697
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.239446142574799,
+      "learning_rate": 8.921084061594608e-06,
+      "loss": 0.694,
+      "step": 3698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7571523324402498,
+      "learning_rate": 8.920440850632395e-06,
+      "loss": 0.7314,
+      "step": 3699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7148606145650473,
+      "learning_rate": 8.919797471200106e-06,
+      "loss": 0.8766,
+      "step": 3700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0801651018681038,
+      "learning_rate": 8.91915392332539e-06,
+      "loss": 0.746,
+      "step": 3701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.971654161906901,
+      "learning_rate": 8.918510207035901e-06,
+      "loss": 0.8225,
+      "step": 3702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5695436092436779,
+      "learning_rate": 8.917866322359303e-06,
+      "loss": 0.8122,
+      "step": 3703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0020430833673593,
+      "learning_rate": 8.917222269323263e-06,
+      "loss": 0.6881,
+      "step": 3704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3508513036468859,
+      "learning_rate": 8.916578047955458e-06,
+      "loss": 0.6449,
+      "step": 3705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7576984645791394,
+      "learning_rate": 8.915933658283572e-06,
+      "loss": 0.8447,
+      "step": 3706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.768049215902588,
+      "learning_rate": 8.915289100335295e-06,
+      "loss": 0.7189,
+      "step": 3707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6045754782446482,
+      "learning_rate": 8.914644374138325e-06,
+      "loss": 0.6976,
+      "step": 3708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.365357284712674,
+      "learning_rate": 8.91399947972037e-06,
+      "loss": 0.6781,
+      "step": 3709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.661552369315311,
+      "learning_rate": 8.913354417109136e-06,
+      "loss": 0.7262,
+      "step": 3710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7101212189366413,
+      "learning_rate": 8.91270918633235e-06,
+      "loss": 0.7202,
+      "step": 3711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0915518058903833,
+      "learning_rate": 8.912063787417738e-06,
+      "loss": 0.8487,
+      "step": 3712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6856797929201894,
+      "learning_rate": 8.911418220393031e-06,
+      "loss": 0.8931,
+      "step": 3713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8802186159613483,
+      "learning_rate": 8.91077248528597e-06,
+      "loss": 0.7665,
+      "step": 3714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6898161764649027,
+      "learning_rate": 8.910126582124307e-06,
+      "loss": 0.7524,
+      "step": 3715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5918655896013438,
+      "learning_rate": 8.909480510935797e-06,
+      "loss": 0.8448,
+      "step": 3716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8149639600150176,
+      "learning_rate": 8.908834271748202e-06,
+      "loss": 0.6991,
+      "step": 3717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8802320074044045,
+      "learning_rate": 8.908187864589292e-06,
+      "loss": 0.7132,
+      "step": 3718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5369553267527614,
+      "learning_rate": 8.907541289486847e-06,
+      "loss": 0.7819,
+      "step": 3719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5351789999006613,
+      "learning_rate": 8.90689454646865e-06,
+      "loss": 0.6961,
+      "step": 3720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.514784514677639,
+      "learning_rate": 8.906247635562493e-06,
+      "loss": 0.7003,
+      "step": 3721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.58772594197105,
+      "learning_rate": 8.905600556796176e-06,
+      "loss": 0.7346,
+      "step": 3722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.916468263614762,
+      "learning_rate": 8.904953310197504e-06,
+      "loss": 0.802,
+      "step": 3723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8737590335133523,
+      "learning_rate": 8.904305895794292e-06,
+      "loss": 0.7924,
+      "step": 3724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9117783257197787,
+      "learning_rate": 8.903658313614358e-06,
+      "loss": 0.808,
+      "step": 3725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.827386861155441,
+      "learning_rate": 8.903010563685534e-06,
+      "loss": 0.8987,
+      "step": 3726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6660464956308534,
+      "learning_rate": 8.902362646035655e-06,
+      "loss": 0.8341,
+      "step": 3727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.845120111793418,
+      "learning_rate": 8.90171456069256e-06,
+      "loss": 0.7642,
+      "step": 3728
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6729513406981036,
+      "learning_rate": 8.901066307684102e-06,
+      "loss": 0.7922,
+      "step": 3729
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1614213331831766,
+      "learning_rate": 8.900417887038136e-06,
+      "loss": 0.7004,
+      "step": 3730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7802343055370753,
+      "learning_rate": 8.899769298782528e-06,
+      "loss": 0.7963,
+      "step": 3731
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0543833677588323,
+      "learning_rate": 8.899120542945147e-06,
+      "loss": 0.8686,
+      "step": 3732
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5666813251263172,
+      "learning_rate": 8.898471619553872e-06,
+      "loss": 0.7385,
+      "step": 3733
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7555710240154192,
+      "learning_rate": 8.89782252863659e-06,
+      "loss": 0.7662,
+      "step": 3734
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7552873258875694,
+      "learning_rate": 8.89717327022119e-06,
+      "loss": 0.6926,
+      "step": 3735
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.796097943480221,
+      "learning_rate": 8.89652384433558e-06,
+      "loss": 0.7769,
+      "step": 3736
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.711308742998957,
+      "learning_rate": 8.89587425100766e-06,
+      "loss": 0.8144,
+      "step": 3737
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2347087705748108,
+      "learning_rate": 8.895224490265346e-06,
+      "loss": 0.692,
+      "step": 3738
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7661775094274135,
+      "learning_rate": 8.894574562136561e-06,
+      "loss": 0.8211,
+      "step": 3739
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6242406105800822,
+      "learning_rate": 8.893924466649235e-06,
+      "loss": 0.7,
+      "step": 3740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0644928261827067,
+      "learning_rate": 8.893274203831302e-06,
+      "loss": 0.8219,
+      "step": 3741
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9547297996859359,
+      "learning_rate": 8.892623773710704e-06,
+      "loss": 0.749,
+      "step": 3742
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8465776304663053,
+      "learning_rate": 8.891973176315393e-06,
+      "loss": 0.6242,
+      "step": 3743
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6872016824607496,
+      "learning_rate": 8.89132241167333e-06,
+      "loss": 0.8035,
+      "step": 3744
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5673329579112436,
+      "learning_rate": 8.890671479812472e-06,
+      "loss": 0.7679,
+      "step": 3745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8101154424776527,
+      "learning_rate": 8.890020380760798e-06,
+      "loss": 0.7047,
+      "step": 3746
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5445841417441715,
+      "learning_rate": 8.889369114546285e-06,
+      "loss": 0.7097,
+      "step": 3747
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1722139007253896,
+      "learning_rate": 8.888717681196918e-06,
+      "loss": 0.806,
+      "step": 3748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5692498881345924,
+      "learning_rate": 8.888066080740692e-06,
+      "loss": 0.7876,
+      "step": 3749
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8378406383851569,
+      "learning_rate": 8.887414313205608e-06,
+      "loss": 0.8582,
+      "step": 3750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9750757173473386,
+      "learning_rate": 8.886762378619674e-06,
+      "loss": 0.7622,
+      "step": 3751
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4689677926374487,
+      "learning_rate": 8.886110277010902e-06,
+      "loss": 0.8185,
+      "step": 3752
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4328954220790333,
+      "learning_rate": 8.885458008407317e-06,
+      "loss": 0.7217,
+      "step": 3753
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9347628570070303,
+      "learning_rate": 8.88480557283695e-06,
+      "loss": 0.7591,
+      "step": 3754
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.874050558084816,
+      "learning_rate": 8.884152970327833e-06,
+      "loss": 0.6548,
+      "step": 3755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1088007202103245,
+      "learning_rate": 8.883500200908013e-06,
+      "loss": 0.7336,
+      "step": 3756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6522401187954592,
+      "learning_rate": 8.882847264605542e-06,
+      "loss": 0.7309,
+      "step": 3757
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.100806693639464,
+      "learning_rate": 8.882194161448476e-06,
+      "loss": 0.7363,
+      "step": 3758
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6643708814500815,
+      "learning_rate": 8.88154089146488e-06,
+      "loss": 0.6488,
+      "step": 3759
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.189413137073641,
+      "learning_rate": 8.880887454682827e-06,
+      "loss": 0.8701,
+      "step": 3760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9374307944809002,
+      "learning_rate": 8.880233851130397e-06,
+      "loss": 0.819,
+      "step": 3761
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.643210787585964,
+      "learning_rate": 8.879580080835677e-06,
+      "loss": 0.7869,
+      "step": 3762
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.707761875180826,
+      "learning_rate": 8.87892614382676e-06,
+      "loss": 0.8407,
+      "step": 3763
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3047760147581022,
+      "learning_rate": 8.878272040131748e-06,
+      "loss": 0.7836,
+      "step": 3764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.806633723705242,
+      "learning_rate": 8.877617769778751e-06,
+      "loss": 0.7173,
+      "step": 3765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9394955041885782,
+      "learning_rate": 8.876963332795881e-06,
+      "loss": 0.759,
+      "step": 3766
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8032794825718723,
+      "learning_rate": 8.876308729211261e-06,
+      "loss": 0.9778,
+      "step": 3767
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8012471477906704,
+      "learning_rate": 8.875653959053023e-06,
+      "loss": 0.7326,
+      "step": 3768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7748719914549804,
+      "learning_rate": 8.874999022349303e-06,
+      "loss": 0.7932,
+      "step": 3769
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8792806890450886,
+      "learning_rate": 8.874343919128245e-06,
+      "loss": 0.7719,
+      "step": 3770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9827888371228426,
+      "learning_rate": 8.873688649417999e-06,
+      "loss": 0.7662,
+      "step": 3771
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5885940325099743,
+      "learning_rate": 8.873033213246726e-06,
+      "loss": 0.7886,
+      "step": 3772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6527777745807897,
+      "learning_rate": 8.87237761064259e-06,
+      "loss": 0.6899,
+      "step": 3773
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7511386573284984,
+      "learning_rate": 8.871721841633762e-06,
+      "loss": 0.9113,
+      "step": 3774
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.70461213711174,
+      "learning_rate": 8.871065906248426e-06,
+      "loss": 0.813,
+      "step": 3775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.658922183737342,
+      "learning_rate": 8.870409804514765e-06,
+      "loss": 0.7947,
+      "step": 3776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.67582102617471,
+      "learning_rate": 8.869753536460976e-06,
+      "loss": 0.8759,
+      "step": 3777
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6184547967283225,
+      "learning_rate": 8.86909710211526e-06,
+      "loss": 0.7252,
+      "step": 3778
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.706505950377408,
+      "learning_rate": 8.868440501505822e-06,
+      "loss": 0.8509,
+      "step": 3779
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6367170304461278,
+      "learning_rate": 8.867783734660883e-06,
+      "loss": 0.7985,
+      "step": 3780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1799381187407687,
+      "learning_rate": 8.86712680160866e-06,
+      "loss": 0.744,
+      "step": 3781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6716784454314044,
+      "learning_rate": 8.866469702377386e-06,
+      "loss": 0.789,
+      "step": 3782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9535002239775288,
+      "learning_rate": 8.8658124369953e-06,
+      "loss": 0.9407,
+      "step": 3783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1083583862366075,
+      "learning_rate": 8.865155005490643e-06,
+      "loss": 0.7178,
+      "step": 3784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6560610266849245,
+      "learning_rate": 8.864497407891668e-06,
+      "loss": 0.939,
+      "step": 3785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.963464093879562,
+      "learning_rate": 8.86383964422663e-06,
+      "loss": 0.8214,
+      "step": 3786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8511506941350875,
+      "learning_rate": 8.8631817145238e-06,
+      "loss": 0.714,
+      "step": 3787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6902599239430138,
+      "learning_rate": 8.862523618811448e-06,
+      "loss": 0.8169,
+      "step": 3788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.78966460079282,
+      "learning_rate": 8.861865357117852e-06,
+      "loss": 0.7853,
+      "step": 3789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.109752735547795,
+      "learning_rate": 8.861206929471302e-06,
+      "loss": 0.5909,
+      "step": 3790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7716330615447706,
+      "learning_rate": 8.86054833590009e-06,
+      "loss": 0.7672,
+      "step": 3791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7779120509843584,
+      "learning_rate": 8.859889576432519e-06,
+      "loss": 0.687,
+      "step": 3792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9871582676439319,
+      "learning_rate": 8.859230651096896e-06,
+      "loss": 0.8153,
+      "step": 3793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0184963677567,
+      "learning_rate": 8.858571559921539e-06,
+      "loss": 0.7449,
+      "step": 3794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8439664833573837,
+      "learning_rate": 8.857912302934766e-06,
+      "loss": 0.8793,
+      "step": 3795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.045039740175378,
+      "learning_rate": 8.857252880164908e-06,
+      "loss": 0.7017,
+      "step": 3796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6990030063909216,
+      "learning_rate": 8.856593291640305e-06,
+      "loss": 0.8363,
+      "step": 3797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9750692926142768,
+      "learning_rate": 8.8559335373893e-06,
+      "loss": 0.7459,
+      "step": 3798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1878802408312494,
+      "learning_rate": 8.855273617440243e-06,
+      "loss": 0.6336,
+      "step": 3799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.759533529223079,
+      "learning_rate": 8.854613531821491e-06,
+      "loss": 0.677,
+      "step": 3800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6711076904586901,
+      "learning_rate": 8.853953280561412e-06,
+      "loss": 0.8093,
+      "step": 3801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2577363053428878,
+      "learning_rate": 8.853292863688378e-06,
+      "loss": 0.7171,
+      "step": 3802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.51544366961259,
+      "learning_rate": 8.852632281230766e-06,
+      "loss": 0.801,
+      "step": 3803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7673824242015466,
+      "learning_rate": 8.851971533216968e-06,
+      "loss": 0.7998,
+      "step": 3804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.11757368057683,
+      "learning_rate": 8.85131061967537e-06,
+      "loss": 0.7842,
+      "step": 3805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9146236834604324,
+      "learning_rate": 8.85064954063438e-06,
+      "loss": 0.8654,
+      "step": 3806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9252066064004314,
+      "learning_rate": 8.849988296122402e-06,
+      "loss": 0.8602,
+      "step": 3807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7023192086655645,
+      "learning_rate": 8.849326886167854e-06,
+      "loss": 0.7523,
+      "step": 3808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8694513188619724,
+      "learning_rate": 8.848665310799156e-06,
+      "loss": 0.7695,
+      "step": 3809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8063954594223517,
+      "learning_rate": 8.84800357004474e-06,
+      "loss": 0.8333,
+      "step": 3810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.936853334527978,
+      "learning_rate": 8.847341663933038e-06,
+      "loss": 0.5924,
+      "step": 3811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.773264856241997,
+      "learning_rate": 8.846679592492498e-06,
+      "loss": 0.643,
+      "step": 3812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0241590934168319,
+      "learning_rate": 8.846017355751567e-06,
+      "loss": 0.6189,
+      "step": 3813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6352780120553887,
+      "learning_rate": 8.845354953738706e-06,
+      "loss": 0.7287,
+      "step": 3814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5748531003426782,
+      "learning_rate": 8.844692386482379e-06,
+      "loss": 0.7787,
+      "step": 3815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8055336235656965,
+      "learning_rate": 8.844029654011055e-06,
+      "loss": 0.8425,
+      "step": 3816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6288561446216088,
+      "learning_rate": 8.84336675635322e-06,
+      "loss": 0.8904,
+      "step": 3817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7529076893664137,
+      "learning_rate": 8.842703693537352e-06,
+      "loss": 0.7937,
+      "step": 3818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7077996773284547,
+      "learning_rate": 8.84204046559195e-06,
+      "loss": 0.7832,
+      "step": 3819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6598396551674681,
+      "learning_rate": 8.841377072545513e-06,
+      "loss": 0.7604,
+      "step": 3820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.115554544288809,
+      "learning_rate": 8.840713514426545e-06,
+      "loss": 0.7373,
+      "step": 3821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.60999561007971,
+      "learning_rate": 8.840049791263567e-06,
+      "loss": 0.7706,
+      "step": 3822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5548847506743457,
+      "learning_rate": 8.839385903085096e-06,
+      "loss": 0.813,
+      "step": 3823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8680225390448637,
+      "learning_rate": 8.83872184991966e-06,
+      "loss": 0.8192,
+      "step": 3824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5311378884193467,
+      "learning_rate": 8.8380576317958e-06,
+      "loss": 0.8076,
+      "step": 3825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8135149010827087,
+      "learning_rate": 8.837393248742054e-06,
+      "loss": 0.8059,
+      "step": 3826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7437848864705618,
+      "learning_rate": 8.836728700786974e-06,
+      "loss": 0.7995,
+      "step": 3827
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7614410062491712,
+      "learning_rate": 8.836063987959116e-06,
+      "loss": 0.77,
+      "step": 3828
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.187438708587635,
+      "learning_rate": 8.835399110287046e-06,
+      "loss": 0.852,
+      "step": 3829
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6441807780360105,
+      "learning_rate": 8.834734067799336e-06,
+      "loss": 0.6983,
+      "step": 3830
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0488416563105267,
+      "learning_rate": 8.83406886052456e-06,
+      "loss": 0.8385,
+      "step": 3831
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5900769637236147,
+      "learning_rate": 8.833403488491308e-06,
+      "loss": 0.719,
+      "step": 3832
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1453795388024488,
+      "learning_rate": 8.83273795172817e-06,
+      "loss": 0.6733,
+      "step": 3833
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8368505632779701,
+      "learning_rate": 8.832072250263746e-06,
+      "loss": 0.7976,
+      "step": 3834
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6677387970541073,
+      "learning_rate": 8.831406384126643e-06,
+      "loss": 0.8035,
+      "step": 3835
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7019380505894988,
+      "learning_rate": 8.830740353345475e-06,
+      "loss": 0.9051,
+      "step": 3836
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5989172887886567,
+      "learning_rate": 8.830074157948862e-06,
+      "loss": 0.8284,
+      "step": 3837
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8372328334138825,
+      "learning_rate": 8.829407797965433e-06,
+      "loss": 0.6989,
+      "step": 3838
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5869590691970574,
+      "learning_rate": 8.82874127342382e-06,
+      "loss": 0.8905,
+      "step": 3839
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2930108858226643,
+      "learning_rate": 8.82807458435267e-06,
+      "loss": 0.8164,
+      "step": 3840
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6573190858423446,
+      "learning_rate": 8.82740773078063e-06,
+      "loss": 0.7354,
+      "step": 3841
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.664344226296595,
+      "learning_rate": 8.826740712736355e-06,
+      "loss": 0.8545,
+      "step": 3842
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7158159367458974,
+      "learning_rate": 8.826073530248508e-06,
+      "loss": 0.7515,
+      "step": 3843
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8589506164080398,
+      "learning_rate": 8.82540618334576e-06,
+      "loss": 0.8566,
+      "step": 3844
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6981336812546408,
+      "learning_rate": 8.82473867205679e-06,
+      "loss": 0.7071,
+      "step": 3845
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1116203252678165,
+      "learning_rate": 8.824070996410281e-06,
+      "loss": 0.7793,
+      "step": 3846
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.717926497439608,
+      "learning_rate": 8.823403156434924e-06,
+      "loss": 0.741,
+      "step": 3847
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9676332819582887,
+      "learning_rate": 8.822735152159418e-06,
+      "loss": 0.6703,
+      "step": 3848
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7385734406810722,
+      "learning_rate": 8.82206698361247e-06,
+      "loss": 0.8739,
+      "step": 3849
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.267285563957948,
+      "learning_rate": 8.82139865082279e-06,
+      "loss": 0.6118,
+      "step": 3850
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7801392036191235,
+      "learning_rate": 8.8207301538191e-06,
+      "loss": 0.8859,
+      "step": 3851
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8271852744640684,
+      "learning_rate": 8.820061492630127e-06,
+      "loss": 0.8663,
+      "step": 3852
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5932991141605435,
+      "learning_rate": 8.819392667284603e-06,
+      "loss": 0.8645,
+      "step": 3853
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7631384558315732,
+      "learning_rate": 8.818723677811269e-06,
+      "loss": 0.7959,
+      "step": 3854
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6455837180133392,
+      "learning_rate": 8.818054524238874e-06,
+      "loss": 0.7203,
+      "step": 3855
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0586539037991183,
+      "learning_rate": 8.817385206596174e-06,
+      "loss": 0.6685,
+      "step": 3856
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.678405022201185,
+      "learning_rate": 8.81671572491193e-06,
+      "loss": 0.7743,
+      "step": 3857
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.272564164604948,
+      "learning_rate": 8.81604607921491e-06,
+      "loss": 0.7405,
+      "step": 3858
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8504516924345926,
+      "learning_rate": 8.815376269533893e-06,
+      "loss": 0.6842,
+      "step": 3859
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3134196004163337,
+      "learning_rate": 8.814706295897659e-06,
+      "loss": 0.8532,
+      "step": 3860
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4890831533782847,
+      "learning_rate": 8.814036158335e-06,
+      "loss": 0.7419,
+      "step": 3861
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6569617691884497,
+      "learning_rate": 8.813365856874711e-06,
+      "loss": 0.7546,
+      "step": 3862
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5136469613953687,
+      "learning_rate": 8.8126953915456e-06,
+      "loss": 0.7497,
+      "step": 3863
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2420808462875055,
+      "learning_rate": 8.812024762376477e-06,
+      "loss": 0.6285,
+      "step": 3864
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.61897795678621,
+      "learning_rate": 8.81135396939616e-06,
+      "loss": 0.7018,
+      "step": 3865
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8041540085052188,
+      "learning_rate": 8.810683012633475e-06,
+      "loss": 0.8378,
+      "step": 3866
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8313279440992083,
+      "learning_rate": 8.810011892117254e-06,
+      "loss": 0.7705,
+      "step": 3867
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5370393308008765,
+      "learning_rate": 8.809340607876335e-06,
+      "loss": 0.7316,
+      "step": 3868
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7222364557344647,
+      "learning_rate": 8.808669159939568e-06,
+      "loss": 0.7627,
+      "step": 3869
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8599049044147584,
+      "learning_rate": 8.807997548335806e-06,
+      "loss": 0.7474,
+      "step": 3870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7702828394901626,
+      "learning_rate": 8.807325773093904e-06,
+      "loss": 0.7865,
+      "step": 3871
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.638204355350188,
+      "learning_rate": 8.806653834242738e-06,
+      "loss": 0.6447,
+      "step": 3872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.700810982639837,
+      "learning_rate": 8.805981731811177e-06,
+      "loss": 0.7709,
+      "step": 3873
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.013727939542002,
+      "learning_rate": 8.805309465828105e-06,
+      "loss": 0.8099,
+      "step": 3874
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.016463935997889,
+      "learning_rate": 8.804637036322411e-06,
+      "loss": 0.7868,
+      "step": 3875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6960936983137478,
+      "learning_rate": 8.80396444332299e-06,
+      "loss": 0.7781,
+      "step": 3876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6406622622086238,
+      "learning_rate": 8.803291686858745e-06,
+      "loss": 0.7633,
+      "step": 3877
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.256943532450742,
+      "learning_rate": 8.802618766958586e-06,
+      "loss": 0.8604,
+      "step": 3878
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8332010163248547,
+      "learning_rate": 8.80194568365143e-06,
+      "loss": 0.6807,
+      "step": 3879
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8217402670539056,
+      "learning_rate": 8.8012724369662e-06,
+      "loss": 0.7957,
+      "step": 3880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6681233825543351,
+      "learning_rate": 8.800599026931827e-06,
+      "loss": 0.8006,
+      "step": 3881
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6511548268372431,
+      "learning_rate": 8.79992545357725e-06,
+      "loss": 0.8393,
+      "step": 3882
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.642212872196117,
+      "learning_rate": 8.799251716931412e-06,
+      "loss": 0.8086,
+      "step": 3883
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7912870818200528,
+      "learning_rate": 8.798577817023269e-06,
+      "loss": 0.836,
+      "step": 3884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2338772545262455,
+      "learning_rate": 8.797903753881775e-06,
+      "loss": 0.7535,
+      "step": 3885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8088438221863645,
+      "learning_rate": 8.7972295275359e-06,
+      "loss": 0.7682,
+      "step": 3886
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.628772888489203,
+      "learning_rate": 8.796555138014613e-06,
+      "loss": 0.7484,
+      "step": 3887
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.623908449827832,
+      "learning_rate": 8.795880585346899e-06,
+      "loss": 0.8061,
+      "step": 3888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6818302475587725,
+      "learning_rate": 8.795205869561742e-06,
+      "loss": 0.7718,
+      "step": 3889
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7488909075989327,
+      "learning_rate": 8.794530990688135e-06,
+      "loss": 0.699,
+      "step": 3890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.922596871044926,
+      "learning_rate": 8.793855948755081e-06,
+      "loss": 0.7325,
+      "step": 3891
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.679878672503642,
+      "learning_rate": 8.793180743791587e-06,
+      "loss": 0.7644,
+      "step": 3892
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.793383860686934,
+      "learning_rate": 8.79250537582667e-06,
+      "loss": 0.7338,
+      "step": 3893
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7538690914680661,
+      "learning_rate": 8.79182984488935e-06,
+      "loss": 0.7347,
+      "step": 3894
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6028416171251259,
+      "learning_rate": 8.791154151008657e-06,
+      "loss": 0.8193,
+      "step": 3895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9150580794511909,
+      "learning_rate": 8.790478294213626e-06,
+      "loss": 0.9006,
+      "step": 3896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5980159101460103,
+      "learning_rate": 8.789802274533301e-06,
+      "loss": 0.6501,
+      "step": 3897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6189283991950945,
+      "learning_rate": 8.789126091996731e-06,
+      "loss": 0.7832,
+      "step": 3898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.93513848720467,
+      "learning_rate": 8.788449746632976e-06,
+      "loss": 0.9467,
+      "step": 3899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8317742673553845,
+      "learning_rate": 8.787773238471098e-06,
+      "loss": 0.8063,
+      "step": 3900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.830230278011175,
+      "learning_rate": 8.787096567540166e-06,
+      "loss": 0.6912,
+      "step": 3901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6479293357520517,
+      "learning_rate": 8.786419733869262e-06,
+      "loss": 0.815,
+      "step": 3902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6572908508606126,
+      "learning_rate": 8.785742737487468e-06,
+      "loss": 0.7359,
+      "step": 3903
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6448834541728503,
+      "learning_rate": 8.78506557842388e-06,
+      "loss": 0.7263,
+      "step": 3904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9673500532420114,
+      "learning_rate": 8.78438825670759e-06,
+      "loss": 0.8493,
+      "step": 3905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7380829633221133,
+      "learning_rate": 8.78371077236771e-06,
+      "loss": 0.7811,
+      "step": 3906
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8174381276607,
+      "learning_rate": 8.783033125433354e-06,
+      "loss": 0.8909,
+      "step": 3907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.143327180177852,
+      "learning_rate": 8.782355315933638e-06,
+      "loss": 0.6121,
+      "step": 3908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6464816250500793,
+      "learning_rate": 8.781677343897687e-06,
+      "loss": 0.7932,
+      "step": 3909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6322691105433587,
+      "learning_rate": 8.780999209354642e-06,
+      "loss": 0.7943,
+      "step": 3910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.613871126320332,
+      "learning_rate": 8.780320912333639e-06,
+      "loss": 0.7531,
+      "step": 3911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.084727811858101,
+      "learning_rate": 8.779642452863827e-06,
+      "loss": 0.6943,
+      "step": 3912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7434904020367596,
+      "learning_rate": 8.778963830974362e-06,
+      "loss": 0.8207,
+      "step": 3913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2243389858599245,
+      "learning_rate": 8.778285046694403e-06,
+      "loss": 0.7536,
+      "step": 3914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7094032066449731,
+      "learning_rate": 8.777606100053123e-06,
+      "loss": 0.7902,
+      "step": 3915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.962461226088439,
+      "learning_rate": 8.776926991079696e-06,
+      "loss": 0.832,
+      "step": 3916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6747793440336742,
+      "learning_rate": 8.7762477198033e-06,
+      "loss": 0.8284,
+      "step": 3917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.063411330332013,
+      "learning_rate": 8.775568286253135e-06,
+      "loss": 0.7448,
+      "step": 3918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.715629374429834,
+      "learning_rate": 8.77488869045839e-06,
+      "loss": 0.7533,
+      "step": 3919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7838620044647309,
+      "learning_rate": 8.77420893244827e-06,
+      "loss": 0.6998,
+      "step": 3920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.593581912286335,
+      "learning_rate": 8.773529012251987e-06,
+      "loss": 0.7506,
+      "step": 3921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7274274404684795,
+      "learning_rate": 8.772848929898758e-06,
+      "loss": 0.6933,
+      "step": 3922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.653494885580054,
+      "learning_rate": 8.77216868541781e-06,
+      "loss": 0.9176,
+      "step": 3923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.742253668196518,
+      "learning_rate": 8.771488278838368e-06,
+      "loss": 0.8062,
+      "step": 3924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7082970745870316,
+      "learning_rate": 8.770807710189679e-06,
+      "loss": 0.7116,
+      "step": 3925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5065855158779733,
+      "learning_rate": 8.770126979500984e-06,
+      "loss": 0.7703,
+      "step": 3926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6955717675463242,
+      "learning_rate": 8.769446086801536e-06,
+      "loss": 0.7061,
+      "step": 3927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8646655742363039,
+      "learning_rate": 8.768765032120595e-06,
+      "loss": 0.7454,
+      "step": 3928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.682655105341969,
+      "learning_rate": 8.768083815487428e-06,
+      "loss": 0.9239,
+      "step": 3929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8415205100312744,
+      "learning_rate": 8.767402436931304e-06,
+      "loss": 0.712,
+      "step": 3930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7812585460212587,
+      "learning_rate": 8.76672089648151e-06,
+      "loss": 0.7292,
+      "step": 3931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7899110742239301,
+      "learning_rate": 8.766039194167328e-06,
+      "loss": 0.8733,
+      "step": 3932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6760666982968142,
+      "learning_rate": 8.765357330018056e-06,
+      "loss": 0.6907,
+      "step": 3933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7843166273899538,
+      "learning_rate": 8.764675304062992e-06,
+      "loss": 0.7364,
+      "step": 3934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6759840225309963,
+      "learning_rate": 8.763993116331448e-06,
+      "loss": 0.7633,
+      "step": 3935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7877425939964633,
+      "learning_rate": 8.763310766852736e-06,
+      "loss": 0.8306,
+      "step": 3936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.516951764225891,
+      "learning_rate": 8.762628255656178e-06,
+      "loss": 0.7326,
+      "step": 3937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6699749260369912,
+      "learning_rate": 8.761945582771104e-06,
+      "loss": 0.8311,
+      "step": 3938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6249939331260432,
+      "learning_rate": 8.76126274822685e-06,
+      "loss": 0.724,
+      "step": 3939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7093058893775106,
+      "learning_rate": 8.76057975205276e-06,
+      "loss": 0.6858,
+      "step": 3940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7673888521574064,
+      "learning_rate": 8.759896594278183e-06,
+      "loss": 0.7998,
+      "step": 3941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6657549028130585,
+      "learning_rate": 8.759213274932476e-06,
+      "loss": 0.7006,
+      "step": 3942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7983127205779736,
+      "learning_rate": 8.758529794045e-06,
+      "loss": 0.7546,
+      "step": 3943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7651192974173935,
+      "learning_rate": 8.75784615164513e-06,
+      "loss": 0.803,
+      "step": 3944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.730271738398202,
+      "learning_rate": 8.757162347762242e-06,
+      "loss": 0.8139,
+      "step": 3945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6438630809517583,
+      "learning_rate": 8.75647838242572e-06,
+      "loss": 0.7704,
+      "step": 3946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6802236136256576,
+      "learning_rate": 8.755794255664954e-06,
+      "loss": 0.7962,
+      "step": 3947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5326471101794785,
+      "learning_rate": 8.755109967509345e-06,
+      "loss": 0.8292,
+      "step": 3948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6503862987959943,
+      "learning_rate": 8.754425517988298e-06,
+      "loss": 0.7506,
+      "step": 3949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4454710053525832,
+      "learning_rate": 8.753740907131226e-06,
+      "loss": 0.8551,
+      "step": 3950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0978835310237975,
+      "learning_rate": 8.753056134967545e-06,
+      "loss": 0.701,
+      "step": 3951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7560146634181053,
+      "learning_rate": 8.752371201526685e-06,
+      "loss": 0.7963,
+      "step": 3952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4390290704535587,
+      "learning_rate": 8.751686106838078e-06,
+      "loss": 0.7528,
+      "step": 3953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0825158911142085,
+      "learning_rate": 8.751000850931162e-06,
+      "loss": 0.8003,
+      "step": 3954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0951293905759694,
+      "learning_rate": 8.750315433835387e-06,
+      "loss": 0.7587,
+      "step": 3955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.797495354999099,
+      "learning_rate": 8.749629855580204e-06,
+      "loss": 0.6687,
+      "step": 3956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9373275797191993,
+      "learning_rate": 8.748944116195076e-06,
+      "loss": 0.8212,
+      "step": 3957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0042801567460966,
+      "learning_rate": 8.74825821570947e-06,
+      "loss": 0.7846,
+      "step": 3958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.764909208662332,
+      "learning_rate": 8.74757215415286e-06,
+      "loss": 0.7364,
+      "step": 3959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6629061492777464,
+      "learning_rate": 8.74688593155473e-06,
+      "loss": 0.744,
+      "step": 3960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2193336776851313,
+      "learning_rate": 8.746199547944565e-06,
+      "loss": 0.8259,
+      "step": 3961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1655330803938293,
+      "learning_rate": 8.745513003351862e-06,
+      "loss": 0.7705,
+      "step": 3962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8300307197457737,
+      "learning_rate": 8.744826297806124e-06,
+      "loss": 0.7918,
+      "step": 3963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2456199065516436,
+      "learning_rate": 8.74413943133686e-06,
+      "loss": 0.6745,
+      "step": 3964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2376143947802785,
+      "learning_rate": 8.743452403973586e-06,
+      "loss": 0.7698,
+      "step": 3965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.017249914979392,
+      "learning_rate": 8.742765215745825e-06,
+      "loss": 0.8428,
+      "step": 3966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.208261569559267,
+      "learning_rate": 8.742077866683108e-06,
+      "loss": 0.7752,
+      "step": 3967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1055589676777777,
+      "learning_rate": 8.74139035681497e-06,
+      "loss": 0.7761,
+      "step": 3968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6893320211131369,
+      "learning_rate": 8.740702686170955e-06,
+      "loss": 0.7441,
+      "step": 3969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0864059130620378,
+      "learning_rate": 8.740014854780616e-06,
+      "loss": 0.7162,
+      "step": 3970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7504457272611087,
+      "learning_rate": 8.739326862673508e-06,
+      "loss": 0.7345,
+      "step": 3971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.010849550428793,
+      "learning_rate": 8.738638709879198e-06,
+      "loss": 0.73,
+      "step": 3972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9774741445529525,
+      "learning_rate": 8.737950396427255e-06,
+      "loss": 0.9295,
+      "step": 3973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9221012724854918,
+      "learning_rate": 8.73726192234726e-06,
+      "loss": 0.7921,
+      "step": 3974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2435294635357081,
+      "learning_rate": 8.736573287668797e-06,
+      "loss": 0.6486,
+      "step": 3975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9203544354824782,
+      "learning_rate": 8.735884492421457e-06,
+      "loss": 0.8677,
+      "step": 3976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.835537870445283,
+      "learning_rate": 8.73519553663484e-06,
+      "loss": 0.8369,
+      "step": 3977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1620483047391414,
+      "learning_rate": 8.734506420338554e-06,
+      "loss": 0.721,
+      "step": 3978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2439867676462593,
+      "learning_rate": 8.733817143562207e-06,
+      "loss": 0.6921,
+      "step": 3979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7160302677516777,
+      "learning_rate": 8.733127706335423e-06,
+      "loss": 0.8456,
+      "step": 3980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.807147111518304,
+      "learning_rate": 8.732438108687829e-06,
+      "loss": 0.8691,
+      "step": 3981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9876666472642464,
+      "learning_rate": 8.731748350649054e-06,
+      "loss": 0.8552,
+      "step": 3982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7349557550944665,
+      "learning_rate": 8.731058432248743e-06,
+      "loss": 0.7839,
+      "step": 3983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.765482827737261,
+      "learning_rate": 8.73036835351654e-06,
+      "loss": 0.8296,
+      "step": 3984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6414238405727237,
+      "learning_rate": 8.729678114482104e-06,
+      "loss": 0.7629,
+      "step": 3985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0025266776821717,
+      "learning_rate": 8.72898771517509e-06,
+      "loss": 0.8821,
+      "step": 3986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.929286784730121,
+      "learning_rate": 8.728297155625171e-06,
+      "loss": 0.803,
+      "step": 3987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5229084654245204,
+      "learning_rate": 8.727606435862018e-06,
+      "loss": 0.7885,
+      "step": 3988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6118120173362358,
+      "learning_rate": 8.726915555915317e-06,
+      "loss": 0.8536,
+      "step": 3989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.926858394755001,
+      "learning_rate": 8.726224515814752e-06,
+      "loss": 0.8279,
+      "step": 3990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1753858766572463,
+      "learning_rate": 8.72553331559002e-06,
+      "loss": 0.678,
+      "step": 3991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8268627154071468,
+      "learning_rate": 8.724841955270827e-06,
+      "loss": 0.7407,
+      "step": 3992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6798833830402637,
+      "learning_rate": 8.724150434886878e-06,
+      "loss": 0.7936,
+      "step": 3993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6220912634904283,
+      "learning_rate": 8.723458754467893e-06,
+      "loss": 0.7764,
+      "step": 3994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9281167906767327,
+      "learning_rate": 8.72276691404359e-06,
+      "loss": 0.8497,
+      "step": 3995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8429169388755506,
+      "learning_rate": 8.722074913643703e-06,
+      "loss": 0.8605,
+      "step": 3996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6360728467598282,
+      "learning_rate": 8.721382753297967e-06,
+      "loss": 0.8179,
+      "step": 3997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0152019840407873,
+      "learning_rate": 8.720690433036125e-06,
+      "loss": 0.8064,
+      "step": 3998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7508931966552224,
+      "learning_rate": 8.719997952887932e-06,
+      "loss": 0.8283,
+      "step": 3999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.126723105291032,
+      "learning_rate": 8.719305312883137e-06,
+      "loss": 0.6797,
+      "step": 4000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.003441947374116,
+      "learning_rate": 8.718612513051513e-06,
+      "loss": 0.7073,
+      "step": 4001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.06077698963385,
+      "learning_rate": 8.717919553422827e-06,
+      "loss": 0.6262,
+      "step": 4002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5604944956566336,
+      "learning_rate": 8.717226434026858e-06,
+      "loss": 0.6879,
+      "step": 4003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.576984247209525,
+      "learning_rate": 8.71653315489339e-06,
+      "loss": 0.7705,
+      "step": 4004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7887152808667106,
+      "learning_rate": 8.715839716052215e-06,
+      "loss": 0.7394,
+      "step": 4005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7668117018358518,
+      "learning_rate": 8.715146117533131e-06,
+      "loss": 0.8002,
+      "step": 4006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8116929754150426,
+      "learning_rate": 8.714452359365943e-06,
+      "loss": 0.8379,
+      "step": 4007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8871880929613258,
+      "learning_rate": 8.713758441580467e-06,
+      "loss": 0.8187,
+      "step": 4008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.444992662130309,
+      "learning_rate": 8.71306436420652e-06,
+      "loss": 0.6606,
+      "step": 4009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8611654312239336,
+      "learning_rate": 8.712370127273927e-06,
+      "loss": 0.7907,
+      "step": 4010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5398626385622394,
+      "learning_rate": 8.711675730812522e-06,
+      "loss": 0.7168,
+      "step": 4011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.882109729795171,
+      "learning_rate": 8.710981174852144e-06,
+      "loss": 0.708,
+      "step": 4012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8092799148177727,
+      "learning_rate": 8.710286459422641e-06,
+      "loss": 0.696,
+      "step": 4013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9220324989142648,
+      "learning_rate": 8.709591584553865e-06,
+      "loss": 0.7886,
+      "step": 4014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4734055964600066,
+      "learning_rate": 8.708896550275675e-06,
+      "loss": 0.7506,
+      "step": 4015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6895013250385582,
+      "learning_rate": 8.708201356617945e-06,
+      "loss": 0.8001,
+      "step": 4016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5688955606740522,
+      "learning_rate": 8.70750600361054e-06,
+      "loss": 0.7302,
+      "step": 4017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.675065940974605,
+      "learning_rate": 8.706810491283346e-06,
+      "loss": 0.6829,
+      "step": 4018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.248765876708306,
+      "learning_rate": 8.706114819666249e-06,
+      "loss": 0.7098,
+      "step": 4019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7839065082078456,
+      "learning_rate": 8.705418988789146e-06,
+      "loss": 0.67,
+      "step": 4020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1753165256338094,
+      "learning_rate": 8.704722998681937e-06,
+      "loss": 0.6617,
+      "step": 4021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6913307897474916,
+      "learning_rate": 8.704026849374527e-06,
+      "loss": 0.8046,
+      "step": 4022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.811818837155296,
+      "learning_rate": 8.703330540896836e-06,
+      "loss": 0.7222,
+      "step": 4023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5471829717917807,
+      "learning_rate": 8.702634073278784e-06,
+      "loss": 0.7699,
+      "step": 4024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8111998968583725,
+      "learning_rate": 8.701937446550298e-06,
+      "loss": 0.7661,
+      "step": 4025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0512873439263974,
+      "learning_rate": 8.701240660741317e-06,
+      "loss": 0.8347,
+      "step": 4026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9931182654028792,
+      "learning_rate": 8.700543715881781e-06,
+      "loss": 0.6416,
+      "step": 4027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1355856250757603,
+      "learning_rate": 8.699846612001638e-06,
+      "loss": 0.6521,
+      "step": 4028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.096669586866785,
+      "learning_rate": 8.699149349130848e-06,
+      "loss": 0.7349,
+      "step": 4029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0199257624059883,
+      "learning_rate": 8.698451927299374e-06,
+      "loss": 0.764,
+      "step": 4030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5976082602328958,
+      "learning_rate": 8.697754346537182e-06,
+      "loss": 0.7918,
+      "step": 4031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8254283635497535,
+      "learning_rate": 8.69705660687425e-06,
+      "loss": 0.8058,
+      "step": 4032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.792476501208208,
+      "learning_rate": 8.696358708340562e-06,
+      "loss": 0.8209,
+      "step": 4033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.682114266652429,
+      "learning_rate": 8.695660650966109e-06,
+      "loss": 0.8695,
+      "step": 4034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7545509161089161,
+      "learning_rate": 8.694962434780885e-06,
+      "loss": 0.7396,
+      "step": 4035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.588439448442217,
+      "learning_rate": 8.694264059814897e-06,
+      "loss": 0.7706,
+      "step": 4036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8125477743988552,
+      "learning_rate": 8.693565526098156e-06,
+      "loss": 0.8609,
+      "step": 4037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.170673146096281,
+      "learning_rate": 8.692866833660679e-06,
+      "loss": 0.7467,
+      "step": 4038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9414793535794572,
+      "learning_rate": 8.692167982532487e-06,
+      "loss": 0.7917,
+      "step": 4039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6137325490301364,
+      "learning_rate": 8.691468972743615e-06,
+      "loss": 0.8181,
+      "step": 4040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1719389205058355,
+      "learning_rate": 8.6907698043241e-06,
+      "loss": 0.5821,
+      "step": 4041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6023389868627949,
+      "learning_rate": 8.690070477303987e-06,
+      "loss": 0.6971,
+      "step": 4042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7507031181633348,
+      "learning_rate": 8.689370991713327e-06,
+      "loss": 0.8434,
+      "step": 4043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7158152035771803,
+      "learning_rate": 8.688671347582178e-06,
+      "loss": 0.7973,
+      "step": 4044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.295884071026563,
+      "learning_rate": 8.687971544940608e-06,
+      "loss": 0.8294,
+      "step": 4045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.156787977637143,
+      "learning_rate": 8.687271583818687e-06,
+      "loss": 0.7849,
+      "step": 4046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6216659299269292,
+      "learning_rate": 8.686571464246491e-06,
+      "loss": 0.6891,
+      "step": 4047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7146007803534986,
+      "learning_rate": 8.685871186254112e-06,
+      "loss": 0.8087,
+      "step": 4048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6365661862385454,
+      "learning_rate": 8.685170749871638e-06,
+      "loss": 0.7686,
+      "step": 4049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7091966178974187,
+      "learning_rate": 8.68447015512917e-06,
+      "loss": 0.8374,
+      "step": 4050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0918218818404744,
+      "learning_rate": 8.683769402056814e-06,
+      "loss": 0.6685,
+      "step": 4051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6588914723284514,
+      "learning_rate": 8.683068490684681e-06,
+      "loss": 0.8015,
+      "step": 4052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8887420650285514,
+      "learning_rate": 8.682367421042895e-06,
+      "loss": 0.7958,
+      "step": 4053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7296913760541983,
+      "learning_rate": 8.681666193161578e-06,
+      "loss": 0.8372,
+      "step": 4054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1211068456957138,
+      "learning_rate": 8.680964807070865e-06,
+      "loss": 0.7617,
+      "step": 4055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7499687703500564,
+      "learning_rate": 8.680263262800897e-06,
+      "loss": 0.7887,
+      "step": 4056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7699375104482153,
+      "learning_rate": 8.679561560381818e-06,
+      "loss": 0.7983,
+      "step": 4057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6936443638987502,
+      "learning_rate": 8.678859699843787e-06,
+      "loss": 0.7675,
+      "step": 4058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8604900039841172,
+      "learning_rate": 8.67815768121696e-06,
+      "loss": 0.7313,
+      "step": 4059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.699727146779617,
+      "learning_rate": 8.677455504531507e-06,
+      "loss": 0.7634,
+      "step": 4060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5438574495031503,
+      "learning_rate": 8.676753169817598e-06,
+      "loss": 0.7707,
+      "step": 4061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4949261539402057,
+      "learning_rate": 8.676050677105419e-06,
+      "loss": 0.816,
+      "step": 4062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8894150333988395,
+      "learning_rate": 8.675348026425154e-06,
+      "loss": 0.7313,
+      "step": 4063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7767568069817121,
+      "learning_rate": 8.674645217807e-06,
+      "loss": 0.8111,
+      "step": 4064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.574686660170029,
+      "learning_rate": 8.673942251281158e-06,
+      "loss": 0.8072,
+      "step": 4065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7190863016072662,
+      "learning_rate": 8.673239126877835e-06,
+      "loss": 0.7066,
+      "step": 4066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6422967847908505,
+      "learning_rate": 8.672535844627243e-06,
+      "loss": 0.7691,
+      "step": 4067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8038303975618346,
+      "learning_rate": 8.67183240455961e-06,
+      "loss": 0.9094,
+      "step": 4068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8556187354301437,
+      "learning_rate": 8.671128806705159e-06,
+      "loss": 0.8173,
+      "step": 4069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.660560617948906,
+      "learning_rate": 8.670425051094128e-06,
+      "loss": 0.7108,
+      "step": 4070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5657951999482154,
+      "learning_rate": 8.669721137756756e-06,
+      "loss": 0.6933,
+      "step": 4071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6634532372203672,
+      "learning_rate": 8.669017066723297e-06,
+      "loss": 0.656,
+      "step": 4072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6278642164504158,
+      "learning_rate": 8.668312838024002e-06,
+      "loss": 0.8069,
+      "step": 4073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2438144880221746,
+      "learning_rate": 8.667608451689135e-06,
+      "loss": 0.6034,
+      "step": 4074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.83652116228529,
+      "learning_rate": 8.666903907748963e-06,
+      "loss": 0.7985,
+      "step": 4075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5836104917731675,
+      "learning_rate": 8.666199206233765e-06,
+      "loss": 0.795,
+      "step": 4076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5715808761255667,
+      "learning_rate": 8.665494347173822e-06,
+      "loss": 0.7308,
+      "step": 4077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6354095672626534,
+      "learning_rate": 8.664789330599423e-06,
+      "loss": 0.9157,
+      "step": 4078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6401280417495017,
+      "learning_rate": 8.664084156540864e-06,
+      "loss": 0.7406,
+      "step": 4079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6425873729736253,
+      "learning_rate": 8.66337882502845e-06,
+      "loss": 0.7179,
+      "step": 4080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4718404688003335,
+      "learning_rate": 8.662673336092487e-06,
+      "loss": 0.7525,
+      "step": 4081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8640937739590044,
+      "learning_rate": 8.661967689763296e-06,
+      "loss": 0.7631,
+      "step": 4082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7392543207198032,
+      "learning_rate": 8.661261886071194e-06,
+      "loss": 0.8644,
+      "step": 4083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.823887833613513,
+      "learning_rate": 8.660555925046518e-06,
+      "loss": 0.7813,
+      "step": 4084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5271454105820936,
+      "learning_rate": 8.659849806719602e-06,
+      "loss": 0.7149,
+      "step": 4085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7180862560120231,
+      "learning_rate": 8.659143531120785e-06,
+      "loss": 0.8046,
+      "step": 4086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9559705682978417,
+      "learning_rate": 8.658437098280426e-06,
+      "loss": 0.6679,
+      "step": 4087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.367052575186595,
+      "learning_rate": 8.657730508228874e-06,
+      "loss": 0.7227,
+      "step": 4088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8027779189394129,
+      "learning_rate": 8.657023760996497e-06,
+      "loss": 0.8092,
+      "step": 4089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5887834198435435,
+      "learning_rate": 8.656316856613662e-06,
+      "loss": 0.6972,
+      "step": 4090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7779879849271538,
+      "learning_rate": 8.655609795110751e-06,
+      "loss": 0.7172,
+      "step": 4091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7334826673373223,
+      "learning_rate": 8.654902576518145e-06,
+      "loss": 0.8017,
+      "step": 4092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0861932192884713,
+      "learning_rate": 8.654195200866236e-06,
+      "loss": 0.6786,
+      "step": 4093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.52442981779097,
+      "learning_rate": 8.653487668185419e-06,
+      "loss": 0.8109,
+      "step": 4094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5758547238541756,
+      "learning_rate": 8.652779978506103e-06,
+      "loss": 0.8579,
+      "step": 4095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.887996032575942,
+      "learning_rate": 8.652072131858694e-06,
+      "loss": 0.8299,
+      "step": 4096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.729050202997483,
+      "learning_rate": 8.651364128273612e-06,
+      "loss": 0.8113,
+      "step": 4097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7833980835854926,
+      "learning_rate": 8.650655967781282e-06,
+      "loss": 0.771,
+      "step": 4098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9584648190426877,
+      "learning_rate": 8.649947650412135e-06,
+      "loss": 0.7546,
+      "step": 4099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0463547406819107,
+      "learning_rate": 8.649239176196609e-06,
+      "loss": 0.7057,
+      "step": 4100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6334577435700364,
+      "learning_rate": 8.648530545165146e-06,
+      "loss": 0.7336,
+      "step": 4101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5285415463018392,
+      "learning_rate": 8.647821757348202e-06,
+      "loss": 0.6892,
+      "step": 4102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5396458819359449,
+      "learning_rate": 8.647112812776231e-06,
+      "loss": 0.801,
+      "step": 4103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8389992979580834,
+      "learning_rate": 8.646403711479702e-06,
+      "loss": 0.787,
+      "step": 4104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6481078341683884,
+      "learning_rate": 8.645694453489085e-06,
+      "loss": 0.7952,
+      "step": 4105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9511926178611418,
+      "learning_rate": 8.644985038834855e-06,
+      "loss": 0.7023,
+      "step": 4106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3674636078436544,
+      "learning_rate": 8.644275467547502e-06,
+      "loss": 0.6977,
+      "step": 4107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.809868939165386,
+      "learning_rate": 8.643565739657515e-06,
+      "loss": 0.8086,
+      "step": 4108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6610508226070975,
+      "learning_rate": 8.642855855195394e-06,
+      "loss": 0.8277,
+      "step": 4109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.724940471152718,
+      "learning_rate": 8.642145814191643e-06,
+      "loss": 0.8528,
+      "step": 4110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5254621671866984,
+      "learning_rate": 8.641435616676776e-06,
+      "loss": 0.8311,
+      "step": 4111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9490338522071464,
+      "learning_rate": 8.64072526268131e-06,
+      "loss": 0.8595,
+      "step": 4112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.531537579697927,
+      "learning_rate": 8.640014752235773e-06,
+      "loss": 0.6957,
+      "step": 4113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0400584143542788,
+      "learning_rate": 8.639304085370692e-06,
+      "loss": 0.639,
+      "step": 4114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2956870473904682,
+      "learning_rate": 8.638593262116612e-06,
+      "loss": 0.6628,
+      "step": 4115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.882377252908706,
+      "learning_rate": 8.637882282504075e-06,
+      "loss": 0.7561,
+      "step": 4116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4214107375176497,
+      "learning_rate": 8.637171146563634e-06,
+      "loss": 0.7014,
+      "step": 4117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5314808055246738,
+      "learning_rate": 8.636459854325849e-06,
+      "loss": 0.7159,
+      "step": 4118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0968636976139907,
+      "learning_rate": 8.635748405821285e-06,
+      "loss": 0.6973,
+      "step": 4119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4963198795105985,
+      "learning_rate": 8.635036801080513e-06,
+      "loss": 0.6822,
+      "step": 4120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1381533942756825,
+      "learning_rate": 8.634325040134117e-06,
+      "loss": 0.6659,
+      "step": 4121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.633133308608957,
+      "learning_rate": 8.63361312301268e-06,
+      "loss": 0.8299,
+      "step": 4122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0622397372340704,
+      "learning_rate": 8.632901049746793e-06,
+      "loss": 0.6392,
+      "step": 4123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.173082256370053,
+      "learning_rate": 8.632188820367056e-06,
+      "loss": 0.7448,
+      "step": 4124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.866579232562236,
+      "learning_rate": 8.631476434904077e-06,
+      "loss": 0.823,
+      "step": 4125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9730999718471198,
+      "learning_rate": 8.63076389338847e-06,
+      "loss": 0.8583,
+      "step": 4126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2302976516699946,
+      "learning_rate": 8.630051195850851e-06,
+      "loss": 0.7645,
+      "step": 4127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5726759222057225,
+      "learning_rate": 8.629338342321846e-06,
+      "loss": 0.6581,
+      "step": 4128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7621507650758133,
+      "learning_rate": 8.62862533283209e-06,
+      "loss": 0.7895,
+      "step": 4129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4954370385872966,
+      "learning_rate": 8.627912167412222e-06,
+      "loss": 0.7295,
+      "step": 4130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6042582928449052,
+      "learning_rate": 8.62719884609289e-06,
+      "loss": 0.7889,
+      "step": 4131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8601918703992908,
+      "learning_rate": 8.626485368904744e-06,
+      "loss": 0.8254,
+      "step": 4132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6932945193712665,
+      "learning_rate": 8.625771735878445e-06,
+      "loss": 0.8485,
+      "step": 4133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4939736022810588,
+      "learning_rate": 8.625057947044662e-06,
+      "loss": 0.6968,
+      "step": 4134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.768333994300861,
+      "learning_rate": 8.624344002434061e-06,
+      "loss": 0.9229,
+      "step": 4135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.020059824622024,
+      "learning_rate": 8.62362990207733e-06,
+      "loss": 0.8456,
+      "step": 4136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.664343283403647,
+      "learning_rate": 8.622915646005152e-06,
+      "loss": 0.7335,
+      "step": 4137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7799778843081722,
+      "learning_rate": 8.622201234248218e-06,
+      "loss": 0.7807,
+      "step": 4138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5463675844341909,
+      "learning_rate": 8.62148666683723e-06,
+      "loss": 0.8223,
+      "step": 4139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.283950521833614,
+      "learning_rate": 8.620771943802895e-06,
+      "loss": 0.6714,
+      "step": 4140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6639365620281483,
+      "learning_rate": 8.620057065175926e-06,
+      "loss": 0.7323,
+      "step": 4141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.602303244882245,
+      "learning_rate": 8.619342030987044e-06,
+      "loss": 0.6943,
+      "step": 4142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.372363472495564,
+      "learning_rate": 8.618626841266972e-06,
+      "loss": 0.5685,
+      "step": 4143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6216123555306639,
+      "learning_rate": 8.617911496046446e-06,
+      "loss": 0.8355,
+      "step": 4144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.692126819050093,
+      "learning_rate": 8.617195995356206e-06,
+      "loss": 0.8517,
+      "step": 4145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9964139101554215,
+      "learning_rate": 8.616480339226998e-06,
+      "loss": 0.7481,
+      "step": 4146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9598868464993469,
+      "learning_rate": 8.615764527689574e-06,
+      "loss": 0.734,
+      "step": 4147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6735823881270313,
+      "learning_rate": 8.615048560774698e-06,
+      "loss": 0.7554,
+      "step": 4148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.319457073189903,
+      "learning_rate": 8.614332438513132e-06,
+      "loss": 0.7319,
+      "step": 4149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8289369910376787,
+      "learning_rate": 8.613616160935652e-06,
+      "loss": 0.8036,
+      "step": 4150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.706204374773201,
+      "learning_rate": 8.612899728073039e-06,
+      "loss": 0.8428,
+      "step": 4151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.515021824937319,
+      "learning_rate": 8.612183139956078e-06,
+      "loss": 0.6351,
+      "step": 4152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5613196755924814,
+      "learning_rate": 8.611466396615562e-06,
+      "loss": 0.8449,
+      "step": 4153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4158570351679296,
+      "learning_rate": 8.610749498082291e-06,
+      "loss": 0.8072,
+      "step": 4154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6065976522848948,
+      "learning_rate": 8.610032444387074e-06,
+      "loss": 0.7909,
+      "step": 4155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3912568487158392,
+      "learning_rate": 8.609315235560722e-06,
+      "loss": 0.7264,
+      "step": 4156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7338042586338764,
+      "learning_rate": 8.608597871634056e-06,
+      "loss": 0.881,
+      "step": 4157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.670173685872696,
+      "learning_rate": 8.607880352637905e-06,
+      "loss": 0.7079,
+      "step": 4158
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9557168129163995,
+      "learning_rate": 8.607162678603097e-06,
+      "loss": 0.7445,
+      "step": 4159
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5600598353938813,
+      "learning_rate": 8.606444849560476e-06,
+      "loss": 0.7428,
+      "step": 4160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4845810163936297,
+      "learning_rate": 8.605726865540889e-06,
+      "loss": 0.7234,
+      "step": 4161
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6492469455091296,
+      "learning_rate": 8.605008726575186e-06,
+      "loss": 0.7887,
+      "step": 4162
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2503478452415686,
+      "learning_rate": 8.60429043269423e-06,
+      "loss": 0.6264,
+      "step": 4163
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5423212680248357,
+      "learning_rate": 8.603571983928888e-06,
+      "loss": 0.6873,
+      "step": 4164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6275934062442525,
+      "learning_rate": 8.602853380310033e-06,
+      "loss": 0.6969,
+      "step": 4165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.659102003494593,
+      "learning_rate": 8.602134621868542e-06,
+      "loss": 0.7477,
+      "step": 4166
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8367222375283594,
+      "learning_rate": 8.601415708635304e-06,
+      "loss": 0.7956,
+      "step": 4167
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.280323103802973,
+      "learning_rate": 8.600696640641213e-06,
+      "loss": 0.7486,
+      "step": 4168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4809602697456634,
+      "learning_rate": 8.599977417917169e-06,
+      "loss": 0.7713,
+      "step": 4169
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7481124880668075,
+      "learning_rate": 8.599258040494078e-06,
+      "loss": 0.7415,
+      "step": 4170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.689796054683759,
+      "learning_rate": 8.59853850840285e-06,
+      "loss": 0.7113,
+      "step": 4171
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.690867317709853,
+      "learning_rate": 8.59781882167441e-06,
+      "loss": 0.8368,
+      "step": 4172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.10576241471904,
+      "learning_rate": 8.597098980339683e-06,
+      "loss": 0.7438,
+      "step": 4173
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6614704630689554,
+      "learning_rate": 8.5963789844296e-06,
+      "loss": 0.7636,
+      "step": 4174
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.124106975055061,
+      "learning_rate": 8.595658833975104e-06,
+      "loss": 0.8758,
+      "step": 4175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8947474689341604,
+      "learning_rate": 8.59493852900714e-06,
+      "loss": 0.8073,
+      "step": 4176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1071137818470032,
+      "learning_rate": 8.59421806955666e-06,
+      "loss": 0.6494,
+      "step": 4177
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8307432246364292,
+      "learning_rate": 8.593497455654627e-06,
+      "loss": 0.7739,
+      "step": 4178
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8771877308427376,
+      "learning_rate": 8.592776687332003e-06,
+      "loss": 0.6153,
+      "step": 4179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3062777263759455,
+      "learning_rate": 8.592055764619762e-06,
+      "loss": 0.66,
+      "step": 4180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7291172294061035,
+      "learning_rate": 8.591334687548888e-06,
+      "loss": 0.8473,
+      "step": 4181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.11207204862944,
+      "learning_rate": 8.590613456150364e-06,
+      "loss": 0.6905,
+      "step": 4182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7771009150242785,
+      "learning_rate": 8.58989207045518e-06,
+      "loss": 0.853,
+      "step": 4183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6223293790101676,
+      "learning_rate": 8.58917053049434e-06,
+      "loss": 0.6615,
+      "step": 4184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9283920263812733,
+      "learning_rate": 8.58844883629885e-06,
+      "loss": 0.7417,
+      "step": 4185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.948492613429596,
+      "learning_rate": 8.58772698789972e-06,
+      "loss": 0.7863,
+      "step": 4186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5213425903993465,
+      "learning_rate": 8.587004985327971e-06,
+      "loss": 0.656,
+      "step": 4187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.616596225328182,
+      "learning_rate": 8.586282828614631e-06,
+      "loss": 0.7819,
+      "step": 4188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6944925805838047,
+      "learning_rate": 8.58556051779073e-06,
+      "loss": 0.7297,
+      "step": 4189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0508734637693562,
+      "learning_rate": 8.584838052887308e-06,
+      "loss": 0.6563,
+      "step": 4190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.523106938243606,
+      "learning_rate": 8.58411543393541e-06,
+      "loss": 0.7882,
+      "step": 4191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.010191338637198,
+      "learning_rate": 8.58339266096609e-06,
+      "loss": 0.7914,
+      "step": 4192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7209378013498433,
+      "learning_rate": 8.582669734010407e-06,
+      "loss": 0.7674,
+      "step": 4193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.923426060272006,
+      "learning_rate": 8.581946653099427e-06,
+      "loss": 0.7529,
+      "step": 4194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6453221326797778,
+      "learning_rate": 8.581223418264222e-06,
+      "loss": 0.7835,
+      "step": 4195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8308292045663659,
+      "learning_rate": 8.58050002953587e-06,
+      "loss": 0.7638,
+      "step": 4196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.799174688043088,
+      "learning_rate": 8.579776486945457e-06,
+      "loss": 0.728,
+      "step": 4197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.96504139653902,
+      "learning_rate": 8.579052790524077e-06,
+      "loss": 0.7745,
+      "step": 4198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7177369580540458,
+      "learning_rate": 8.578328940302827e-06,
+      "loss": 0.7749,
+      "step": 4199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8571451034171758,
+      "learning_rate": 8.577604936312813e-06,
+      "loss": 0.735,
+      "step": 4200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6686611404202465,
+      "learning_rate": 8.576880778585148e-06,
+      "loss": 0.7775,
+      "step": 4201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.821752450144449,
+      "learning_rate": 8.57615646715095e-06,
+      "loss": 0.7256,
+      "step": 4202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9826308796907848,
+      "learning_rate": 8.575432002041341e-06,
+      "loss": 0.742,
+      "step": 4203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7928872651926286,
+      "learning_rate": 8.574707383287459e-06,
+      "loss": 0.7096,
+      "step": 4204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.592562743640663,
+      "learning_rate": 8.57398261092044e-06,
+      "loss": 0.7673,
+      "step": 4205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5141625625994852,
+      "learning_rate": 8.573257684971425e-06,
+      "loss": 0.7133,
+      "step": 4206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8679914869093162,
+      "learning_rate": 8.572532605471572e-06,
+      "loss": 0.8165,
+      "step": 4207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7866697980202149,
+      "learning_rate": 8.571807372452036e-06,
+      "loss": 0.7922,
+      "step": 4208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.792662437264148,
+      "learning_rate": 8.571081985943984e-06,
+      "loss": 0.7317,
+      "step": 4209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7561753673168343,
+      "learning_rate": 8.570356445978583e-06,
+      "loss": 0.7311,
+      "step": 4210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7896744647569116,
+      "learning_rate": 8.569630752587014e-06,
+      "loss": 0.7303,
+      "step": 4211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.737727330985594,
+      "learning_rate": 8.568904905800464e-06,
+      "loss": 0.8337,
+      "step": 4212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6097943512481863,
+      "learning_rate": 8.56817890565012e-06,
+      "loss": 0.7469,
+      "step": 4213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2752827846675818,
+      "learning_rate": 8.567452752167183e-06,
+      "loss": 0.7274,
+      "step": 4214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4784339294824256,
+      "learning_rate": 8.566726445382854e-06,
+      "loss": 0.7984,
+      "step": 4215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7278427870118125,
+      "learning_rate": 8.565999985328348e-06,
+      "loss": 0.6815,
+      "step": 4216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8459618213661393,
+      "learning_rate": 8.565273372034879e-06,
+      "loss": 0.7407,
+      "step": 4217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4354500255938778,
+      "learning_rate": 8.564546605533674e-06,
+      "loss": 0.8072,
+      "step": 4218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.776821549810797,
+      "learning_rate": 8.563819685855963e-06,
+      "loss": 0.78,
+      "step": 4219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8344628610419906,
+      "learning_rate": 8.563092613032981e-06,
+      "loss": 0.86,
+      "step": 4220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2096336126726273,
+      "learning_rate": 8.562365387095977e-06,
+      "loss": 0.8119,
+      "step": 4221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.779720983238152,
+      "learning_rate": 8.561638008076197e-06,
+      "loss": 0.8754,
+      "step": 4222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.757974640486943,
+      "learning_rate": 8.5609104760049e-06,
+      "loss": 0.7078,
+      "step": 4223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0807015368361443,
+      "learning_rate": 8.560182790913349e-06,
+      "loss": 0.6877,
+      "step": 4224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.646718888362747,
+      "learning_rate": 8.559454952832815e-06,
+      "loss": 0.8386,
+      "step": 4225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.892307945757563,
+      "learning_rate": 8.558726961794573e-06,
+      "loss": 0.7692,
+      "step": 4226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1043193850000583,
+      "learning_rate": 8.557998817829909e-06,
+      "loss": 0.6868,
+      "step": 4227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7867983745020182,
+      "learning_rate": 8.557270520970111e-06,
+      "loss": 0.8828,
+      "step": 4228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.821203505205144,
+      "learning_rate": 8.556542071246476e-06,
+      "loss": 0.7745,
+      "step": 4229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8122269720893822,
+      "learning_rate": 8.555813468690309e-06,
+      "loss": 0.8395,
+      "step": 4230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.683771234996193,
+      "learning_rate": 8.555084713332917e-06,
+      "loss": 0.7709,
+      "step": 4231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3520382950709082,
+      "learning_rate": 8.554355805205616e-06,
+      "loss": 0.7501,
+      "step": 4232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2095727889231593,
+      "learning_rate": 8.553626744339732e-06,
+      "loss": 0.7706,
+      "step": 4233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1855897189532696,
+      "learning_rate": 8.552897530766592e-06,
+      "loss": 0.6681,
+      "step": 4234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7116740565434445,
+      "learning_rate": 8.552168164517532e-06,
+      "loss": 0.8228,
+      "step": 4235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5717658286543954,
+      "learning_rate": 8.551438645623896e-06,
+      "loss": 0.7264,
+      "step": 4236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1114830430514138,
+      "learning_rate": 8.55070897411703e-06,
+      "loss": 0.7031,
+      "step": 4237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6824980029611054,
+      "learning_rate": 8.549979150028292e-06,
+      "loss": 0.8802,
+      "step": 4238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5927151593221076,
+      "learning_rate": 8.549249173389045e-06,
+      "loss": 0.6804,
+      "step": 4239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7706192808251338,
+      "learning_rate": 8.548519044230654e-06,
+      "loss": 0.8328,
+      "step": 4240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6017444446912676,
+      "learning_rate": 8.5477887625845e-06,
+      "loss": 0.905,
+      "step": 4241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6494873727230654,
+      "learning_rate": 8.547058328481959e-06,
+      "loss": 0.6562,
+      "step": 4242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4963093124618696,
+      "learning_rate": 8.546327741954423e-06,
+      "loss": 0.7042,
+      "step": 4243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.553613837971926,
+      "learning_rate": 8.545597003033286e-06,
+      "loss": 0.7635,
+      "step": 4244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6193916637215786,
+      "learning_rate": 8.544866111749948e-06,
+      "loss": 0.7301,
+      "step": 4245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0754587262884392,
+      "learning_rate": 8.544135068135819e-06,
+      "loss": 0.5704,
+      "step": 4246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9142252725640325,
+      "learning_rate": 8.543403872222313e-06,
+      "loss": 0.7525,
+      "step": 4247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8458600475836169,
+      "learning_rate": 8.54267252404085e-06,
+      "loss": 0.8124,
+      "step": 4248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5941124712384374,
+      "learning_rate": 8.54194102362286e-06,
+      "loss": 0.8024,
+      "step": 4249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8625067527768033,
+      "learning_rate": 8.541209370999777e-06,
+      "loss": 0.7421,
+      "step": 4250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6370115109572325,
+      "learning_rate": 8.540477566203039e-06,
+      "loss": 0.7605,
+      "step": 4251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8665881809301144,
+      "learning_rate": 8.539745609264094e-06,
+      "loss": 0.8813,
+      "step": 4252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.577476308383605,
+      "learning_rate": 8.539013500214399e-06,
+      "loss": 0.8556,
+      "step": 4253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6731443596082896,
+      "learning_rate": 8.538281239085411e-06,
+      "loss": 0.6643,
+      "step": 4254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.032016818420302,
+      "learning_rate": 8.537548825908597e-06,
+      "loss": 0.7586,
+      "step": 4255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7788187651545366,
+      "learning_rate": 8.536816260715433e-06,
+      "loss": 0.7296,
+      "step": 4256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5781579824312952,
+      "learning_rate": 8.536083543537396e-06,
+      "loss": 0.7594,
+      "step": 4257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6783438446281354,
+      "learning_rate": 8.535350674405975e-06,
+      "loss": 0.81,
+      "step": 4258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.027469944639651,
+      "learning_rate": 8.534617653352661e-06,
+      "loss": 0.8738,
+      "step": 4259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8019604659162207,
+      "learning_rate": 8.533884480408955e-06,
+      "loss": 0.7615,
+      "step": 4260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0706532900150694,
+      "learning_rate": 8.533151155606364e-06,
+      "loss": 0.6448,
+      "step": 4261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8859495078601975,
+      "learning_rate": 8.532417678976398e-06,
+      "loss": 0.6933,
+      "step": 4262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5339667130991026,
+      "learning_rate": 8.531684050550575e-06,
+      "loss": 0.7644,
+      "step": 4263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3892346470687984,
+      "learning_rate": 8.530950270360425e-06,
+      "loss": 0.7115,
+      "step": 4264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.965567319509786,
+      "learning_rate": 8.530216338437478e-06,
+      "loss": 0.844,
+      "step": 4265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.101854559552234,
+      "learning_rate": 8.529482254813272e-06,
+      "loss": 0.8664,
+      "step": 4266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6034192100460234,
+      "learning_rate": 8.528748019519352e-06,
+      "loss": 0.7911,
+      "step": 4267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2154067917911846,
+      "learning_rate": 8.528013632587272e-06,
+      "loss": 0.6586,
+      "step": 4268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6202115923540485,
+      "learning_rate": 8.52727909404859e-06,
+      "loss": 0.7846,
+      "step": 4269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7219079672219408,
+      "learning_rate": 8.526544403934868e-06,
+      "loss": 0.7874,
+      "step": 4270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9399704127439912,
+      "learning_rate": 8.52580956227768e-06,
+      "loss": 0.8493,
+      "step": 4271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.819147511961464,
+      "learning_rate": 8.525074569108603e-06,
+      "loss": 0.8111,
+      "step": 4272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0074287430209554,
+      "learning_rate": 8.524339424459219e-06,
+      "loss": 0.7877,
+      "step": 4273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6939129601703096,
+      "learning_rate": 8.523604128361123e-06,
+      "loss": 0.7905,
+      "step": 4274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5159805282992416,
+      "learning_rate": 8.522868680845908e-06,
+      "loss": 0.7708,
+      "step": 4275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7113449355559758,
+      "learning_rate": 8.52213308194518e-06,
+      "loss": 0.6481,
+      "step": 4276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6303481917433764,
+      "learning_rate": 8.521397331690551e-06,
+      "loss": 0.8298,
+      "step": 4277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.325972186692281,
+      "learning_rate": 8.520661430113637e-06,
+      "loss": 0.7589,
+      "step": 4278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8995873613086207,
+      "learning_rate": 8.519925377246057e-06,
+      "loss": 0.6881,
+      "step": 4279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.634059779394506,
+      "learning_rate": 8.519189173119446e-06,
+      "loss": 0.7342,
+      "step": 4280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7408685201648777,
+      "learning_rate": 8.518452817765438e-06,
+      "loss": 0.8299,
+      "step": 4281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0171482750433007,
+      "learning_rate": 8.517716311215678e-06,
+      "loss": 0.8612,
+      "step": 4282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.486547668235507,
+      "learning_rate": 8.516979653501813e-06,
+      "loss": 0.6953,
+      "step": 4283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2451123050253345,
+      "learning_rate": 8.516242844655498e-06,
+      "loss": 0.6781,
+      "step": 4284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9692319839958468,
+      "learning_rate": 8.515505884708399e-06,
+      "loss": 0.7136,
+      "step": 4285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8135995930752067,
+      "learning_rate": 8.514768773692182e-06,
+      "loss": 0.6463,
+      "step": 4286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6346382432573618,
+      "learning_rate": 8.514031511638524e-06,
+      "loss": 0.6198,
+      "step": 4287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7833806457560224,
+      "learning_rate": 8.513294098579103e-06,
+      "loss": 0.7358,
+      "step": 4288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.090947782306075,
+      "learning_rate": 8.512556534545612e-06,
+      "loss": 0.5846,
+      "step": 4289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0969995207747525,
+      "learning_rate": 8.511818819569743e-06,
+      "loss": 0.5572,
+      "step": 4290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.652424850795297,
+      "learning_rate": 8.5110809536832e-06,
+      "loss": 0.8433,
+      "step": 4291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7392710872117003,
+      "learning_rate": 8.510342936917685e-06,
+      "loss": 0.7389,
+      "step": 4292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4341279186992952,
+      "learning_rate": 8.50960476930492e-06,
+      "loss": 0.7168,
+      "step": 4293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7683845022380795,
+      "learning_rate": 8.50886645087662e-06,
+      "loss": 0.7855,
+      "step": 4294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7909391413848246,
+      "learning_rate": 8.508127981664514e-06,
+      "loss": 0.806,
+      "step": 4295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7940103714878033,
+      "learning_rate": 8.507389361700335e-06,
+      "loss": 0.6635,
+      "step": 4296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7233290556966403,
+      "learning_rate": 8.506650591015825e-06,
+      "loss": 0.9642,
+      "step": 4297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4863177266326963,
+      "learning_rate": 8.50591166964273e-06,
+      "loss": 0.6711,
+      "step": 4298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5936218284732775,
+      "learning_rate": 8.5051725976128e-06,
+      "loss": 0.7718,
+      "step": 4299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8780999304512491,
+      "learning_rate": 8.504433374957799e-06,
+      "loss": 0.6821,
+      "step": 4300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5931036144990107,
+      "learning_rate": 8.50369400170949e-06,
+      "loss": 0.7779,
+      "step": 4301
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2195866942502667,
+      "learning_rate": 8.502954477899647e-06,
+      "loss": 0.7444,
+      "step": 4302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6115338331225364,
+      "learning_rate": 8.50221480356005e-06,
+      "loss": 0.7189,
+      "step": 4303
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3248215820188207,
+      "learning_rate": 8.50147497872248e-06,
+      "loss": 0.6886,
+      "step": 4304
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9880843101059846,
+      "learning_rate": 8.500735003418734e-06,
+      "loss": 0.9641,
+      "step": 4305
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9825012961390758,
+      "learning_rate": 8.499994877680609e-06,
+      "loss": 0.7213,
+      "step": 4306
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5712934998876775,
+      "learning_rate": 8.499254601539908e-06,
+      "loss": 0.7675,
+      "step": 4307
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6763503377536482,
+      "learning_rate": 8.498514175028442e-06,
+      "loss": 0.6631,
+      "step": 4308
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.931792785720067,
+      "learning_rate": 8.497773598178033e-06,
+      "loss": 0.7092,
+      "step": 4309
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7828624081900686,
+      "learning_rate": 8.497032871020501e-06,
+      "loss": 0.7496,
+      "step": 4310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5255491966166723,
+      "learning_rate": 8.49629199358768e-06,
+      "loss": 0.6789,
+      "step": 4311
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8254594147134369,
+      "learning_rate": 8.495550965911403e-06,
+      "loss": 0.8969,
+      "step": 4312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.763805665959888,
+      "learning_rate": 8.494809788023518e-06,
+      "loss": 0.7883,
+      "step": 4313
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6296053651409157,
+      "learning_rate": 8.494068459955871e-06,
+      "loss": 0.7242,
+      "step": 4314
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7842124620580644,
+      "learning_rate": 8.493326981740322e-06,
+      "loss": 0.7491,
+      "step": 4315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.662832669391192,
+      "learning_rate": 8.492585353408732e-06,
+      "loss": 0.7659,
+      "step": 4316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1116651804293225,
+      "learning_rate": 8.491843574992971e-06,
+      "loss": 0.6889,
+      "step": 4317
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4867368651221713,
+      "learning_rate": 8.491101646524916e-06,
+      "loss": 0.8332,
+      "step": 4318
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6640392820165186,
+      "learning_rate": 8.490359568036446e-06,
+      "loss": 0.7704,
+      "step": 4319
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.231377322980365,
+      "learning_rate": 8.489617339559455e-06,
+      "loss": 0.7212,
+      "step": 4320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7660939656403642,
+      "learning_rate": 8.488874961125832e-06,
+      "loss": 0.7641,
+      "step": 4321
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5629697071476572,
+      "learning_rate": 8.488132432767483e-06,
+      "loss": 0.7924,
+      "step": 4322
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.791415464940472,
+      "learning_rate": 8.487389754516315e-06,
+      "loss": 0.7815,
+      "step": 4323
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5495018394704254,
+      "learning_rate": 8.486646926404243e-06,
+      "loss": 0.742,
+      "step": 4324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.323687624900751,
+      "learning_rate": 8.485903948463185e-06,
+      "loss": 0.8223,
+      "step": 4325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.764376277586376,
+      "learning_rate": 8.485160820725073e-06,
+      "loss": 0.7368,
+      "step": 4326
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7696763074653994,
+      "learning_rate": 8.484417543221839e-06,
+      "loss": 0.7366,
+      "step": 4327
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6145543961039392,
+      "learning_rate": 8.483674115985421e-06,
+      "loss": 0.8552,
+      "step": 4328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2438950380137237,
+      "learning_rate": 8.48293053904777e-06,
+      "loss": 0.581,
+      "step": 4329
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8120359353277709,
+      "learning_rate": 8.482186812440836e-06,
+      "loss": 0.7256,
+      "step": 4330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7256931336207113,
+      "learning_rate": 8.481442936196578e-06,
+      "loss": 0.7396,
+      "step": 4331
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7358127301673159,
+      "learning_rate": 8.480698910346965e-06,
+      "loss": 0.8174,
+      "step": 4332
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5947230787873332,
+      "learning_rate": 8.479954734923967e-06,
+      "loss": 0.6897,
+      "step": 4333
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4649422071333487,
+      "learning_rate": 8.479210409959565e-06,
+      "loss": 0.7265,
+      "step": 4334
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8812343171293695,
+      "learning_rate": 8.478465935485741e-06,
+      "loss": 0.7595,
+      "step": 4335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.754769678206916,
+      "learning_rate": 8.47772131153449e-06,
+      "loss": 0.7025,
+      "step": 4336
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6492036719124927,
+      "learning_rate": 8.476976538137809e-06,
+      "loss": 0.7973,
+      "step": 4337
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3736472412880705,
+      "learning_rate": 8.476231615327703e-06,
+      "loss": 0.5999,
+      "step": 4338
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.638358599827256,
+      "learning_rate": 8.475486543136181e-06,
+      "loss": 0.7477,
+      "step": 4339
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8080683509954663,
+      "learning_rate": 8.474741321595263e-06,
+      "loss": 0.8302,
+      "step": 4340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7478676327553844,
+      "learning_rate": 8.473995950736973e-06,
+      "loss": 0.674,
+      "step": 4341
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0179306532062062,
+      "learning_rate": 8.473250430593338e-06,
+      "loss": 0.5748,
+      "step": 4342
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6222924392339417,
+      "learning_rate": 8.472504761196397e-06,
+      "loss": 0.7067,
+      "step": 4343
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8668439528506358,
+      "learning_rate": 8.471758942578193e-06,
+      "loss": 0.714,
+      "step": 4344
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.365628106844699,
+      "learning_rate": 8.471012974770776e-06,
+      "loss": 0.7645,
+      "step": 4345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4199492308906017,
+      "learning_rate": 8.4702668578062e-06,
+      "loss": 0.7446,
+      "step": 4346
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8324120805271997,
+      "learning_rate": 8.46952059171653e-06,
+      "loss": 0.7877,
+      "step": 4347
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5915132790486124,
+      "learning_rate": 8.468774176533831e-06,
+      "loss": 0.7585,
+      "step": 4348
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8097604438990864,
+      "learning_rate": 8.46802761229018e-06,
+      "loss": 0.7243,
+      "step": 4349
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8513207743535867,
+      "learning_rate": 8.46728089901766e-06,
+      "loss": 0.7703,
+      "step": 4350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.561986653059071,
+      "learning_rate": 8.466534036748359e-06,
+      "loss": 0.6283,
+      "step": 4351
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4123670348853488,
+      "learning_rate": 8.465787025514368e-06,
+      "loss": 0.7005,
+      "step": 4352
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.030492664051349,
+      "learning_rate": 8.465039865347791e-06,
+      "loss": 0.821,
+      "step": 4353
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7910157773216655,
+      "learning_rate": 8.464292556280734e-06,
+      "loss": 0.7557,
+      "step": 4354
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.523009713479757,
+      "learning_rate": 8.463545098345308e-06,
+      "loss": 0.5972,
+      "step": 4355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7463740358709245,
+      "learning_rate": 8.462797491573638e-06,
+      "loss": 0.7902,
+      "step": 4356
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.591922513116588,
+      "learning_rate": 8.462049735997848e-06,
+      "loss": 0.7153,
+      "step": 4357
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8215361720811296,
+      "learning_rate": 8.46130183165007e-06,
+      "loss": 0.6945,
+      "step": 4358
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9475936070587638,
+      "learning_rate": 8.46055377856244e-06,
+      "loss": 0.7197,
+      "step": 4359
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8841255755761586,
+      "learning_rate": 8.459805576767111e-06,
+      "loss": 0.6988,
+      "step": 4360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.087974560733898,
+      "learning_rate": 8.459057226296232e-06,
+      "loss": 0.7798,
+      "step": 4361
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8962017459315956,
+      "learning_rate": 8.458308727181956e-06,
+      "loss": 0.7597,
+      "step": 4362
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.625164907645903,
+      "learning_rate": 8.457560079456455e-06,
+      "loss": 0.7404,
+      "step": 4363
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9456622105988437,
+      "learning_rate": 8.456811283151896e-06,
+      "loss": 0.7445,
+      "step": 4364
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4497821837544853,
+      "learning_rate": 8.456062338300458e-06,
+      "loss": 0.6877,
+      "step": 4365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.032443928776238,
+      "learning_rate": 8.455313244934324e-06,
+      "loss": 0.8435,
+      "step": 4366
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6753911775385015,
+      "learning_rate": 8.454564003085685e-06,
+      "loss": 0.7996,
+      "step": 4367
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6656638679532725,
+      "learning_rate": 8.453814612786736e-06,
+      "loss": 0.7065,
+      "step": 4368
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.075012346924941,
+      "learning_rate": 8.453065074069682e-06,
+      "loss": 0.7438,
+      "step": 4369
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2401517121151984,
+      "learning_rate": 8.45231538696673e-06,
+      "loss": 0.9297,
+      "step": 4370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7365983166292718,
+      "learning_rate": 8.451565551510097e-06,
+      "loss": 0.768,
+      "step": 4371
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2672860524020606,
+      "learning_rate": 8.450815567732007e-06,
+      "loss": 0.6423,
+      "step": 4372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6049679066087543,
+      "learning_rate": 8.450065435664686e-06,
+      "loss": 0.6182,
+      "step": 4373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6915321102281116,
+      "learning_rate": 8.449315155340369e-06,
+      "loss": 0.7752,
+      "step": 4374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0846551619550067,
+      "learning_rate": 8.4485647267913e-06,
+      "loss": 0.5557,
+      "step": 4375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.597148300935026,
+      "learning_rate": 8.44781415004972e-06,
+      "loss": 0.8045,
+      "step": 4376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7101221838497325,
+      "learning_rate": 8.447063425147891e-06,
+      "loss": 0.8715,
+      "step": 4377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7085932851753176,
+      "learning_rate": 8.446312552118068e-06,
+      "loss": 0.7906,
+      "step": 4378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6034442028348916,
+      "learning_rate": 8.44556153099252e-06,
+      "loss": 0.7473,
+      "step": 4379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7723308975537233,
+      "learning_rate": 8.44481036180352e-06,
+      "loss": 0.826,
+      "step": 4380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9322978407044178,
+      "learning_rate": 8.444059044583344e-06,
+      "loss": 0.7507,
+      "step": 4381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6155128313190417,
+      "learning_rate": 8.443307579364282e-06,
+      "loss": 0.7974,
+      "step": 4382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6015386102737506,
+      "learning_rate": 8.442555966178626e-06,
+      "loss": 0.6613,
+      "step": 4383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8098416592433075,
+      "learning_rate": 8.441804205058672e-06,
+      "loss": 0.8713,
+      "step": 4384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.754244812077981,
+      "learning_rate": 8.441052296036724e-06,
+      "loss": 0.7084,
+      "step": 4385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5729258538823065,
+      "learning_rate": 8.440300239145098e-06,
+      "loss": 0.7897,
+      "step": 4386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.674708142731406,
+      "learning_rate": 8.439548034416108e-06,
+      "loss": 0.8104,
+      "step": 4387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.631500984777853,
+      "learning_rate": 8.43879568188208e-06,
+      "loss": 0.8691,
+      "step": 4388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5877103460022335,
+      "learning_rate": 8.43804318157534e-06,
+      "loss": 0.5925,
+      "step": 4389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6937716875453885,
+      "learning_rate": 8.437290533528231e-06,
+      "loss": 0.6866,
+      "step": 4390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8155514596260358,
+      "learning_rate": 8.436537737773089e-06,
+      "loss": 0.7457,
+      "step": 4391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.408188688296035,
+      "learning_rate": 8.43578479434227e-06,
+      "loss": 0.6603,
+      "step": 4392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0850837471137322,
+      "learning_rate": 8.435031703268123e-06,
+      "loss": 0.6559,
+      "step": 4393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.020575744973154,
+      "learning_rate": 8.434278464583018e-06,
+      "loss": 0.8202,
+      "step": 4394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.830709186772871,
+      "learning_rate": 8.433525078319316e-06,
+      "loss": 0.888,
+      "step": 4395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7180080009384429,
+      "learning_rate": 8.432771544509395e-06,
+      "loss": 0.6731,
+      "step": 4396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8668946166589928,
+      "learning_rate": 8.432017863185635e-06,
+      "loss": 0.7342,
+      "step": 4397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1589604882426223,
+      "learning_rate": 8.431264034380424e-06,
+      "loss": 0.7007,
+      "step": 4398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6630266496425747,
+      "learning_rate": 8.430510058126156e-06,
+      "loss": 0.7841,
+      "step": 4399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7700710916113611,
+      "learning_rate": 8.42975593445523e-06,
+      "loss": 0.8464,
+      "step": 4400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.535450610514273,
+      "learning_rate": 8.429001663400054e-06,
+      "loss": 0.7544,
+      "step": 4401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6576216408090176,
+      "learning_rate": 8.428247244993038e-06,
+      "loss": 0.688,
+      "step": 4402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7733769513158237,
+      "learning_rate": 8.427492679266605e-06,
+      "loss": 0.7184,
+      "step": 4403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.186134655294516,
+      "learning_rate": 8.426737966253176e-06,
+      "loss": 0.6147,
+      "step": 4404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6624684050447665,
+      "learning_rate": 8.425983105985188e-06,
+      "loss": 0.6999,
+      "step": 4405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5791200593067316,
+      "learning_rate": 8.425228098495073e-06,
+      "loss": 0.7405,
+      "step": 4406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1131079027468103,
+      "learning_rate": 8.424472943815278e-06,
+      "loss": 0.6826,
+      "step": 4407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5661332711719071,
+      "learning_rate": 8.423717641978255e-06,
+      "loss": 0.7651,
+      "step": 4408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5715361673947086,
+      "learning_rate": 8.422962193016459e-06,
+      "loss": 0.6432,
+      "step": 4409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.559236312419136,
+      "learning_rate": 8.422206596962357e-06,
+      "loss": 0.6639,
+      "step": 4410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.654443103175172,
+      "learning_rate": 8.421450853848414e-06,
+      "loss": 0.6944,
+      "step": 4411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7545988498913196,
+      "learning_rate": 8.420694963707108e-06,
+      "loss": 0.9063,
+      "step": 4412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5911901658383878,
+      "learning_rate": 8.419938926570921e-06,
+      "loss": 0.7817,
+      "step": 4413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7939860943501833,
+      "learning_rate": 8.41918274247234e-06,
+      "loss": 0.9398,
+      "step": 4414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.454091733781156,
+      "learning_rate": 8.418426411443864e-06,
+      "loss": 0.7381,
+      "step": 4415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8056279075070356,
+      "learning_rate": 8.41766993351799e-06,
+      "loss": 0.7945,
+      "step": 4416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7107438089823466,
+      "learning_rate": 8.416913308727229e-06,
+      "loss": 0.6845,
+      "step": 4417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8129810983270378,
+      "learning_rate": 8.416156537104092e-06,
+      "loss": 0.7555,
+      "step": 4418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.588745416242504,
+      "learning_rate": 8.415399618681101e-06,
+      "loss": 0.8093,
+      "step": 4419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9601712518008636,
+      "learning_rate": 8.414642553490783e-06,
+      "loss": 0.7179,
+      "step": 4420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5201983831274346,
+      "learning_rate": 8.413885341565668e-06,
+      "loss": 0.816,
+      "step": 4421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2737227156279558,
+      "learning_rate": 8.413127982938299e-06,
+      "loss": 0.6032,
+      "step": 4422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7419445201132822,
+      "learning_rate": 8.412370477641215e-06,
+      "loss": 0.622,
+      "step": 4423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6845614028921103,
+      "learning_rate": 8.411612825706976e-06,
+      "loss": 0.7719,
+      "step": 4424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6765755449889685,
+      "learning_rate": 8.410855027168134e-06,
+      "loss": 0.7106,
+      "step": 4425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.91942143232303,
+      "learning_rate": 8.410097082057256e-06,
+      "loss": 0.7882,
+      "step": 4426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4496602233574958,
+      "learning_rate": 8.40933899040691e-06,
+      "loss": 0.6549,
+      "step": 4427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5447044240371246,
+      "learning_rate": 8.408580752249676e-06,
+      "loss": 0.7376,
+      "step": 4428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9883617636635436,
+      "learning_rate": 8.407822367618135e-06,
+      "loss": 0.7863,
+      "step": 4429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8897190024100516,
+      "learning_rate": 8.407063836544877e-06,
+      "loss": 0.8136,
+      "step": 4430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6620575665691624,
+      "learning_rate": 8.4063051590625e-06,
+      "loss": 0.7761,
+      "step": 4431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4465336490070086,
+      "learning_rate": 8.405546335203602e-06,
+      "loss": 0.6935,
+      "step": 4432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0010520334939823,
+      "learning_rate": 8.404787365000796e-06,
+      "loss": 0.8,
+      "step": 4433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5589776259614507,
+      "learning_rate": 8.40402824848669e-06,
+      "loss": 0.7625,
+      "step": 4434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5885684745135498,
+      "learning_rate": 8.403268985693912e-06,
+      "loss": 0.7058,
+      "step": 4435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8170410397914847,
+      "learning_rate": 8.402509576655085e-06,
+      "loss": 0.7138,
+      "step": 4436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6673352973753788,
+      "learning_rate": 8.401750021402843e-06,
+      "loss": 0.7136,
+      "step": 4437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9110527951479086,
+      "learning_rate": 8.400990319969829e-06,
+      "loss": 0.7372,
+      "step": 4438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.742558411599327,
+      "learning_rate": 8.400230472388684e-06,
+      "loss": 0.8147,
+      "step": 4439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0197254284774115,
+      "learning_rate": 8.399470478692064e-06,
+      "loss": 0.7257,
+      "step": 4440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5347536222078728,
+      "learning_rate": 8.398710338912626e-06,
+      "loss": 0.7904,
+      "step": 4441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.774380040251668,
+      "learning_rate": 8.397950053083036e-06,
+      "loss": 0.7764,
+      "step": 4442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1718101130482448,
+      "learning_rate": 8.397189621235964e-06,
+      "loss": 0.6579,
+      "step": 4443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.850240215597529,
+      "learning_rate": 8.396429043404088e-06,
+      "loss": 0.7263,
+      "step": 4444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4370763259802315,
+      "learning_rate": 8.395668319620092e-06,
+      "loss": 0.7938,
+      "step": 4445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9417163207288608,
+      "learning_rate": 8.394907449916667e-06,
+      "loss": 0.7459,
+      "step": 4446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6416315110849373,
+      "learning_rate": 8.394146434326507e-06,
+      "loss": 0.7614,
+      "step": 4447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.650819562519553,
+      "learning_rate": 8.393385272882315e-06,
+      "loss": 0.7994,
+      "step": 4448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8873040312118285,
+      "learning_rate": 8.3926239656168e-06,
+      "loss": 0.7979,
+      "step": 4449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8780621426223614,
+      "learning_rate": 8.391862512562679e-06,
+      "loss": 0.7121,
+      "step": 4450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7831115840922682,
+      "learning_rate": 8.39110091375267e-06,
+      "loss": 0.7946,
+      "step": 4451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1113008781276648,
+      "learning_rate": 8.390339169219504e-06,
+      "loss": 0.6384,
+      "step": 4452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1355573025545525,
+      "learning_rate": 8.389577278995913e-06,
+      "loss": 0.6086,
+      "step": 4453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6333802138262512,
+      "learning_rate": 8.388815243114637e-06,
+      "loss": 0.7906,
+      "step": 4454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8401770330938372,
+      "learning_rate": 8.388053061608421e-06,
+      "loss": 0.7346,
+      "step": 4455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4251342074975657,
+      "learning_rate": 8.387290734510022e-06,
+      "loss": 0.6881,
+      "step": 4456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9329117833800407,
+      "learning_rate": 8.386528261852196e-06,
+      "loss": 0.8392,
+      "step": 4457
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.478539178359957,
+      "learning_rate": 8.385765643667707e-06,
+      "loss": 0.6782,
+      "step": 4458
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.717404576707778,
+      "learning_rate": 8.385002879989328e-06,
+      "loss": 0.7581,
+      "step": 4459
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7509695726523535,
+      "learning_rate": 8.384239970849837e-06,
+      "loss": 0.6049,
+      "step": 4460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.272492323679003,
+      "learning_rate": 8.383476916282015e-06,
+      "loss": 0.784,
+      "step": 4461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.494829281360511,
+      "learning_rate": 8.382713716318658e-06,
+      "loss": 0.7721,
+      "step": 4462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5618637556460444,
+      "learning_rate": 8.381950370992557e-06,
+      "loss": 0.8176,
+      "step": 4463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3726907749534365,
+      "learning_rate": 8.381186880336518e-06,
+      "loss": 0.8224,
+      "step": 4464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6852352118139353,
+      "learning_rate": 8.380423244383347e-06,
+      "loss": 0.8818,
+      "step": 4465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.547970392900871,
+      "learning_rate": 8.37965946316586e-06,
+      "loss": 0.677,
+      "step": 4466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6996801376786974,
+      "learning_rate": 8.378895536716882e-06,
+      "loss": 0.8903,
+      "step": 4467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4911167251393316,
+      "learning_rate": 8.378131465069235e-06,
+      "loss": 0.6387,
+      "step": 4468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6340498211476324,
+      "learning_rate": 8.377367248255757e-06,
+      "loss": 0.7938,
+      "step": 4469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.116462943964901,
+      "learning_rate": 8.376602886309285e-06,
+      "loss": 0.5984,
+      "step": 4470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7141754245621137,
+      "learning_rate": 8.375838379262667e-06,
+      "loss": 0.7064,
+      "step": 4471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0359148859128497,
+      "learning_rate": 8.375073727148756e-06,
+      "loss": 0.7727,
+      "step": 4472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1604481223545426,
+      "learning_rate": 8.37430893000041e-06,
+      "loss": 0.8198,
+      "step": 4473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5261079615302269,
+      "learning_rate": 8.373543987850494e-06,
+      "loss": 0.8127,
+      "step": 4474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6147721460130482,
+      "learning_rate": 8.372778900731882e-06,
+      "loss": 0.7896,
+      "step": 4475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5083711445741836,
+      "learning_rate": 8.372013668677446e-06,
+      "loss": 0.5942,
+      "step": 4476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2359502225522676,
+      "learning_rate": 8.371248291720073e-06,
+      "loss": 0.7963,
+      "step": 4477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6684193156039309,
+      "learning_rate": 8.370482769892654e-06,
+      "loss": 0.6794,
+      "step": 4478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9907151862591924,
+      "learning_rate": 8.369717103228084e-06,
+      "loss": 0.5926,
+      "step": 4479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.908357973559347,
+      "learning_rate": 8.368951291759264e-06,
+      "loss": 0.7825,
+      "step": 4480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6474092354162049,
+      "learning_rate": 8.368185335519106e-06,
+      "loss": 0.8999,
+      "step": 4481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7634658937007608,
+      "learning_rate": 8.367419234540522e-06,
+      "loss": 0.8485,
+      "step": 4482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.570507076501418,
+      "learning_rate": 8.366652988856432e-06,
+      "loss": 0.6865,
+      "step": 4483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6857246458911779,
+      "learning_rate": 8.365886598499766e-06,
+      "loss": 0.7536,
+      "step": 4484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6071399432905513,
+      "learning_rate": 8.365120063503458e-06,
+      "loss": 0.8217,
+      "step": 4485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7861004995242158,
+      "learning_rate": 8.364353383900445e-06,
+      "loss": 0.7746,
+      "step": 4486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8234839214601204,
+      "learning_rate": 8.363586559723675e-06,
+      "loss": 0.6652,
+      "step": 4487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8982270510976735,
+      "learning_rate": 8.3628195910061e-06,
+      "loss": 0.8295,
+      "step": 4488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6127150442662403,
+      "learning_rate": 8.362052477780677e-06,
+      "loss": 0.7828,
+      "step": 4489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7820308791699941,
+      "learning_rate": 8.36128522008037e-06,
+      "loss": 0.7581,
+      "step": 4490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5851683745210536,
+      "learning_rate": 8.360517817938154e-06,
+      "loss": 0.713,
+      "step": 4491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.781690747011097,
+      "learning_rate": 8.359750271386999e-06,
+      "loss": 0.6564,
+      "step": 4492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.60739096866426,
+      "learning_rate": 8.358982580459896e-06,
+      "loss": 0.8768,
+      "step": 4493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6265332035285929,
+      "learning_rate": 8.35821474518983e-06,
+      "loss": 0.7511,
+      "step": 4494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8633471131704478,
+      "learning_rate": 8.357446765609796e-06,
+      "loss": 0.6974,
+      "step": 4495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7129971729836238,
+      "learning_rate": 8.356678641752797e-06,
+      "loss": 0.7352,
+      "step": 4496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6717516397072847,
+      "learning_rate": 8.355910373651844e-06,
+      "loss": 0.7026,
+      "step": 4497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7851544694587607,
+      "learning_rate": 8.355141961339945e-06,
+      "loss": 0.7463,
+      "step": 4498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7933570934666374,
+      "learning_rate": 8.354373404850124e-06,
+      "loss": 0.8238,
+      "step": 4499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.565649681301601,
+      "learning_rate": 8.353604704215408e-06,
+      "loss": 0.7053,
+      "step": 4500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6524938765389185,
+      "learning_rate": 8.352835859468829e-06,
+      "loss": 0.7899,
+      "step": 4501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6772222032049142,
+      "learning_rate": 8.352066870643424e-06,
+      "loss": 0.7353,
+      "step": 4502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.798776158629308,
+      "learning_rate": 8.351297737772244e-06,
+      "loss": 0.8364,
+      "step": 4503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.749159013427915,
+      "learning_rate": 8.350528460888334e-06,
+      "loss": 0.6822,
+      "step": 4504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.193929002207952,
+      "learning_rate": 8.349759040024753e-06,
+      "loss": 0.5792,
+      "step": 4505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4985218718763755,
+      "learning_rate": 8.348989475214568e-06,
+      "loss": 0.8579,
+      "step": 4506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6033329017510944,
+      "learning_rate": 8.348219766490845e-06,
+      "loss": 0.729,
+      "step": 4507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6885770421311865,
+      "learning_rate": 8.347449913886662e-06,
+      "loss": 0.6507,
+      "step": 4508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1855844855646247,
+      "learning_rate": 8.346679917435104e-06,
+      "loss": 0.6317,
+      "step": 4509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8091904582088472,
+      "learning_rate": 8.345909777169252e-06,
+      "loss": 0.8681,
+      "step": 4510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6954966281585593,
+      "learning_rate": 8.345139493122208e-06,
+      "loss": 0.8834,
+      "step": 4511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8703779613989326,
+      "learning_rate": 8.34436906532707e-06,
+      "loss": 0.7523,
+      "step": 4512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.738325438026822,
+      "learning_rate": 8.343598493816944e-06,
+      "loss": 0.779,
+      "step": 4513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.846240087953041,
+      "learning_rate": 8.342827778624943e-06,
+      "loss": 0.6483,
+      "step": 4514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.739011976259812,
+      "learning_rate": 8.34205691978419e-06,
+      "loss": 0.8201,
+      "step": 4515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.047984789091543,
+      "learning_rate": 8.341285917327807e-06,
+      "loss": 0.6266,
+      "step": 4516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6347306505410353,
+      "learning_rate": 8.340514771288926e-06,
+      "loss": 0.7396,
+      "step": 4517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9433912895214767,
+      "learning_rate": 8.339743481700685e-06,
+      "loss": 0.8005,
+      "step": 4518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5234073256517313,
+      "learning_rate": 8.33897204859623e-06,
+      "loss": 0.7623,
+      "step": 4519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.037671855285985,
+      "learning_rate": 8.338200472008708e-06,
+      "loss": 0.7899,
+      "step": 4520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6003856132831997,
+      "learning_rate": 8.337428751971279e-06,
+      "loss": 0.7342,
+      "step": 4521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7946319377067312,
+      "learning_rate": 8.336656888517103e-06,
+      "loss": 0.8659,
+      "step": 4522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5582314338863046,
+      "learning_rate": 8.33588488167935e-06,
+      "loss": 0.7415,
+      "step": 4523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.852288059133615,
+      "learning_rate": 8.335112731491192e-06,
+      "loss": 0.8063,
+      "step": 4524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.108391173721592,
+      "learning_rate": 8.334340437985814e-06,
+      "loss": 0.6549,
+      "step": 4525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5087374897832417,
+      "learning_rate": 8.333568001196402e-06,
+      "loss": 0.6183,
+      "step": 4526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.586918097099132,
+      "learning_rate": 8.332795421156147e-06,
+      "loss": 0.5753,
+      "step": 4527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6219781189418607,
+      "learning_rate": 8.332022697898253e-06,
+      "loss": 0.7909,
+      "step": 4528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.608393563132217,
+      "learning_rate": 8.331249831455921e-06,
+      "loss": 0.7639,
+      "step": 4529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8255723098041883,
+      "learning_rate": 8.330476821862366e-06,
+      "loss": 0.801,
+      "step": 4530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6330307535016628,
+      "learning_rate": 8.329703669150801e-06,
+      "loss": 0.6499,
+      "step": 4531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.84525719387495,
+      "learning_rate": 8.328930373354457e-06,
+      "loss": 0.7846,
+      "step": 4532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0460394866280514,
+      "learning_rate": 8.328156934506559e-06,
+      "loss": 0.6531,
+      "step": 4533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5588993047101785,
+      "learning_rate": 8.327383352640347e-06,
+      "loss": 0.6891,
+      "step": 4534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.518082987498205,
+      "learning_rate": 8.32660962778906e-06,
+      "loss": 0.7932,
+      "step": 4535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.242074075217815,
+      "learning_rate": 8.325835759985951e-06,
+      "loss": 0.7684,
+      "step": 4536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6209763289730743,
+      "learning_rate": 8.32506174926427e-06,
+      "loss": 0.8336,
+      "step": 4537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8344418525602304,
+      "learning_rate": 8.324287595657284e-06,
+      "loss": 0.7728,
+      "step": 4538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6165903466130245,
+      "learning_rate": 8.323513299198252e-06,
+      "loss": 0.7369,
+      "step": 4539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.068564524736894,
+      "learning_rate": 8.322738859920453e-06,
+      "loss": 0.8014,
+      "step": 4540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9049799977689774,
+      "learning_rate": 8.321964277857167e-06,
+      "loss": 0.8504,
+      "step": 4541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5520419802362644,
+      "learning_rate": 8.321189553041675e-06,
+      "loss": 0.8544,
+      "step": 4542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.507339884707921,
+      "learning_rate": 8.320414685507272e-06,
+      "loss": 0.9155,
+      "step": 4543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6047069906439466,
+      "learning_rate": 8.319639675287255e-06,
+      "loss": 0.5836,
+      "step": 4544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2364170134357018,
+      "learning_rate": 8.318864522414928e-06,
+      "loss": 0.5371,
+      "step": 4545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8846366728163808,
+      "learning_rate": 8.318089226923602e-06,
+      "loss": 0.7603,
+      "step": 4546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1995463378963394,
+      "learning_rate": 8.317313788846591e-06,
+      "loss": 0.553,
+      "step": 4547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8500907154905735,
+      "learning_rate": 8.31653820821722e-06,
+      "loss": 0.8198,
+      "step": 4548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.751440029097319,
+      "learning_rate": 8.315762485068815e-06,
+      "loss": 0.8529,
+      "step": 4549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9591307329162773,
+      "learning_rate": 8.31498661943471e-06,
+      "loss": 0.6685,
+      "step": 4550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7971591673825613,
+      "learning_rate": 8.314210611348249e-06,
+      "loss": 0.8224,
+      "step": 4551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5601508502926682,
+      "learning_rate": 8.313434460842775e-06,
+      "loss": 0.8462,
+      "step": 4552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.570963430962924,
+      "learning_rate": 8.312658167951644e-06,
+      "loss": 0.7931,
+      "step": 4553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6324120634082893,
+      "learning_rate": 8.311881732708213e-06,
+      "loss": 0.7833,
+      "step": 4554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2051385811537676,
+      "learning_rate": 8.311105155145849e-06,
+      "loss": 0.7787,
+      "step": 4555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9620273745897787,
+      "learning_rate": 8.310328435297923e-06,
+      "loss": 0.6586,
+      "step": 4556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9568844184551473,
+      "learning_rate": 8.309551573197809e-06,
+      "loss": 0.7604,
+      "step": 4557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2644340028636265,
+      "learning_rate": 8.308774568878896e-06,
+      "loss": 0.6835,
+      "step": 4558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.543051984718732,
+      "learning_rate": 8.307997422374569e-06,
+      "loss": 0.8199,
+      "step": 4559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7725056150843985,
+      "learning_rate": 8.307220133718225e-06,
+      "loss": 0.8532,
+      "step": 4560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7351279214010733,
+      "learning_rate": 8.306442702943268e-06,
+      "loss": 0.7618,
+      "step": 4561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8463254051299536,
+      "learning_rate": 8.305665130083102e-06,
+      "loss": 0.7893,
+      "step": 4562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9953583582553225,
+      "learning_rate": 8.304887415171146e-06,
+      "loss": 0.7208,
+      "step": 4563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5798928001495047,
+      "learning_rate": 8.304109558240817e-06,
+      "loss": 0.6758,
+      "step": 4564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7423394779086776,
+      "learning_rate": 8.30333155932554e-06,
+      "loss": 0.8096,
+      "step": 4565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6932991237962243,
+      "learning_rate": 8.302553418458749e-06,
+      "loss": 0.6672,
+      "step": 4566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.673395440692939,
+      "learning_rate": 8.301775135673884e-06,
+      "loss": 0.7188,
+      "step": 4567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.989616554893317,
+      "learning_rate": 8.300996711004387e-06,
+      "loss": 0.7006,
+      "step": 4568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1276676173476083,
+      "learning_rate": 8.300218144483709e-06,
+      "loss": 0.6937,
+      "step": 4569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0550437743903756,
+      "learning_rate": 8.299439436145309e-06,
+      "loss": 0.6725,
+      "step": 4570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.713981371406246,
+      "learning_rate": 8.298660586022646e-06,
+      "loss": 0.734,
+      "step": 4571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6060482838165042,
+      "learning_rate": 8.297881594149193e-06,
+      "loss": 0.7927,
+      "step": 4572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8244794498569625,
+      "learning_rate": 8.297102460558421e-06,
+      "loss": 0.7515,
+      "step": 4573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8118785234403445,
+      "learning_rate": 8.296323185283816e-06,
+      "loss": 0.7916,
+      "step": 4574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.518296662300008,
+      "learning_rate": 8.29554376835886e-06,
+      "loss": 0.6735,
+      "step": 4575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7803715657549957,
+      "learning_rate": 8.29476420981705e-06,
+      "loss": 0.939,
+      "step": 4576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.533872882503958,
+      "learning_rate": 8.293984509691885e-06,
+      "loss": 0.8638,
+      "step": 4577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7423535866746005,
+      "learning_rate": 8.293204668016867e-06,
+      "loss": 0.8197,
+      "step": 4578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0632570138433262,
+      "learning_rate": 8.292424684825514e-06,
+      "loss": 0.7442,
+      "step": 4579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2962398082369953,
+      "learning_rate": 8.291644560151335e-06,
+      "loss": 0.5646,
+      "step": 4580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6715225091303634,
+      "learning_rate": 8.29086429402786e-06,
+      "loss": 0.761,
+      "step": 4581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.616490930534518,
+      "learning_rate": 8.290083886488618e-06,
+      "loss": 0.7548,
+      "step": 4582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6254373528155588,
+      "learning_rate": 8.289303337567145e-06,
+      "loss": 0.7537,
+      "step": 4583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2453327067804767,
+      "learning_rate": 8.28852264729698e-06,
+      "loss": 0.7988,
+      "step": 4584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.728446038039224,
+      "learning_rate": 8.287741815711674e-06,
+      "loss": 0.8828,
+      "step": 4585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6656177062924218,
+      "learning_rate": 8.286960842844779e-06,
+      "loss": 0.7363,
+      "step": 4586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7493758700762783,
+      "learning_rate": 8.286179728729856e-06,
+      "loss": 0.7625,
+      "step": 4587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9380766571442305,
+      "learning_rate": 8.285398473400471e-06,
+      "loss": 0.7616,
+      "step": 4588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5589243201290215,
+      "learning_rate": 8.284617076890199e-06,
+      "loss": 0.7612,
+      "step": 4589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7728741660482041,
+      "learning_rate": 8.283835539232614e-06,
+      "loss": 0.7904,
+      "step": 4590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5645373169879238,
+      "learning_rate": 8.283053860461303e-06,
+      "loss": 0.7915,
+      "step": 4591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9209926214374822,
+      "learning_rate": 8.282272040609855e-06,
+      "loss": 0.7214,
+      "step": 4592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1202265878103628,
+      "learning_rate": 8.281490079711866e-06,
+      "loss": 0.6578,
+      "step": 4593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.476978685125889,
+      "learning_rate": 8.280707977800944e-06,
+      "loss": 1.0106,
+      "step": 4594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6251926101820011,
+      "learning_rate": 8.279925734910691e-06,
+      "loss": 0.6983,
+      "step": 4595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4747907831909548,
+      "learning_rate": 8.279143351074726e-06,
+      "loss": 0.7318,
+      "step": 4596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6559682576855599,
+      "learning_rate": 8.278360826326669e-06,
+      "loss": 0.7561,
+      "step": 4597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.487504046928779,
+      "learning_rate": 8.277578160700143e-06,
+      "loss": 0.7104,
+      "step": 4598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3528593283665042,
+      "learning_rate": 8.276795354228785e-06,
+      "loss": 0.6612,
+      "step": 4599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9332594413168958,
+      "learning_rate": 8.276012406946232e-06,
+      "loss": 0.6939,
+      "step": 4600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.723450198166268,
+      "learning_rate": 8.275229318886132e-06,
+      "loss": 0.6994,
+      "step": 4601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1832359010635147,
+      "learning_rate": 8.274446090082133e-06,
+      "loss": 0.7702,
+      "step": 4602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0335259641148942,
+      "learning_rate": 8.273662720567892e-06,
+      "loss": 0.5897,
+      "step": 4603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9679175391851385,
+      "learning_rate": 8.272879210377074e-06,
+      "loss": 0.7418,
+      "step": 4604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9044654588196364,
+      "learning_rate": 8.272095559543346e-06,
+      "loss": 0.6888,
+      "step": 4605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.647955176621358,
+      "learning_rate": 8.271311768100386e-06,
+      "loss": 0.7032,
+      "step": 4606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.812139472250341,
+      "learning_rate": 8.270527836081872e-06,
+      "loss": 0.7922,
+      "step": 4607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8615089648333816,
+      "learning_rate": 8.269743763521495e-06,
+      "loss": 0.7699,
+      "step": 4608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5436346759825612,
+      "learning_rate": 8.268959550452946e-06,
+      "loss": 0.6959,
+      "step": 4609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.682549902341748,
+      "learning_rate": 8.268175196909924e-06,
+      "loss": 0.7714,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9586516786462824,
+      "learning_rate": 8.267390702926135e-06,
+      "loss": 0.8197,
+      "step": 4611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6795556985563476,
+      "learning_rate": 8.26660606853529e-06,
+      "loss": 0.7876,
+      "step": 4612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.119480742355903,
+      "learning_rate": 8.26582129377111e-06,
+      "loss": 0.6516,
+      "step": 4613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3362766552453658,
+      "learning_rate": 8.265036378667312e-06,
+      "loss": 0.7758,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2739495370141498,
+      "learning_rate": 8.264251323257633e-06,
+      "loss": 0.7362,
+      "step": 4615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.18919669030536,
+      "learning_rate": 8.263466127575801e-06,
+      "loss": 0.7684,
+      "step": 4616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5522958102176916,
+      "learning_rate": 8.262680791655565e-06,
+      "loss": 0.8526,
+      "step": 4617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6830987381242293,
+      "learning_rate": 8.261895315530667e-06,
+      "loss": 0.7598,
+      "step": 4618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.357841543764718,
+      "learning_rate": 8.261109699234862e-06,
+      "loss": 0.6775,
+      "step": 4619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7193317847700293,
+      "learning_rate": 8.26032394280191e-06,
+      "loss": 0.7156,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6541342112310766,
+      "learning_rate": 8.259538046265578e-06,
+      "loss": 0.6746,
+      "step": 4621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8469893570446863,
+      "learning_rate": 8.258752009659638e-06,
+      "loss": 0.7752,
+      "step": 4622
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5688779977359146,
+      "learning_rate": 8.257965833017864e-06,
+      "loss": 0.7734,
+      "step": 4623
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.138988695713818,
+      "learning_rate": 8.257179516374045e-06,
+      "loss": 0.8225,
+      "step": 4624
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.782137822320572,
+      "learning_rate": 8.256393059761966e-06,
+      "loss": 0.7751,
+      "step": 4625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1979695175034077,
+      "learning_rate": 8.255606463215426e-06,
+      "loss": 0.7735,
+      "step": 4626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.724431185748028,
+      "learning_rate": 8.254819726768224e-06,
+      "loss": 0.6937,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5360829415810622,
+      "learning_rate": 8.254032850454173e-06,
+      "loss": 0.8052,
+      "step": 4628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8564023366617157,
+      "learning_rate": 8.253245834307079e-06,
+      "loss": 0.7685,
+      "step": 4629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1825034760940822,
+      "learning_rate": 8.252458678360769e-06,
+      "loss": 0.7111,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5215511931750496,
+      "learning_rate": 8.251671382649065e-06,
+      "loss": 0.7666,
+      "step": 4631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8166853931037856,
+      "learning_rate": 8.250883947205799e-06,
+      "loss": 0.7203,
+      "step": 4632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6404882249537134,
+      "learning_rate": 8.25009637206481e-06,
+      "loss": 0.8573,
+      "step": 4633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8655589307206035,
+      "learning_rate": 8.249308657259943e-06,
+      "loss": 0.7897,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2433848007400687,
+      "learning_rate": 8.248520802825046e-06,
+      "loss": 0.7157,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6844882566907895,
+      "learning_rate": 8.247732808793975e-06,
+      "loss": 0.7782,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6027657280503536,
+      "learning_rate": 8.246944675200592e-06,
+      "loss": 0.7762,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5401337823983499,
+      "learning_rate": 8.246156402078765e-06,
+      "loss": 0.7141,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6497167503714742,
+      "learning_rate": 8.245367989462368e-06,
+      "loss": 0.7696,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6727091920643289,
+      "learning_rate": 8.244579437385281e-06,
+      "loss": 0.8285,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8091707877554872,
+      "learning_rate": 8.243790745881389e-06,
+      "loss": 0.8062,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.064615157868781,
+      "learning_rate": 8.243001914984583e-06,
+      "loss": 0.8097,
+      "step": 4642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6898688041830185,
+      "learning_rate": 8.242212944728762e-06,
+      "loss": 0.6906,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6821877136456063,
+      "learning_rate": 8.241423835147833e-06,
+      "loss": 0.81,
+      "step": 4644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5585208248116098,
+      "learning_rate": 8.240634586275701e-06,
+      "loss": 0.635,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7093851954487598,
+      "learning_rate": 8.239845198146284e-06,
+      "loss": 0.7249,
+      "step": 4646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0902028304759326,
+      "learning_rate": 8.239055670793503e-06,
+      "loss": 0.7357,
+      "step": 4647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.095752379575049,
+      "learning_rate": 8.238266004251284e-06,
+      "loss": 0.7294,
+      "step": 4648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.485205762985859,
+      "learning_rate": 8.237476198553567e-06,
+      "loss": 0.8059,
+      "step": 4649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.282292918516944,
+      "learning_rate": 8.236686253734285e-06,
+      "loss": 0.8764,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.209279452087081,
+      "learning_rate": 8.235896169827386e-06,
+      "loss": 0.6584,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6679072211210253,
+      "learning_rate": 8.235105946866823e-06,
+      "loss": 0.6734,
+      "step": 4652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1952705993441066,
+      "learning_rate": 8.234315584886551e-06,
+      "loss": 0.6367,
+      "step": 4653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1936145607807462,
+      "learning_rate": 8.233525083920536e-06,
+      "loss": 0.606,
+      "step": 4654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8941954376546872,
+      "learning_rate": 8.232734444002748e-06,
+      "loss": 0.7639,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6548337178317685,
+      "learning_rate": 8.23194366516716e-06,
+      "loss": 0.7995,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6365179497118887,
+      "learning_rate": 8.231152747447753e-06,
+      "loss": 0.7215,
+      "step": 4657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6137350390084813,
+      "learning_rate": 8.230361690878519e-06,
+      "loss": 0.7184,
+      "step": 4658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7772207538673335,
+      "learning_rate": 8.229570495493447e-06,
+      "loss": 0.7165,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5298319130009599,
+      "learning_rate": 8.22877916132654e-06,
+      "loss": 0.6808,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.258251908119367,
+      "learning_rate": 8.2279876884118e-06,
+      "loss": 0.67,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.00995044155686,
+      "learning_rate": 8.22719607678324e-06,
+      "loss": 0.7028,
+      "step": 4662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6394051516246506,
+      "learning_rate": 8.226404326474878e-06,
+      "loss": 0.7439,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.184695682410509,
+      "learning_rate": 8.225612437520736e-06,
+      "loss": 0.6629,
+      "step": 4664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7847575167235206,
+      "learning_rate": 8.224820409954844e-06,
+      "loss": 0.8607,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.903858973403514,
+      "learning_rate": 8.224028243811237e-06,
+      "loss": 0.8407,
+      "step": 4666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.776968954091543,
+      "learning_rate": 8.223235939123955e-06,
+      "loss": 0.7408,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7121756191942472,
+      "learning_rate": 8.222443495927049e-06,
+      "loss": 0.8307,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7517840054182217,
+      "learning_rate": 8.221650914254566e-06,
+      "loss": 0.726,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.776760510046748,
+      "learning_rate": 8.220858194140568e-06,
+      "loss": 0.8286,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4758871950952306,
+      "learning_rate": 8.220065335619123e-06,
+      "loss": 0.8072,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7599881255596193,
+      "learning_rate": 8.219272338724297e-06,
+      "loss": 0.7853,
+      "step": 4672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7366879290105954,
+      "learning_rate": 8.21847920349017e-06,
+      "loss": 0.8053,
+      "step": 4673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.773055364621588,
+      "learning_rate": 8.217685929950823e-06,
+      "loss": 0.8015,
+      "step": 4674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1049082660767078,
+      "learning_rate": 8.216892518140346e-06,
+      "loss": 0.6632,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.725008624084496,
+      "learning_rate": 8.216098968092833e-06,
+      "loss": 0.7392,
+      "step": 4676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7972350608843761,
+      "learning_rate": 8.215305279842385e-06,
+      "loss": 0.7768,
+      "step": 4677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9938485072320052,
+      "learning_rate": 8.214511453423107e-06,
+      "loss": 0.5524,
+      "step": 4678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6582153225451224,
+      "learning_rate": 8.213717488869113e-06,
+      "loss": 0.7074,
+      "step": 4679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.762467313898682,
+      "learning_rate": 8.212923386214522e-06,
+      "loss": 0.7173,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7145888118320922,
+      "learning_rate": 8.212129145493457e-06,
+      "loss": 0.7352,
+      "step": 4681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.787562549356914,
+      "learning_rate": 8.211334766740048e-06,
+      "loss": 0.8351,
+      "step": 4682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6659716111000202,
+      "learning_rate": 8.210540249988435e-06,
+      "loss": 0.74,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7544838649891956,
+      "learning_rate": 8.209745595272755e-06,
+      "loss": 0.7643,
+      "step": 4684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.75051549418668,
+      "learning_rate": 8.208950802627157e-06,
+      "loss": 0.882,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.022610796996644,
+      "learning_rate": 8.208155872085799e-06,
+      "loss": 0.7254,
+      "step": 4686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6003636278763769,
+      "learning_rate": 8.207360803682837e-06,
+      "loss": 0.7938,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4782562639850945,
+      "learning_rate": 8.206565597452438e-06,
+      "loss": 0.8272,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.191099868410819,
+      "learning_rate": 8.205770253428775e-06,
+      "loss": 0.8548,
+      "step": 4689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.721264133659505,
+      "learning_rate": 8.204974771646023e-06,
+      "loss": 0.6301,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.740995244660774,
+      "learning_rate": 8.204179152138367e-06,
+      "loss": 0.754,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.548385416126818,
+      "learning_rate": 8.203383394939998e-06,
+      "loss": 0.6941,
+      "step": 4692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6750980081932059,
+      "learning_rate": 8.20258750008511e-06,
+      "loss": 0.7482,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5239911088629767,
+      "learning_rate": 8.201791467607905e-06,
+      "loss": 0.7025,
+      "step": 4694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6471405086037478,
+      "learning_rate": 8.200995297542589e-06,
+      "loss": 0.8076,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8488696005761847,
+      "learning_rate": 8.200198989923376e-06,
+      "loss": 0.8433,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6829093932447323,
+      "learning_rate": 8.199402544784485e-06,
+      "loss": 0.721,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7298265794981214,
+      "learning_rate": 8.198605962160143e-06,
+      "loss": 0.6393,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3873936221964498,
+      "learning_rate": 8.197809242084575e-06,
+      "loss": 0.6408,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6325139746388297,
+      "learning_rate": 8.197012384592027e-06,
+      "loss": 0.7144,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.94109776449095,
+      "learning_rate": 8.196215389716732e-06,
+      "loss": 0.8285,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2595610555916357,
+      "learning_rate": 8.195418257492948e-06,
+      "loss": 0.6457,
+      "step": 4702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.997192110883261,
+      "learning_rate": 8.194620987954922e-06,
+      "loss": 0.8665,
+      "step": 4703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6361511121699426,
+      "learning_rate": 8.193823581136919e-06,
+      "loss": 0.7914,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1466544735470134,
+      "learning_rate": 8.193026037073201e-06,
+      "loss": 0.7196,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8331611413139892,
+      "learning_rate": 8.192228355798045e-06,
+      "loss": 0.8027,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.809495148587336,
+      "learning_rate": 8.191430537345728e-06,
+      "loss": 0.7555,
+      "step": 4707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1249104798849898,
+      "learning_rate": 8.19063258175053e-06,
+      "loss": 0.5238,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7203049050520458,
+      "learning_rate": 8.189834489046746e-06,
+      "loss": 0.7057,
+      "step": 4709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7248156401800117,
+      "learning_rate": 8.18903625926867e-06,
+      "loss": 0.7895,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8704136619236702,
+      "learning_rate": 8.188237892450603e-06,
+      "loss": 0.6692,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9176545279355648,
+      "learning_rate": 8.187439388626855e-06,
+      "loss": 0.6444,
+      "step": 4712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8272913342386572,
+      "learning_rate": 8.186640747831735e-06,
+      "loss": 0.7511,
+      "step": 4713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2044598833548872,
+      "learning_rate": 8.185841970099566e-06,
+      "loss": 0.8213,
+      "step": 4714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6836969755314197,
+      "learning_rate": 8.185043055464673e-06,
+      "loss": 0.7722,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.14415456489064,
+      "learning_rate": 8.184244003961386e-06,
+      "loss": 0.6817,
+      "step": 4716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1913343456548278,
+      "learning_rate": 8.183444815624041e-06,
+      "loss": 0.8432,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5755620801582875,
+      "learning_rate": 8.182645490486986e-06,
+      "loss": 0.7734,
+      "step": 4718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5517825573206334,
+      "learning_rate": 8.181846028584563e-06,
+      "loss": 0.839,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6466745686479518,
+      "learning_rate": 8.181046429951131e-06,
+      "loss": 0.9337,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0378223440010101,
+      "learning_rate": 8.180246694621048e-06,
+      "loss": 0.5961,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7607408386766144,
+      "learning_rate": 8.179446822628684e-06,
+      "loss": 0.703,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5521687197400686,
+      "learning_rate": 8.178646814008407e-06,
+      "loss": 0.7175,
+      "step": 4723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7759483880193265,
+      "learning_rate": 8.177846668794598e-06,
+      "loss": 0.7817,
+      "step": 4724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8093372680153381,
+      "learning_rate": 8.177046387021641e-06,
+      "loss": 0.7532,
+      "step": 4725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7085482814483313,
+      "learning_rate": 8.176245968723924e-06,
+      "loss": 0.8245,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0394520958046203,
+      "learning_rate": 8.175445413935844e-06,
+      "loss": 0.6764,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8465777210229457,
+      "learning_rate": 8.174644722691802e-06,
+      "loss": 0.7387,
+      "step": 4728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.899570448448205,
+      "learning_rate": 8.173843895026207e-06,
+      "loss": 0.8005,
+      "step": 4729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6353499529349138,
+      "learning_rate": 8.17304293097347e-06,
+      "loss": 0.811,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.733501920985293,
+      "learning_rate": 8.172241830568012e-06,
+      "loss": 0.7489,
+      "step": 4731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8032927531509446,
+      "learning_rate": 8.17144059384426e-06,
+      "loss": 0.742,
+      "step": 4732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.576809088208534,
+      "learning_rate": 8.170639220836639e-06,
+      "loss": 0.7109,
+      "step": 4733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1777927830723556,
+      "learning_rate": 8.169837711579591e-06,
+      "loss": 0.6073,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.534643115617676,
+      "learning_rate": 8.169036066107556e-06,
+      "loss": 0.7312,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6232311822001821,
+      "learning_rate": 8.168234284454984e-06,
+      "loss": 0.6506,
+      "step": 4736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6578728796564894,
+      "learning_rate": 8.16743236665633e-06,
+      "loss": 0.7185,
+      "step": 4737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5835681646954118,
+      "learning_rate": 8.166630312746051e-06,
+      "loss": 0.7692,
+      "step": 4738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1500080310931406,
+      "learning_rate": 8.165828122758615e-06,
+      "loss": 0.7974,
+      "step": 4739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.54642274294998,
+      "learning_rate": 8.165025796728495e-06,
+      "loss": 0.8104,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4470458317276578,
+      "learning_rate": 8.164223334690168e-06,
+      "loss": 0.7655,
+      "step": 4741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.715084619048964,
+      "learning_rate": 8.163420736678115e-06,
+      "loss": 0.8123,
+      "step": 4742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0221807805866696,
+      "learning_rate": 8.162618002726829e-06,
+      "loss": 0.7795,
+      "step": 4743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4730703567091665,
+      "learning_rate": 8.161815132870806e-06,
+      "loss": 0.8492,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8303929419796252,
+      "learning_rate": 8.161012127144543e-06,
+      "loss": 0.7688,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5792887045552235,
+      "learning_rate": 8.160208985582547e-06,
+      "loss": 0.7911,
+      "step": 4746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6143738285951856,
+      "learning_rate": 8.159405708219335e-06,
+      "loss": 0.709,
+      "step": 4747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7846905757260576,
+      "learning_rate": 8.158602295089423e-06,
+      "loss": 0.6952,
+      "step": 4748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6880024928059063,
+      "learning_rate": 8.157798746227337e-06,
+      "loss": 0.8741,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8370712748198896,
+      "learning_rate": 8.156995061667605e-06,
+      "loss": 0.7964,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8647678174307822,
+      "learning_rate": 8.156191241444763e-06,
+      "loss": 0.6887,
+      "step": 4751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8656885016990161,
+      "learning_rate": 8.155387285593356e-06,
+      "loss": 0.9021,
+      "step": 4752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1369759773436383,
+      "learning_rate": 8.154583194147929e-06,
+      "loss": 0.6403,
+      "step": 4753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8806223875960724,
+      "learning_rate": 8.153778967143035e-06,
+      "loss": 0.7769,
+      "step": 4754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4413634715221706,
+      "learning_rate": 8.152974604613238e-06,
+      "loss": 0.7577,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0094309828026256,
+      "learning_rate": 8.152170106593099e-06,
+      "loss": 0.6038,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6342445135613342,
+      "learning_rate": 8.15136547311719e-06,
+      "loss": 0.7,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.402658755104219,
+      "learning_rate": 8.150560704220087e-06,
+      "loss": 0.7878,
+      "step": 4758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5342217935232356,
+      "learning_rate": 8.149755799936377e-06,
+      "loss": 0.753,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.665360431396829,
+      "learning_rate": 8.148950760300642e-06,
+      "loss": 0.6872,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.085188949667242,
+      "learning_rate": 8.148145585347482e-06,
+      "loss": 0.6371,
+      "step": 4761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8717868427400515,
+      "learning_rate": 8.147340275111492e-06,
+      "loss": 0.7135,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5601155154626987,
+      "learning_rate": 8.146534829627285e-06,
+      "loss": 0.6553,
+      "step": 4763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.75927533190347,
+      "learning_rate": 8.145729248929466e-06,
+      "loss": 0.7608,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0760395407425078,
+      "learning_rate": 8.144923533052655e-06,
+      "loss": 0.6969,
+      "step": 4765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.76285122736149,
+      "learning_rate": 8.144117682031478e-06,
+      "loss": 0.7985,
+      "step": 4766
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9834025003348863,
+      "learning_rate": 8.14331169590056e-06,
+      "loss": 0.7883,
+      "step": 4767
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2767138236035815,
+      "learning_rate": 8.142505574694539e-06,
+      "loss": 0.7127,
+      "step": 4768
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1310902322952212,
+      "learning_rate": 8.141699318448053e-06,
+      "loss": 0.7189,
+      "step": 4769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4003872306476308,
+      "learning_rate": 8.14089292719575e-06,
+      "loss": 0.6828,
+      "step": 4770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8489724840722377,
+      "learning_rate": 8.140086400972285e-06,
+      "loss": 0.8118,
+      "step": 4771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.739169296732448,
+      "learning_rate": 8.139279739812314e-06,
+      "loss": 0.778,
+      "step": 4772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8614808400734346,
+      "learning_rate": 8.1384729437505e-06,
+      "loss": 0.7577,
+      "step": 4773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6734149361983444,
+      "learning_rate": 8.137666012821514e-06,
+      "loss": 0.8763,
+      "step": 4774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6151689349851925,
+      "learning_rate": 8.136858947060032e-06,
+      "loss": 0.7157,
+      "step": 4775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6701291391727207,
+      "learning_rate": 8.136051746500735e-06,
+      "loss": 0.6927,
+      "step": 4776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0821063592450235,
+      "learning_rate": 8.135244411178309e-06,
+      "loss": 0.5435,
+      "step": 4777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6240278607587368,
+      "learning_rate": 8.134436941127448e-06,
+      "loss": 0.816,
+      "step": 4778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2955791922214501,
+      "learning_rate": 8.13362933638285e-06,
+      "loss": 0.7466,
+      "step": 4779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.754791306184478,
+      "learning_rate": 8.132821596979223e-06,
+      "loss": 0.8188,
+      "step": 4780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9648460694571455,
+      "learning_rate": 8.132013722951275e-06,
+      "loss": 0.7064,
+      "step": 4781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.063588368328651,
+      "learning_rate": 8.13120571433372e-06,
+      "loss": 0.699,
+      "step": 4782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6472833216331526,
+      "learning_rate": 8.130397571161283e-06,
+      "loss": 0.6802,
+      "step": 4783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.773157408995865,
+      "learning_rate": 8.129589293468689e-06,
+      "loss": 0.7625,
+      "step": 4784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6306474784804226,
+      "learning_rate": 8.128780881290674e-06,
+      "loss": 0.7608,
+      "step": 4785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2797865781254503,
+      "learning_rate": 8.127972334661978e-06,
+      "loss": 0.6444,
+      "step": 4786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8897726725196105,
+      "learning_rate": 8.127163653617342e-06,
+      "loss": 0.837,
+      "step": 4787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6526930940824505,
+      "learning_rate": 8.12635483819152e-06,
+      "loss": 0.7836,
+      "step": 4788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6537358750610451,
+      "learning_rate": 8.125545888419269e-06,
+      "loss": 0.75,
+      "step": 4789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4053482170159834,
+      "learning_rate": 8.124736804335348e-06,
+      "loss": 0.6037,
+      "step": 4790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2342863624903646,
+      "learning_rate": 8.123927585974529e-06,
+      "loss": 0.6266,
+      "step": 4791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7498420541124728,
+      "learning_rate": 8.123118233371585e-06,
+      "loss": 0.7913,
+      "step": 4792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6873152275191055,
+      "learning_rate": 8.122308746561294e-06,
+      "loss": 0.7182,
+      "step": 4793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.015540953945574,
+      "learning_rate": 8.12149912557844e-06,
+      "loss": 0.7209,
+      "step": 4794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5080791806323297,
+      "learning_rate": 8.12068937045782e-06,
+      "loss": 0.6772,
+      "step": 4795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0856018075400837,
+      "learning_rate": 8.119879481234228e-06,
+      "loss": 0.6781,
+      "step": 4796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7682158705162394,
+      "learning_rate": 8.119069457942464e-06,
+      "loss": 0.7632,
+      "step": 4797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5439900113725542,
+      "learning_rate": 8.118259300617339e-06,
+      "loss": 0.7578,
+      "step": 4798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5460450832099497,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.7659,
+      "step": 4799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8868927192279767,
+      "learning_rate": 8.11663858400627e-06,
+      "loss": 0.7567,
+      "step": 4800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0399772818610997,
+      "learning_rate": 8.115828024789972e-06,
+      "loss": 0.6904,
+      "step": 4801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7771639624601014,
+      "learning_rate": 8.115017331679602e-06,
+      "loss": 0.7962,
+      "step": 4802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6222414838526347,
+      "learning_rate": 8.114206504710002e-06,
+      "loss": 0.8756,
+      "step": 4803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4655420051122263,
+      "learning_rate": 8.113395543916012e-06,
+      "loss": 0.6623,
+      "step": 4804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5771444933398633,
+      "learning_rate": 8.112584449332481e-06,
+      "loss": 0.8384,
+      "step": 4805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.640910991208986,
+      "learning_rate": 8.111773220994264e-06,
+      "loss": 0.7838,
+      "step": 4806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2149058852729289,
+      "learning_rate": 8.110961858936222e-06,
+      "loss": 0.5686,
+      "step": 4807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0576242884634344,
+      "learning_rate": 8.11015036319322e-06,
+      "loss": 0.7227,
+      "step": 4808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2569596299409094,
+      "learning_rate": 8.109338733800132e-06,
+      "loss": 0.687,
+      "step": 4809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7526853210854363,
+      "learning_rate": 8.108526970791832e-06,
+      "loss": 0.8189,
+      "step": 4810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6048127391196552,
+      "learning_rate": 8.107715074203205e-06,
+      "loss": 0.8113,
+      "step": 4811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6337895023226576,
+      "learning_rate": 8.10690304406914e-06,
+      "loss": 0.8518,
+      "step": 4812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7445631848963719,
+      "learning_rate": 8.106090880424534e-06,
+      "loss": 0.7157,
+      "step": 4813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.635439039272422,
+      "learning_rate": 8.10527858330428e-06,
+      "loss": 0.7731,
+      "step": 4814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5379436111229152,
+      "learning_rate": 8.104466152743295e-06,
+      "loss": 0.7572,
+      "step": 4815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9357527081398165,
+      "learning_rate": 8.103653588776483e-06,
+      "loss": 0.6631,
+      "step": 4816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7082381140065426,
+      "learning_rate": 8.102840891438764e-06,
+      "loss": 0.5616,
+      "step": 4817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9834773284167149,
+      "learning_rate": 8.10202806076506e-06,
+      "loss": 0.8494,
+      "step": 4818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.788379362236626,
+      "learning_rate": 8.101215096790305e-06,
+      "loss": 0.713,
+      "step": 4819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6803224123223242,
+      "learning_rate": 8.10040199954943e-06,
+      "loss": 0.7598,
+      "step": 4820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7961599118749354,
+      "learning_rate": 8.099588769077374e-06,
+      "loss": 0.7209,
+      "step": 4821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.652106499752561,
+      "learning_rate": 8.098775405409087e-06,
+      "loss": 0.859,
+      "step": 4822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5591536277511162,
+      "learning_rate": 8.09796190857952e-06,
+      "loss": 0.7257,
+      "step": 4823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6147855373688411,
+      "learning_rate": 8.097148278623628e-06,
+      "loss": 0.6999,
+      "step": 4824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8700052293111087,
+      "learning_rate": 8.096334515576379e-06,
+      "loss": 0.7963,
+      "step": 4825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1455400205564783,
+      "learning_rate": 8.095520619472739e-06,
+      "loss": 0.7543,
+      "step": 4826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0623457903174733,
+      "learning_rate": 8.094706590347683e-06,
+      "loss": 0.5443,
+      "step": 4827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7043206478327875,
+      "learning_rate": 8.093892428236192e-06,
+      "loss": 0.7919,
+      "step": 4828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.607254099495692,
+      "learning_rate": 8.093078133173256e-06,
+      "loss": 0.8968,
+      "step": 4829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0211972984059974,
+      "learning_rate": 8.09226370519386e-06,
+      "loss": 0.6232,
+      "step": 4830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6968972967120826,
+      "learning_rate": 8.091449144333009e-06,
+      "loss": 0.7295,
+      "step": 4831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4350217601675477,
+      "learning_rate": 8.090634450625704e-06,
+      "loss": 0.7265,
+      "step": 4832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5774160595678772,
+      "learning_rate": 8.08981962410695e-06,
+      "loss": 0.67,
+      "step": 4833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0532328276482068,
+      "learning_rate": 8.089004664811767e-06,
+      "loss": 0.6285,
+      "step": 4834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7452604651662074,
+      "learning_rate": 8.088189572775173e-06,
+      "loss": 0.8476,
+      "step": 4835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9115699111913578,
+      "learning_rate": 8.087374348032198e-06,
+      "loss": 0.7495,
+      "step": 4836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5156384824079203,
+      "learning_rate": 8.08655899061787e-06,
+      "loss": 0.8037,
+      "step": 4837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6763606658388288,
+      "learning_rate": 8.085743500567228e-06,
+      "loss": 0.8262,
+      "step": 4838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.568890312325404,
+      "learning_rate": 8.084927877915314e-06,
+      "loss": 0.821,
+      "step": 4839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5505205589905136,
+      "learning_rate": 8.084112122697178e-06,
+      "loss": 0.7912,
+      "step": 4840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5427354110803118,
+      "learning_rate": 8.083296234947878e-06,
+      "loss": 0.7369,
+      "step": 4841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5229440995138497,
+      "learning_rate": 8.082480214702472e-06,
+      "loss": 0.682,
+      "step": 4842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5797074461055338,
+      "learning_rate": 8.081664061996024e-06,
+      "loss": 0.7165,
+      "step": 4843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9045291992122038,
+      "learning_rate": 8.080847776863609e-06,
+      "loss": 0.698,
+      "step": 4844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6770225473758968,
+      "learning_rate": 8.080031359340303e-06,
+      "loss": 0.6961,
+      "step": 4845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6577045459279327,
+      "learning_rate": 8.079214809461192e-06,
+      "loss": 0.7303,
+      "step": 4846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6597854218850812,
+      "learning_rate": 8.07839812726136e-06,
+      "loss": 0.799,
+      "step": 4847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5841035791322056,
+      "learning_rate": 8.077581312775908e-06,
+      "loss": 0.8164,
+      "step": 4848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5321183753473198,
+      "learning_rate": 8.07676436603993e-06,
+      "loss": 0.7204,
+      "step": 4849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.015644186547971,
+      "learning_rate": 8.075947287088535e-06,
+      "loss": 0.6951,
+      "step": 4850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.730635229390961,
+      "learning_rate": 8.075130075956836e-06,
+      "loss": 0.7541,
+      "step": 4851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.768294901970862,
+      "learning_rate": 8.074312732679948e-06,
+      "loss": 0.7651,
+      "step": 4852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6983568268944194,
+      "learning_rate": 8.073495257292996e-06,
+      "loss": 0.7444,
+      "step": 4853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6643355209207034,
+      "learning_rate": 8.072677649831107e-06,
+      "loss": 0.6139,
+      "step": 4854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8614013567573222,
+      "learning_rate": 8.071859910329416e-06,
+      "loss": 0.7774,
+      "step": 4855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6200249966920266,
+      "learning_rate": 8.071042038823063e-06,
+      "loss": 0.6397,
+      "step": 4856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.59578779292382,
+      "learning_rate": 8.070224035347193e-06,
+      "loss": 0.8392,
+      "step": 4857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6158680542864852,
+      "learning_rate": 8.069405899936961e-06,
+      "loss": 0.8917,
+      "step": 4858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4394448611508552,
+      "learning_rate": 8.068587632627521e-06,
+      "loss": 0.748,
+      "step": 4859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.58155867040709,
+      "learning_rate": 8.067769233454037e-06,
+      "loss": 0.6823,
+      "step": 4860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3374125442904972,
+      "learning_rate": 8.066950702451673e-06,
+      "loss": 0.7152,
+      "step": 4861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.923867113969024,
+      "learning_rate": 8.066132039655609e-06,
+      "loss": 0.7626,
+      "step": 4862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7091158530963912,
+      "learning_rate": 8.065313245101024e-06,
+      "loss": 0.7803,
+      "step": 4863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4366654994908847,
+      "learning_rate": 8.064494318823102e-06,
+      "loss": 0.7387,
+      "step": 4864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6431921209702411,
+      "learning_rate": 8.06367526085703e-06,
+      "loss": 0.8206,
+      "step": 4865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6612274602798534,
+      "learning_rate": 8.062856071238014e-06,
+      "loss": 0.7002,
+      "step": 4866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8787976076168835,
+      "learning_rate": 8.062036750001249e-06,
+      "loss": 0.7253,
+      "step": 4867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8862309220184061,
+      "learning_rate": 8.061217297181946e-06,
+      "loss": 0.8163,
+      "step": 4868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0091660036848096,
+      "learning_rate": 8.060397712815318e-06,
+      "loss": 0.648,
+      "step": 4869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5705774334895133,
+      "learning_rate": 8.059577996936583e-06,
+      "loss": 0.7718,
+      "step": 4870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1926605414074354,
+      "learning_rate": 8.05875814958097e-06,
+      "loss": 0.7242,
+      "step": 4871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0147958567791304,
+      "learning_rate": 8.057938170783704e-06,
+      "loss": 0.6156,
+      "step": 4872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6888088427752328,
+      "learning_rate": 8.057118060580027e-06,
+      "loss": 0.6741,
+      "step": 4873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.191614388684196,
+      "learning_rate": 8.056297819005177e-06,
+      "loss": 0.7155,
+      "step": 4874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0994746310948535,
+      "learning_rate": 8.055477446094404e-06,
+      "loss": 0.703,
+      "step": 4875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9475907236411272,
+      "learning_rate": 8.054656941882959e-06,
+      "loss": 0.9034,
+      "step": 4876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5701894158420253,
+      "learning_rate": 8.053836306406102e-06,
+      "loss": 0.7308,
+      "step": 4877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7238642426667452,
+      "learning_rate": 8.053015539699099e-06,
+      "loss": 0.7778,
+      "step": 4878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5432370957721815,
+      "learning_rate": 8.052194641797217e-06,
+      "loss": 0.6983,
+      "step": 4879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5559304772659452,
+      "learning_rate": 8.051373612735736e-06,
+      "loss": 0.7209,
+      "step": 4880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9998902366189683,
+      "learning_rate": 8.050552452549933e-06,
+      "loss": 0.6455,
+      "step": 4881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.466076011969828,
+      "learning_rate": 8.049731161275099e-06,
+      "loss": 0.5977,
+      "step": 4882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4407623005417107,
+      "learning_rate": 8.048909738946523e-06,
+      "loss": 0.738,
+      "step": 4883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4526737417822235,
+      "learning_rate": 8.048088185599507e-06,
+      "loss": 0.7582,
+      "step": 4884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6278686549152017,
+      "learning_rate": 8.047266501269352e-06,
+      "loss": 0.6952,
+      "step": 4885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6819430862550382,
+      "learning_rate": 8.046444685991369e-06,
+      "loss": 0.8734,
+      "step": 4886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7066095623550435,
+      "learning_rate": 8.045622739800873e-06,
+      "loss": 0.769,
+      "step": 4887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9669005404720008,
+      "learning_rate": 8.044800662733185e-06,
+      "loss": 0.8549,
+      "step": 4888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7979057340311033,
+      "learning_rate": 8.043978454823632e-06,
+      "loss": 0.8448,
+      "step": 4889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0752317284080832,
+      "learning_rate": 8.043156116107545e-06,
+      "loss": 0.6948,
+      "step": 4890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7954694186581646,
+      "learning_rate": 8.042333646620262e-06,
+      "loss": 0.794,
+      "step": 4891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5166122105388824,
+      "learning_rate": 8.041511046397127e-06,
+      "loss": 0.7186,
+      "step": 4892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9533541468823548,
+      "learning_rate": 8.040688315473489e-06,
+      "loss": 0.7451,
+      "step": 4893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.733942495889719,
+      "learning_rate": 8.0398654538847e-06,
+      "loss": 0.7206,
+      "step": 4894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1401809189760395,
+      "learning_rate": 8.039042461666125e-06,
+      "loss": 0.7272,
+      "step": 4895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6439106047943461,
+      "learning_rate": 8.038219338853126e-06,
+      "loss": 0.753,
+      "step": 4896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7021894383203324,
+      "learning_rate": 8.037396085481075e-06,
+      "loss": 0.7529,
+      "step": 4897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9172352936226071,
+      "learning_rate": 8.03657270158535e-06,
+      "loss": 0.7596,
+      "step": 4898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6070660584962888,
+      "learning_rate": 8.035749187201333e-06,
+      "loss": 0.7279,
+      "step": 4899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5737143198359935,
+      "learning_rate": 8.034925542364412e-06,
+      "loss": 0.8093,
+      "step": 4900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7444942590323331,
+      "learning_rate": 8.034101767109982e-06,
+      "loss": 0.7452,
+      "step": 4901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9422695470282683,
+      "learning_rate": 8.033277861473441e-06,
+      "loss": 0.6945,
+      "step": 4902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8549863565393216,
+      "learning_rate": 8.032453825490197e-06,
+      "loss": 0.6569,
+      "step": 4903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7431072395109595,
+      "learning_rate": 8.031629659195657e-06,
+      "loss": 0.734,
+      "step": 4904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9764298688171664,
+      "learning_rate": 8.030805362625238e-06,
+      "loss": 0.8691,
+      "step": 4905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6057844043321547,
+      "learning_rate": 8.029980935814363e-06,
+      "loss": 0.7276,
+      "step": 4906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3060727540899724,
+      "learning_rate": 8.029156378798459e-06,
+      "loss": 0.6747,
+      "step": 4907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.076849940654713,
+      "learning_rate": 8.028331691612961e-06,
+      "loss": 0.693,
+      "step": 4908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6421213729768374,
+      "learning_rate": 8.027506874293304e-06,
+      "loss": 0.7077,
+      "step": 4909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.694032095287676,
+      "learning_rate": 8.026681926874932e-06,
+      "loss": 0.672,
+      "step": 4910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6550173920131672,
+      "learning_rate": 8.025856849393301e-06,
+      "loss": 0.722,
+      "step": 4911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.386656637057385,
+      "learning_rate": 8.02503164188386e-06,
+      "loss": 0.6577,
+      "step": 4912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6109575909803564,
+      "learning_rate": 8.024206304382072e-06,
+      "loss": 0.7817,
+      "step": 4913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.714093782930831,
+      "learning_rate": 8.023380836923404e-06,
+      "loss": 0.7995,
+      "step": 4914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8023630464393925,
+      "learning_rate": 8.022555239543328e-06,
+      "loss": 0.7261,
+      "step": 4915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.661264551682016,
+      "learning_rate": 8.021729512277323e-06,
+      "loss": 0.7705,
+      "step": 4916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.651713287436719,
+      "learning_rate": 8.02090365516087e-06,
+      "loss": 0.6868,
+      "step": 4917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1871450077564543,
+      "learning_rate": 8.02007766822946e-06,
+      "loss": 0.7573,
+      "step": 4918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7106678224370606,
+      "learning_rate": 8.019251551518585e-06,
+      "loss": 0.7007,
+      "step": 4919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.841064930889779,
+      "learning_rate": 8.018425305063749e-06,
+      "loss": 0.8201,
+      "step": 4920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8098363859551907,
+      "learning_rate": 8.017598928900452e-06,
+      "loss": 0.8282,
+      "step": 4921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7689724091642591,
+      "learning_rate": 8.01677242306421e-06,
+      "loss": 0.8417,
+      "step": 4922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6293882114860692,
+      "learning_rate": 8.015945787590537e-06,
+      "loss": 0.7493,
+      "step": 4923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.713521876875804,
+      "learning_rate": 8.015119022514958e-06,
+      "loss": 0.8818,
+      "step": 4924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5069858929730233,
+      "learning_rate": 8.014292127873002e-06,
+      "loss": 0.7326,
+      "step": 4925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7166891097456694,
+      "learning_rate": 8.013465103700197e-06,
+      "loss": 0.8332,
+      "step": 4926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6440338258449547,
+      "learning_rate": 8.012637950032086e-06,
+      "loss": 0.6641,
+      "step": 4927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7012301032502644,
+      "learning_rate": 8.011810666904212e-06,
+      "loss": 0.7442,
+      "step": 4928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1245821094239796,
+      "learning_rate": 8.010983254352127e-06,
+      "loss": 0.6167,
+      "step": 4929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.202648217407046,
+      "learning_rate": 8.010155712411386e-06,
+      "loss": 0.818,
+      "step": 4930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6590855783990368,
+      "learning_rate": 8.009328041117547e-06,
+      "loss": 0.7206,
+      "step": 4931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.83220836287284,
+      "learning_rate": 8.008500240506182e-06,
+      "loss": 0.7372,
+      "step": 4932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.794035480888466,
+      "learning_rate": 8.007672310612863e-06,
+      "loss": 0.7861,
+      "step": 4933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9200617999342524,
+      "learning_rate": 8.006844251473165e-06,
+      "loss": 0.7777,
+      "step": 4934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0308193450856629,
+      "learning_rate": 8.006016063122672e-06,
+      "loss": 0.7146,
+      "step": 4935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4881157637152354,
+      "learning_rate": 8.005187745596976e-06,
+      "loss": 0.7633,
+      "step": 4936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8168480135261666,
+      "learning_rate": 8.004359298931668e-06,
+      "loss": 0.6442,
+      "step": 4937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5834048938077119,
+      "learning_rate": 8.003530723162352e-06,
+      "loss": 0.6714,
+      "step": 4938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7779673699337986,
+      "learning_rate": 8.002702018324629e-06,
+      "loss": 0.8267,
+      "step": 4939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7029317691548278,
+      "learning_rate": 8.001873184454114e-06,
+      "loss": 0.8648,
+      "step": 4940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9162470294934224,
+      "learning_rate": 8.001044221586423e-06,
+      "loss": 0.8224,
+      "step": 4941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1580659625332586,
+      "learning_rate": 8.000215129757178e-06,
+      "loss": 0.7251,
+      "step": 4942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.617716870012279,
+      "learning_rate": 7.999385909002007e-06,
+      "loss": 0.7848,
+      "step": 4943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5949969624655167,
+      "learning_rate": 7.998556559356543e-06,
+      "loss": 0.7362,
+      "step": 4944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6322624520627653,
+      "learning_rate": 7.997727080856425e-06,
+      "loss": 0.8575,
+      "step": 4945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5859789269707985,
+      "learning_rate": 7.9968974735373e-06,
+      "loss": 0.621,
+      "step": 4946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0097414876864192,
+      "learning_rate": 7.996067737434815e-06,
+      "loss": 0.7898,
+      "step": 4947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5915118165285873,
+      "learning_rate": 7.995237872584627e-06,
+      "loss": 0.7723,
+      "step": 4948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.609592234636802,
+      "learning_rate": 7.994407879022397e-06,
+      "loss": 0.7534,
+      "step": 4949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9733148881903182,
+      "learning_rate": 7.99357775678379e-06,
+      "loss": 0.8011,
+      "step": 4950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5305855172668743,
+      "learning_rate": 7.99274750590448e-06,
+      "loss": 0.7512,
+      "step": 4951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8189613475203852,
+      "learning_rate": 7.991917126420147e-06,
+      "loss": 0.7707,
+      "step": 4952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.513700521323628,
+      "learning_rate": 7.99108661836647e-06,
+      "loss": 0.7512,
+      "step": 4953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9338680355258036,
+      "learning_rate": 7.990255981779139e-06,
+      "loss": 0.7892,
+      "step": 4954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6214834909013955,
+      "learning_rate": 7.989425216693849e-06,
+      "loss": 0.7879,
+      "step": 4955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2316522477597796,
+      "learning_rate": 7.9885943231463e-06,
+      "loss": 0.6171,
+      "step": 4956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.188406719012613,
+      "learning_rate": 7.987763301172197e-06,
+      "loss": 0.606,
+      "step": 4957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6705455803223734,
+      "learning_rate": 7.986932150807253e-06,
+      "loss": 0.7459,
+      "step": 4958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.189324039328686,
+      "learning_rate": 7.986100872087177e-06,
+      "loss": 0.6232,
+      "step": 4959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5634947540976891,
+      "learning_rate": 7.9852694650477e-06,
+      "loss": 0.7119,
+      "step": 4960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7268272828127744,
+      "learning_rate": 7.984437929724546e-06,
+      "loss": 0.7249,
+      "step": 4961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5867748106082045,
+      "learning_rate": 7.983606266153445e-06,
+      "loss": 0.7369,
+      "step": 4962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5345467672581776,
+      "learning_rate": 7.98277447437014e-06,
+      "loss": 0.786,
+      "step": 4963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6422779934582161,
+      "learning_rate": 7.981942554410371e-06,
+      "loss": 0.7265,
+      "step": 4964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9457854376605237,
+      "learning_rate": 7.98111050630989e-06,
+      "loss": 0.8778,
+      "step": 4965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.690192548365369,
+      "learning_rate": 7.980278330104452e-06,
+      "loss": 0.6294,
+      "step": 4966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2932771843031141,
+      "learning_rate": 7.979446025829815e-06,
+      "loss": 0.7054,
+      "step": 4967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0479230775279642,
+      "learning_rate": 7.978613593521748e-06,
+      "loss": 0.6083,
+      "step": 4968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3112714108951746,
+      "learning_rate": 7.97778103321602e-06,
+      "loss": 0.6275,
+      "step": 4969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6786845763144567,
+      "learning_rate": 7.976948344948412e-06,
+      "loss": 0.873,
+      "step": 4970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9877752280010388,
+      "learning_rate": 7.976115528754702e-06,
+      "loss": 0.7748,
+      "step": 4971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6687270186562968,
+      "learning_rate": 7.975282584670679e-06,
+      "loss": 0.7932,
+      "step": 4972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6748735938764594,
+      "learning_rate": 7.974449512732137e-06,
+      "loss": 0.6832,
+      "step": 4973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8216125722950542,
+      "learning_rate": 7.973616312974876e-06,
+      "loss": 0.7266,
+      "step": 4974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3920604209639778,
+      "learning_rate": 7.972782985434697e-06,
+      "loss": 0.6638,
+      "step": 4975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6965145693630412,
+      "learning_rate": 7.971949530147417e-06,
+      "loss": 0.866,
+      "step": 4976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9469250916075311,
+      "learning_rate": 7.971115947148842e-06,
+      "loss": 0.7779,
+      "step": 4977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8012279818128658,
+      "learning_rate": 7.9702822364748e-06,
+      "loss": 0.711,
+      "step": 4978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6679687122486118,
+      "learning_rate": 7.969448398161115e-06,
+      "loss": 0.7892,
+      "step": 4979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0909649181862227,
+      "learning_rate": 7.968614432243617e-06,
+      "loss": 0.8444,
+      "step": 4980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.643963886622507,
+      "learning_rate": 7.96778033875815e-06,
+      "loss": 0.7609,
+      "step": 4981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.61496712752907,
+      "learning_rate": 7.966946117740548e-06,
+      "loss": 0.7579,
+      "step": 4982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.873467407077143,
+      "learning_rate": 7.966111769226666e-06,
+      "loss": 0.6727,
+      "step": 4983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3892545509804561,
+      "learning_rate": 7.965277293252354e-06,
+      "loss": 0.7017,
+      "step": 4984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6682609176854564,
+      "learning_rate": 7.964442689853473e-06,
+      "loss": 0.8093,
+      "step": 4985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6065449634032765,
+      "learning_rate": 7.963607959065887e-06,
+      "loss": 0.7511,
+      "step": 4986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.707473640955341,
+      "learning_rate": 7.962773100925467e-06,
+      "loss": 0.8649,
+      "step": 4987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4709265267526916,
+      "learning_rate": 7.961938115468088e-06,
+      "loss": 0.7264,
+      "step": 4988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.772712744637312,
+      "learning_rate": 7.961103002729634e-06,
+      "loss": 0.7489,
+      "step": 4989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1650764306064787,
+      "learning_rate": 7.960267762745985e-06,
+      "loss": 0.6921,
+      "step": 4990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.654858894445648,
+      "learning_rate": 7.95943239555304e-06,
+      "loss": 0.7534,
+      "step": 4991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7160685305798422,
+      "learning_rate": 7.958596901186695e-06,
+      "loss": 0.84,
+      "step": 4992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.658502268525517,
+      "learning_rate": 7.95776127968285e-06,
+      "loss": 0.8026,
+      "step": 4993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6879138575014823,
+      "learning_rate": 7.956925531077417e-06,
+      "loss": 0.7948,
+      "step": 4994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.680193296285951,
+      "learning_rate": 7.956089655406307e-06,
+      "loss": 0.763,
+      "step": 4995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.906542499023394,
+      "learning_rate": 7.955253652705442e-06,
+      "loss": 0.8078,
+      "step": 4996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5532853100560926,
+      "learning_rate": 7.954417523010747e-06,
+      "loss": 0.6883,
+      "step": 4997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2820738257153173,
+      "learning_rate": 7.953581266358148e-06,
+      "loss": 0.6755,
+      "step": 4998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8407968867803148,
+      "learning_rate": 7.952744882783587e-06,
+      "loss": 0.8263,
+      "step": 4999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0528643776985773,
+      "learning_rate": 7.951908372323001e-06,
+      "loss": 0.8481,
+      "step": 5000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6420814199372875,
+      "learning_rate": 7.951071735012337e-06,
+      "loss": 0.7598,
+      "step": 5001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6059201592516439,
+      "learning_rate": 7.95023497088755e-06,
+      "loss": 0.8303,
+      "step": 5002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.549564828359537,
+      "learning_rate": 7.949398079984594e-06,
+      "loss": 0.7669,
+      "step": 5003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6365397539371351,
+      "learning_rate": 7.948561062339435e-06,
+      "loss": 0.7721,
+      "step": 5004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.798345403634553,
+      "learning_rate": 7.94772391798804e-06,
+      "loss": 0.7159,
+      "step": 5005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0293052679633243,
+      "learning_rate": 7.946886646966384e-06,
+      "loss": 0.6519,
+      "step": 5006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2124239767960763,
+      "learning_rate": 7.946049249310445e-06,
+      "loss": 0.732,
+      "step": 5007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7197111593247867,
+      "learning_rate": 7.945211725056208e-06,
+      "loss": 0.8892,
+      "step": 5008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7149379516750813,
+      "learning_rate": 7.944374074239665e-06,
+      "loss": 0.609,
+      "step": 5009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4203829838809858,
+      "learning_rate": 7.94353629689681e-06,
+      "loss": 0.6555,
+      "step": 5010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6760164353506524,
+      "learning_rate": 7.942698393063646e-06,
+      "loss": 0.7566,
+      "step": 5011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1980966115275296,
+      "learning_rate": 7.941860362776176e-06,
+      "loss": 0.8155,
+      "step": 5012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.657926506524567,
+      "learning_rate": 7.941022206070415e-06,
+      "loss": 0.8407,
+      "step": 5013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.260005266335191,
+      "learning_rate": 7.940183922982381e-06,
+      "loss": 0.7122,
+      "step": 5014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6010938421071432,
+      "learning_rate": 7.939345513548095e-06,
+      "loss": 0.8532,
+      "step": 5015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7844237848328208,
+      "learning_rate": 7.938506977803586e-06,
+      "loss": 0.8024,
+      "step": 5016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6188428900290859,
+      "learning_rate": 7.937668315784888e-06,
+      "loss": 0.7037,
+      "step": 5017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7524458118088133,
+      "learning_rate": 7.93682952752804e-06,
+      "loss": 0.7504,
+      "step": 5018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7668562632729305,
+      "learning_rate": 7.935990613069087e-06,
+      "loss": 0.6594,
+      "step": 5019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9784692073043324,
+      "learning_rate": 7.935151572444076e-06,
+      "loss": 0.7786,
+      "step": 5020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.752134048740872,
+      "learning_rate": 7.934312405689067e-06,
+      "loss": 0.7839,
+      "step": 5021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5603199497572606,
+      "learning_rate": 7.933473112840117e-06,
+      "loss": 0.7607,
+      "step": 5022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9123818559496983,
+      "learning_rate": 7.932633693933298e-06,
+      "loss": 0.9009,
+      "step": 5023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7107355258354338,
+      "learning_rate": 7.931794149004675e-06,
+      "loss": 0.7661,
+      "step": 5024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6514761818068018,
+      "learning_rate": 7.930954478090329e-06,
+      "loss": 0.8112,
+      "step": 5025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7808700587577373,
+      "learning_rate": 7.930114681226341e-06,
+      "loss": 0.6925,
+      "step": 5026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9734453596441401,
+      "learning_rate": 7.9292747584488e-06,
+      "loss": 0.821,
+      "step": 5027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.433367310004602,
+      "learning_rate": 7.928434709793798e-06,
+      "loss": 0.7556,
+      "step": 5028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0905136174123389,
+      "learning_rate": 7.927594535297433e-06,
+      "loss": 0.5228,
+      "step": 5029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5601742851050782,
+      "learning_rate": 7.926754234995814e-06,
+      "loss": 0.7101,
+      "step": 5030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5743035340936493,
+      "learning_rate": 7.925913808925045e-06,
+      "loss": 0.8059,
+      "step": 5031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6128572706308597,
+      "learning_rate": 7.925073257121245e-06,
+      "loss": 0.8817,
+      "step": 5032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2971424509450935,
+      "learning_rate": 7.924232579620533e-06,
+      "loss": 0.5852,
+      "step": 5033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0782710884983673,
+      "learning_rate": 7.923391776459031e-06,
+      "loss": 0.6526,
+      "step": 5034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1741295879150817,
+      "learning_rate": 7.922550847672876e-06,
+      "loss": 0.8591,
+      "step": 5035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0319907240556385,
+      "learning_rate": 7.921709793298202e-06,
+      "loss": 0.6194,
+      "step": 5036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5443684936669115,
+      "learning_rate": 7.920868613371151e-06,
+      "loss": 0.7227,
+      "step": 5037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8833289674855043,
+      "learning_rate": 7.920027307927872e-06,
+      "loss": 0.6715,
+      "step": 5038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7173714740834511,
+      "learning_rate": 7.919185877004515e-06,
+      "loss": 0.8261,
+      "step": 5039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1690634575857541,
+      "learning_rate": 7.91834432063724e-06,
+      "loss": 0.7156,
+      "step": 5040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6349634176740913,
+      "learning_rate": 7.917502638862209e-06,
+      "loss": 0.6633,
+      "step": 5041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1684342763050624,
+      "learning_rate": 7.916660831715592e-06,
+      "loss": 0.6461,
+      "step": 5042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.598645628717302,
+      "learning_rate": 7.915818899233564e-06,
+      "loss": 0.7882,
+      "step": 5043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1835923258154493,
+      "learning_rate": 7.914976841452304e-06,
+      "loss": 0.8363,
+      "step": 5044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1911295472021948,
+      "learning_rate": 7.914134658407997e-06,
+      "loss": 0.692,
+      "step": 5045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9924468139461555,
+      "learning_rate": 7.913292350136834e-06,
+      "loss": 0.8396,
+      "step": 5046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2281229905057427,
+      "learning_rate": 7.912449916675008e-06,
+      "loss": 0.7318,
+      "step": 5047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.19420315914226,
+      "learning_rate": 7.911607358058726e-06,
+      "loss": 0.7175,
+      "step": 5048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5223435404047854,
+      "learning_rate": 7.91076467432419e-06,
+      "loss": 0.7592,
+      "step": 5049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6110564079604637,
+      "learning_rate": 7.909921865507614e-06,
+      "loss": 0.7383,
+      "step": 5050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5071431766804397,
+      "learning_rate": 7.909078931645215e-06,
+      "loss": 0.6851,
+      "step": 5051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9900213404361466,
+      "learning_rate": 7.908235872773214e-06,
+      "loss": 0.674,
+      "step": 5052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.921353491892052,
+      "learning_rate": 7.907392688927842e-06,
+      "loss": 0.7643,
+      "step": 5053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0742095510670262,
+      "learning_rate": 7.90654938014533e-06,
+      "loss": 0.7097,
+      "step": 5054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.276183442865872,
+      "learning_rate": 7.905705946461918e-06,
+      "loss": 0.7075,
+      "step": 5055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7402225366643767,
+      "learning_rate": 7.90486238791385e-06,
+      "loss": 0.7242,
+      "step": 5056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5697057534963894,
+      "learning_rate": 7.904018704537376e-06,
+      "loss": 0.8431,
+      "step": 5057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.561837162339394,
+      "learning_rate": 7.90317489636875e-06,
+      "loss": 0.7436,
+      "step": 5058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3734269213196604,
+      "learning_rate": 7.902330963444234e-06,
+      "loss": 0.6707,
+      "step": 5059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7631430267288182,
+      "learning_rate": 7.901486905800092e-06,
+      "loss": 0.7074,
+      "step": 5060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.710180178712513,
+      "learning_rate": 7.900642723472596e-06,
+      "loss": 0.7363,
+      "step": 5061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8338229101074734,
+      "learning_rate": 7.899798416498023e-06,
+      "loss": 0.7635,
+      "step": 5062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0665990931419103,
+      "learning_rate": 7.898953984912652e-06,
+      "loss": 0.8262,
+      "step": 5063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6311550957652146,
+      "learning_rate": 7.898109428752773e-06,
+      "loss": 0.8037,
+      "step": 5064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6975200178233794,
+      "learning_rate": 7.897264748054679e-06,
+      "loss": 0.8286,
+      "step": 5065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9613440552568824,
+      "learning_rate": 7.896419942854662e-06,
+      "loss": 0.8831,
+      "step": 5066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0242337304572313,
+      "learning_rate": 7.895575013189032e-06,
+      "loss": 0.6439,
+      "step": 5067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7378997911561431,
+      "learning_rate": 7.894729959094097e-06,
+      "loss": 0.7218,
+      "step": 5068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7933689409052074,
+      "learning_rate": 7.893884780606164e-06,
+      "loss": 0.8767,
+      "step": 5069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8823686750649564,
+      "learning_rate": 7.893039477761562e-06,
+      "loss": 0.8504,
+      "step": 5070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8598165102491253,
+      "learning_rate": 7.892194050596606e-06,
+      "loss": 0.8002,
+      "step": 5071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.443029468095854,
+      "learning_rate": 7.891348499147634e-06,
+      "loss": 0.6908,
+      "step": 5072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2798963004686683,
+      "learning_rate": 7.890502823450975e-06,
+      "loss": 0.6842,
+      "step": 5073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4532909401927963,
+      "learning_rate": 7.889657023542973e-06,
+      "loss": 0.6564,
+      "step": 5074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7193219259518553,
+      "learning_rate": 7.888811099459974e-06,
+      "loss": 0.6868,
+      "step": 5075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4333254049963298,
+      "learning_rate": 7.887965051238326e-06,
+      "loss": 0.6843,
+      "step": 5076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6625845368808803,
+      "learning_rate": 7.88711887891439e-06,
+      "loss": 0.8211,
+      "step": 5077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1474037184220658,
+      "learning_rate": 7.886272582524525e-06,
+      "loss": 0.7153,
+      "step": 5078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7843743145128355,
+      "learning_rate": 7.885426162105101e-06,
+      "loss": 0.7403,
+      "step": 5079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7035590767755413,
+      "learning_rate": 7.884579617692487e-06,
+      "loss": 0.8291,
+      "step": 5080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7724401029394108,
+      "learning_rate": 7.883732949323061e-06,
+      "loss": 0.8214,
+      "step": 5081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8116589791145024,
+      "learning_rate": 7.882886157033209e-06,
+      "loss": 0.713,
+      "step": 5082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.156963229748554,
+      "learning_rate": 7.882039240859318e-06,
+      "loss": 0.7241,
+      "step": 5083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6052380337594667,
+      "learning_rate": 7.881192200837785e-06,
+      "loss": 0.9357,
+      "step": 5084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.442043093428582,
+      "learning_rate": 7.880345037005002e-06,
+      "loss": 0.7652,
+      "step": 5085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7074657413273922,
+      "learning_rate": 7.879497749397382e-06,
+      "loss": 0.7335,
+      "step": 5086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7160830012529045,
+      "learning_rate": 7.878650338051328e-06,
+      "loss": 0.7843,
+      "step": 5087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.723194898592235,
+      "learning_rate": 7.877802803003258e-06,
+      "loss": 0.7412,
+      "step": 5088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7834576012787697,
+      "learning_rate": 7.876955144289594e-06,
+      "loss": 0.846,
+      "step": 5089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6577383476835381,
+      "learning_rate": 7.876107361946758e-06,
+      "loss": 0.6503,
+      "step": 5090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7278385750275056,
+      "learning_rate": 7.875259456011185e-06,
+      "loss": 0.8545,
+      "step": 5091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2636940180271228,
+      "learning_rate": 7.874411426519308e-06,
+      "loss": 0.6835,
+      "step": 5092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5527483793093282,
+      "learning_rate": 7.873563273507571e-06,
+      "loss": 0.7166,
+      "step": 5093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8338344015808512,
+      "learning_rate": 7.872714997012421e-06,
+      "loss": 0.7337,
+      "step": 5094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.864649173311467,
+      "learning_rate": 7.87186659707031e-06,
+      "loss": 0.634,
+      "step": 5095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6849772924074717,
+      "learning_rate": 7.871018073717693e-06,
+      "loss": 0.6643,
+      "step": 5096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.941507018531259,
+      "learning_rate": 7.870169426991038e-06,
+      "loss": 0.6745,
+      "step": 5097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9320577323900083,
+      "learning_rate": 7.86932065692681e-06,
+      "loss": 0.9255,
+      "step": 5098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7239062349526002,
+      "learning_rate": 7.868471763561482e-06,
+      "loss": 0.6988,
+      "step": 5099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7099549073800335,
+      "learning_rate": 7.867622746931533e-06,
+      "loss": 0.7811,
+      "step": 5100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8202249140403515,
+      "learning_rate": 7.86677360707345e-06,
+      "loss": 0.9203,
+      "step": 5101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6942139410674053,
+      "learning_rate": 7.865924344023722e-06,
+      "loss": 0.717,
+      "step": 5102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.679717512833047,
+      "learning_rate": 7.865074957818839e-06,
+      "loss": 0.7877,
+      "step": 5103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8410203332341075,
+      "learning_rate": 7.864225448495304e-06,
+      "loss": 0.7732,
+      "step": 5104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.087197835742401,
+      "learning_rate": 7.863375816089624e-06,
+      "loss": 0.7479,
+      "step": 5105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.494192789923904,
+      "learning_rate": 7.862526060638309e-06,
+      "loss": 0.6749,
+      "step": 5106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6903702308663653,
+      "learning_rate": 7.861676182177873e-06,
+      "loss": 0.6893,
+      "step": 5107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4373102605377386,
+      "learning_rate": 7.860826180744837e-06,
+      "loss": 0.7225,
+      "step": 5108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.907864173603959,
+      "learning_rate": 7.85997605637573e-06,
+      "loss": 0.7447,
+      "step": 5109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7885444650484954,
+      "learning_rate": 7.859125809107082e-06,
+      "loss": 0.8527,
+      "step": 5110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0463808595078223,
+      "learning_rate": 7.85827543897543e-06,
+      "loss": 0.7546,
+      "step": 5111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5554725205261573,
+      "learning_rate": 7.857424946017317e-06,
+      "loss": 0.7195,
+      "step": 5112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4768457084390385,
+      "learning_rate": 7.85657433026929e-06,
+      "loss": 0.6637,
+      "step": 5113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4601255789706717,
+      "learning_rate": 7.855723591767903e-06,
+      "loss": 0.7246,
+      "step": 5114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5939884598124157,
+      "learning_rate": 7.854872730549712e-06,
+      "loss": 0.6821,
+      "step": 5115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.197958745880302,
+      "learning_rate": 7.854021746651283e-06,
+      "loss": 0.7732,
+      "step": 5116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9329939658144364,
+      "learning_rate": 7.853170640109182e-06,
+      "loss": 0.7904,
+      "step": 5117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4071458298992403,
+      "learning_rate": 7.852319410959984e-06,
+      "loss": 0.6854,
+      "step": 5118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5384711870214796,
+      "learning_rate": 7.85146805924027e-06,
+      "loss": 0.6835,
+      "step": 5119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.555919936734464,
+      "learning_rate": 7.850616584986621e-06,
+      "loss": 0.7138,
+      "step": 5120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5567314675661124,
+      "learning_rate": 7.84976498823563e-06,
+      "loss": 0.6778,
+      "step": 5121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7352674856924257,
+      "learning_rate": 7.84891326902389e-06,
+      "loss": 0.8299,
+      "step": 5122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5337786855392692,
+      "learning_rate": 7.848061427388002e-06,
+      "loss": 0.7195,
+      "step": 5123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.557988371360905,
+      "learning_rate": 7.847209463364574e-06,
+      "loss": 0.6828,
+      "step": 5124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6519145073548491,
+      "learning_rate": 7.846357376990213e-06,
+      "loss": 0.7602,
+      "step": 5125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.712457371171235,
+      "learning_rate": 7.845505168301533e-06,
+      "loss": 0.7102,
+      "step": 5126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.987506977336771,
+      "learning_rate": 7.844652837335162e-06,
+      "loss": 0.7388,
+      "step": 5127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8059968656682626,
+      "learning_rate": 7.843800384127721e-06,
+      "loss": 0.6845,
+      "step": 5128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1491811345419891,
+      "learning_rate": 7.842947808715848e-06,
+      "loss": 0.6185,
+      "step": 5129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5303269081263267,
+      "learning_rate": 7.842095111136173e-06,
+      "loss": 0.7255,
+      "step": 5130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1735650765853234,
+      "learning_rate": 7.841242291425342e-06,
+      "loss": 0.7443,
+      "step": 5131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1201298458550628,
+      "learning_rate": 7.840389349620002e-06,
+      "loss": 0.6836,
+      "step": 5132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3025078620338106,
+      "learning_rate": 7.839536285756804e-06,
+      "loss": 0.6908,
+      "step": 5133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.605996259536892,
+      "learning_rate": 7.83868309987241e-06,
+      "loss": 0.6658,
+      "step": 5134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0295619647449556,
+      "learning_rate": 7.837829792003479e-06,
+      "loss": 0.8981,
+      "step": 5135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3318376241283605,
+      "learning_rate": 7.836976362186684e-06,
+      "loss": 0.6842,
+      "step": 5136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5508475532538837,
+      "learning_rate": 7.836122810458696e-06,
+      "loss": 0.7147,
+      "step": 5137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9252856020657083,
+      "learning_rate": 7.835269136856194e-06,
+      "loss": 0.8036,
+      "step": 5138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5699817283235107,
+      "learning_rate": 7.834415341415862e-06,
+      "loss": 0.8687,
+      "step": 5139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.625815530353136,
+      "learning_rate": 7.833561424174392e-06,
+      "loss": 0.8503,
+      "step": 5140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.782977669511771,
+      "learning_rate": 7.832707385168474e-06,
+      "loss": 0.789,
+      "step": 5141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9554775469611896,
+      "learning_rate": 7.831853224434815e-06,
+      "loss": 0.6666,
+      "step": 5142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2533402487004368,
+      "learning_rate": 7.830998942010113e-06,
+      "loss": 0.6426,
+      "step": 5143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8200850697717195,
+      "learning_rate": 7.830144537931082e-06,
+      "loss": 0.8252,
+      "step": 5144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.027123041592555,
+      "learning_rate": 7.829290012234438e-06,
+      "loss": 0.7988,
+      "step": 5145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8351954787581064,
+      "learning_rate": 7.8284353649569e-06,
+      "loss": 0.7425,
+      "step": 5146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.551627442700939,
+      "learning_rate": 7.827580596135196e-06,
+      "loss": 0.6359,
+      "step": 5147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5798984217200684,
+      "learning_rate": 7.826725705806056e-06,
+      "loss": 0.8092,
+      "step": 5148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.911455602702293,
+      "learning_rate": 7.825870694006217e-06,
+      "loss": 0.7565,
+      "step": 5149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1442080371838341,
+      "learning_rate": 7.82501556077242e-06,
+      "loss": 0.6214,
+      "step": 5150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5845123342952931,
+      "learning_rate": 7.824160306141413e-06,
+      "loss": 0.8344,
+      "step": 5151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.831272606451045,
+      "learning_rate": 7.823304930149949e-06,
+      "loss": 0.8962,
+      "step": 5152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6793537760151556,
+      "learning_rate": 7.822449432834784e-06,
+      "loss": 0.9006,
+      "step": 5153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5682290622621735,
+      "learning_rate": 7.82159381423268e-06,
+      "loss": 0.7204,
+      "step": 5154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1132517228119816,
+      "learning_rate": 7.820738074380404e-06,
+      "loss": 0.7176,
+      "step": 5155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4248279690409,
+      "learning_rate": 7.819882213314736e-06,
+      "loss": 0.6413,
+      "step": 5156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3040325813086993,
+      "learning_rate": 7.819026231072445e-06,
+      "loss": 0.7893,
+      "step": 5157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8432128225608713,
+      "learning_rate": 7.81817012769032e-06,
+      "loss": 0.7052,
+      "step": 5158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.060136227792833,
+      "learning_rate": 7.817313903205148e-06,
+      "loss": 0.7117,
+      "step": 5159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0357696128643386,
+      "learning_rate": 7.816457557653722e-06,
+      "loss": 0.6143,
+      "step": 5160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.997141146585168,
+      "learning_rate": 7.815601091072845e-06,
+      "loss": 0.8007,
+      "step": 5161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5207757010380605,
+      "learning_rate": 7.814744503499317e-06,
+      "loss": 0.7796,
+      "step": 5162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.835237522570645,
+      "learning_rate": 7.813887794969948e-06,
+      "loss": 0.7476,
+      "step": 5163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6805188169701493,
+      "learning_rate": 7.813030965521554e-06,
+      "loss": 0.7357,
+      "step": 5164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7326902155201662,
+      "learning_rate": 7.812174015190955e-06,
+      "loss": 0.7871,
+      "step": 5165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7524095319554482,
+      "learning_rate": 7.811316944014974e-06,
+      "loss": 0.7297,
+      "step": 5166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.021101224962927,
+      "learning_rate": 7.810459752030444e-06,
+      "loss": 0.7355,
+      "step": 5167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1160023410188529,
+      "learning_rate": 7.809602439274198e-06,
+      "loss": 0.565,
+      "step": 5168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.623266494287545,
+      "learning_rate": 7.80874500578308e-06,
+      "loss": 0.6794,
+      "step": 5169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9835794708646013,
+      "learning_rate": 7.807887451593932e-06,
+      "loss": 0.7635,
+      "step": 5170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6995329904810377,
+      "learning_rate": 7.807029776743608e-06,
+      "loss": 0.6691,
+      "step": 5171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7065643396655594,
+      "learning_rate": 7.806171981268962e-06,
+      "loss": 0.7514,
+      "step": 5172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0486386923339865,
+      "learning_rate": 7.805314065206857e-06,
+      "loss": 0.7107,
+      "step": 5173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4009479909252358,
+      "learning_rate": 7.804456028594158e-06,
+      "loss": 0.7004,
+      "step": 5174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7765872536572047,
+      "learning_rate": 7.803597871467738e-06,
+      "loss": 0.8274,
+      "step": 5175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3250276070956455,
+      "learning_rate": 7.802739593864475e-06,
+      "loss": 0.7416,
+      "step": 5176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6160912909197538,
+      "learning_rate": 7.801881195821246e-06,
+      "loss": 0.8654,
+      "step": 5177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.661642324480457,
+      "learning_rate": 7.801022677374945e-06,
+      "loss": 0.8485,
+      "step": 5178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6095761438072695,
+      "learning_rate": 7.80016403856246e-06,
+      "loss": 0.7509,
+      "step": 5179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.116315674763856,
+      "learning_rate": 7.799305279420691e-06,
+      "loss": 0.6298,
+      "step": 5180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.117342306474534,
+      "learning_rate": 7.79844639998654e-06,
+      "loss": 0.8168,
+      "step": 5181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6315507887933498,
+      "learning_rate": 7.797587400296913e-06,
+      "loss": 0.7289,
+      "step": 5182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.280566405591747,
+      "learning_rate": 7.796728280388725e-06,
+      "loss": 0.7393,
+      "step": 5183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7068365011036426,
+      "learning_rate": 7.795869040298895e-06,
+      "loss": 0.7938,
+      "step": 5184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5487708336656911,
+      "learning_rate": 7.795009680064344e-06,
+      "loss": 0.7058,
+      "step": 5185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8070089153330833,
+      "learning_rate": 7.794150199722004e-06,
+      "loss": 0.7467,
+      "step": 5186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9657575144890773,
+      "learning_rate": 7.793290599308807e-06,
+      "loss": 0.8461,
+      "step": 5187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.513871314936803,
+      "learning_rate": 7.792430878861691e-06,
+      "loss": 0.6955,
+      "step": 5188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5884109370281443,
+      "learning_rate": 7.791571038417602e-06,
+      "loss": 0.7356,
+      "step": 5189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2272098291531586,
+      "learning_rate": 7.79071107801349e-06,
+      "loss": 0.7054,
+      "step": 5190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6203384727937176,
+      "learning_rate": 7.789850997686303e-06,
+      "loss": 0.7202,
+      "step": 5191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.561342989037223,
+      "learning_rate": 7.78899079747301e-06,
+      "loss": 0.7519,
+      "step": 5192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0147739107284672,
+      "learning_rate": 7.788130477410567e-06,
+      "loss": 0.6631,
+      "step": 5193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8491121696471824,
+      "learning_rate": 7.78727003753595e-06,
+      "loss": 0.8061,
+      "step": 5194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7343279360301875,
+      "learning_rate": 7.786409477886133e-06,
+      "loss": 0.8108,
+      "step": 5195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0001423117810906,
+      "learning_rate": 7.785548798498093e-06,
+      "loss": 0.8708,
+      "step": 5196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6137207202019452,
+      "learning_rate": 7.784687999408817e-06,
+      "loss": 0.7596,
+      "step": 5197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1845450406378983,
+      "learning_rate": 7.7838270806553e-06,
+      "loss": 0.6653,
+      "step": 5198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6994178331809067,
+      "learning_rate": 7.782966042274529e-06,
+      "loss": 0.6464,
+      "step": 5199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6952936671820733,
+      "learning_rate": 7.782104884303512e-06,
+      "loss": 0.7366,
+      "step": 5200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.700646464320394,
+      "learning_rate": 7.78124360677925e-06,
+      "loss": 0.7271,
+      "step": 5201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.62843561178569,
+      "learning_rate": 7.780382209738756e-06,
+      "loss": 0.8258,
+      "step": 5202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4267746052731554,
+      "learning_rate": 7.779520693219047e-06,
+      "loss": 0.6534,
+      "step": 5203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7092484593102808,
+      "learning_rate": 7.778659057257144e-06,
+      "loss": 0.8377,
+      "step": 5204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2353931426774662,
+      "learning_rate": 7.777797301890073e-06,
+      "loss": 0.7959,
+      "step": 5205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6057384451990322,
+      "learning_rate": 7.776935427154864e-06,
+      "loss": 0.8742,
+      "step": 5206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8505254165761267,
+      "learning_rate": 7.776073433088555e-06,
+      "loss": 0.9321,
+      "step": 5207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4837004968343015,
+      "learning_rate": 7.775211319728191e-06,
+      "loss": 0.7767,
+      "step": 5208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.713113979010448,
+      "learning_rate": 7.774349087110813e-06,
+      "loss": 0.7375,
+      "step": 5209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.648902957673403,
+      "learning_rate": 7.773486735273476e-06,
+      "loss": 0.7844,
+      "step": 5210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4791807109481923,
+      "learning_rate": 7.772624264253237e-06,
+      "loss": 0.7011,
+      "step": 5211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6290853062361093,
+      "learning_rate": 7.77176167408716e-06,
+      "loss": 0.6871,
+      "step": 5212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7355579412753794,
+      "learning_rate": 7.77089896481231e-06,
+      "loss": 0.8306,
+      "step": 5213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.19489320513279,
+      "learning_rate": 7.77003613646576e-06,
+      "loss": 0.7637,
+      "step": 5214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8273998393305766,
+      "learning_rate": 7.769173189084589e-06,
+      "loss": 0.8291,
+      "step": 5215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9312278324819305,
+      "learning_rate": 7.768310122705877e-06,
+      "loss": 0.9512,
+      "step": 5216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4054670658873905,
+      "learning_rate": 7.767446937366716e-06,
+      "loss": 0.6056,
+      "step": 5217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7843266892575707,
+      "learning_rate": 7.766583633104195e-06,
+      "loss": 0.7653,
+      "step": 5218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.864642289236451,
+      "learning_rate": 7.765720209955414e-06,
+      "loss": 0.7898,
+      "step": 5219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9145656432915268,
+      "learning_rate": 7.764856667957477e-06,
+      "loss": 0.8334,
+      "step": 5220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4462455489247419,
+      "learning_rate": 7.76399300714749e-06,
+      "loss": 0.6927,
+      "step": 5221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4499233303170707,
+      "learning_rate": 7.763129227562568e-06,
+      "loss": 0.687,
+      "step": 5222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7404331402027544,
+      "learning_rate": 7.762265329239829e-06,
+      "loss": 0.7237,
+      "step": 5223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4116017037011872,
+      "learning_rate": 7.761401312216398e-06,
+      "loss": 0.7294,
+      "step": 5224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7885611480750587,
+      "learning_rate": 7.760537176529402e-06,
+      "loss": 0.8005,
+      "step": 5225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6835489605709775,
+      "learning_rate": 7.759672922215976e-06,
+      "loss": 0.6729,
+      "step": 5226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.802296566007712,
+      "learning_rate": 7.758808549313258e-06,
+      "loss": 0.7707,
+      "step": 5227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2988116259167166,
+      "learning_rate": 7.757944057858392e-06,
+      "loss": 0.6227,
+      "step": 5228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8141394256727466,
+      "learning_rate": 7.757079447888529e-06,
+      "loss": 0.7172,
+      "step": 5229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.791705554910903,
+      "learning_rate": 7.75621471944082e-06,
+      "loss": 0.7106,
+      "step": 5230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.660725484485216,
+      "learning_rate": 7.755349872552426e-06,
+      "loss": 0.6382,
+      "step": 5231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8885894158585537,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 0.8072,
+      "step": 5232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9003294356685723,
+      "learning_rate": 7.753619823602249e-06,
+      "loss": 0.735,
+      "step": 5233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.060042520034867,
+      "learning_rate": 7.752754621614807e-06,
+      "loss": 0.8026,
+      "step": 5234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6514927781137045,
+      "learning_rate": 7.751889301335369e-06,
+      "loss": 0.7855,
+      "step": 5235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9560055880724392,
+      "learning_rate": 7.75102386280112e-06,
+      "loss": 0.5976,
+      "step": 5236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4165271052509742,
+      "learning_rate": 7.750158306049247e-06,
+      "loss": 0.7526,
+      "step": 5237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0683018201970071,
+      "learning_rate": 7.749292631116946e-06,
+      "loss": 0.6619,
+      "step": 5238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2327056872381752,
+      "learning_rate": 7.748426838041421e-06,
+      "loss": 0.6849,
+      "step": 5239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.887583013848584,
+      "learning_rate": 7.747560926859872e-06,
+      "loss": 0.7622,
+      "step": 5240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7318640973069785,
+      "learning_rate": 7.74669489760951e-06,
+      "loss": 0.8651,
+      "step": 5241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.551402664362832,
+      "learning_rate": 7.745828750327551e-06,
+      "loss": 0.7569,
+      "step": 5242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.503779154118664,
+      "learning_rate": 7.744962485051217e-06,
+      "loss": 0.7196,
+      "step": 5243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4734910006854287,
+      "learning_rate": 7.744096101817731e-06,
+      "loss": 0.7689,
+      "step": 5244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1614222758960506,
+      "learning_rate": 7.743229600664322e-06,
+      "loss": 0.6557,
+      "step": 5245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4971357484190009,
+      "learning_rate": 7.742362981628229e-06,
+      "loss": 0.6885,
+      "step": 5246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3600366516436422,
+      "learning_rate": 7.741496244746692e-06,
+      "loss": 0.7083,
+      "step": 5247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.755134338347107,
+      "learning_rate": 7.740629390056955e-06,
+      "loss": 0.7455,
+      "step": 5248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0355737408572367,
+      "learning_rate": 7.73976241759627e-06,
+      "loss": 0.6095,
+      "step": 5249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5133405035439185,
+      "learning_rate": 7.738895327401891e-06,
+      "loss": 0.8259,
+      "step": 5250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.562682506215058,
+      "learning_rate": 7.738028119511081e-06,
+      "loss": 0.7312,
+      "step": 5251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8850667311571319,
+      "learning_rate": 7.737160793961103e-06,
+      "loss": 0.65,
+      "step": 5252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5176432677667298,
+      "learning_rate": 7.736293350789232e-06,
+      "loss": 0.7704,
+      "step": 5253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2677162722695434,
+      "learning_rate": 7.73542579003274e-06,
+      "loss": 0.6481,
+      "step": 5254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8175572224315042,
+      "learning_rate": 7.734558111728911e-06,
+      "loss": 0.7583,
+      "step": 5255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.776259280247447,
+      "learning_rate": 7.733690315915028e-06,
+      "loss": 0.7642,
+      "step": 5256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6274362398097986,
+      "learning_rate": 7.732822402628385e-06,
+      "loss": 0.7663,
+      "step": 5257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7316499496670297,
+      "learning_rate": 7.731954371906279e-06,
+      "loss": 0.8428,
+      "step": 5258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5821460973310455,
+      "learning_rate": 7.731086223786006e-06,
+      "loss": 0.7099,
+      "step": 5259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9413654576199495,
+      "learning_rate": 7.730217958304876e-06,
+      "loss": 0.9236,
+      "step": 5260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9849708594318487,
+      "learning_rate": 7.729349575500202e-06,
+      "loss": 0.7126,
+      "step": 5261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7668882014160272,
+      "learning_rate": 7.728481075409297e-06,
+      "loss": 0.8834,
+      "step": 5262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4789892038891743,
+      "learning_rate": 7.727612458069482e-06,
+      "loss": 0.6402,
+      "step": 5263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5350697352302434,
+      "learning_rate": 7.726743723518087e-06,
+      "loss": 0.7468,
+      "step": 5264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.704400016135552,
+      "learning_rate": 7.72587487179244e-06,
+      "loss": 0.6895,
+      "step": 5265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1885363038156043,
+      "learning_rate": 7.72500590292988e-06,
+      "loss": 0.7124,
+      "step": 5266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7300566256787997,
+      "learning_rate": 7.724136816967746e-06,
+      "loss": 0.8615,
+      "step": 5267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.185300406562066,
+      "learning_rate": 7.723267613943388e-06,
+      "loss": 0.8431,
+      "step": 5268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6775213215611586,
+      "learning_rate": 7.722398293894153e-06,
+      "loss": 0.8287,
+      "step": 5269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5568366977175787,
+      "learning_rate": 7.721528856857403e-06,
+      "loss": 0.8255,
+      "step": 5270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5247457385889795,
+      "learning_rate": 7.720659302870496e-06,
+      "loss": 0.757,
+      "step": 5271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.70560334987983,
+      "learning_rate": 7.719789631970798e-06,
+      "loss": 0.7986,
+      "step": 5272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9634168488381167,
+      "learning_rate": 7.718919844195685e-06,
+      "loss": 0.7461,
+      "step": 5273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1012416819686244,
+      "learning_rate": 7.718049939582529e-06,
+      "loss": 0.7171,
+      "step": 5274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5628104909963896,
+      "learning_rate": 7.717179918168713e-06,
+      "loss": 0.7473,
+      "step": 5275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.577513172011016,
+      "learning_rate": 7.716309779991625e-06,
+      "loss": 0.78,
+      "step": 5276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.598151868312437,
+      "learning_rate": 7.715439525088657e-06,
+      "loss": 0.8032,
+      "step": 5277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7660848618155744,
+      "learning_rate": 7.714569153497204e-06,
+      "loss": 0.7262,
+      "step": 5278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4755356272369307,
+      "learning_rate": 7.713698665254669e-06,
+      "loss": 0.6867,
+      "step": 5279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4988556266987205,
+      "learning_rate": 7.71282806039846e-06,
+      "loss": 0.7229,
+      "step": 5280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.724634833795722,
+      "learning_rate": 7.711957338965986e-06,
+      "loss": 0.7191,
+      "step": 5281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9974464671213643,
+      "learning_rate": 7.711086500994667e-06,
+      "loss": 0.7033,
+      "step": 5282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8723337408348848,
+      "learning_rate": 7.710215546521921e-06,
+      "loss": 0.727,
+      "step": 5283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7507399284083462,
+      "learning_rate": 7.70934447558518e-06,
+      "loss": 0.8324,
+      "step": 5284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5066222209297508,
+      "learning_rate": 7.708473288221868e-06,
+      "loss": 0.7555,
+      "step": 5285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.605559385399205,
+      "learning_rate": 7.70760198446943e-06,
+      "loss": 0.7524,
+      "step": 5286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9477266966483262,
+      "learning_rate": 7.706730564365304e-06,
+      "loss": 0.6585,
+      "step": 5287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6544148125907312,
+      "learning_rate": 7.705859027946939e-06,
+      "loss": 0.9185,
+      "step": 5288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7878236333570139,
+      "learning_rate": 7.704987375251782e-06,
+      "loss": 0.7624,
+      "step": 5289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7244191531854458,
+      "learning_rate": 7.704115606317296e-06,
+      "loss": 0.7179,
+      "step": 5290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.558817161386269,
+      "learning_rate": 7.703243721180938e-06,
+      "loss": 0.7317,
+      "step": 5291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.774270083127057,
+      "learning_rate": 7.702371719880178e-06,
+      "loss": 0.871,
+      "step": 5292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6229332848766862,
+      "learning_rate": 7.701499602452487e-06,
+      "loss": 0.6772,
+      "step": 5293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6732505632749257,
+      "learning_rate": 7.70062736893534e-06,
+      "loss": 0.8253,
+      "step": 5294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4429534430045408,
+      "learning_rate": 7.699755019366221e-06,
+      "loss": 0.8105,
+      "step": 5295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9725009909686142,
+      "learning_rate": 7.698882553782617e-06,
+      "loss": 0.7134,
+      "step": 5296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7520212432728104,
+      "learning_rate": 7.698009972222017e-06,
+      "loss": 0.8342,
+      "step": 5297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5599049148718267,
+      "learning_rate": 7.697137274721922e-06,
+      "loss": 0.7194,
+      "step": 5298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.643701930149702,
+      "learning_rate": 7.696264461319831e-06,
+      "loss": 0.8489,
+      "step": 5299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8368939959723531,
+      "learning_rate": 7.69539153205325e-06,
+      "loss": 0.8753,
+      "step": 5300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6229294685824758,
+      "learning_rate": 7.694518486959695e-06,
+      "loss": 0.8088,
+      "step": 5301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0948837533903055,
+      "learning_rate": 7.693645326076677e-06,
+      "loss": 0.7459,
+      "step": 5302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8853482183351078,
+      "learning_rate": 7.692772049441718e-06,
+      "loss": 0.7622,
+      "step": 5303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6078273989083849,
+      "learning_rate": 7.69189865709235e-06,
+      "loss": 0.8973,
+      "step": 5304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7165617740805852,
+      "learning_rate": 7.691025149066103e-06,
+      "loss": 0.6204,
+      "step": 5305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2770318780600323,
+      "learning_rate": 7.69015152540051e-06,
+      "loss": 0.7605,
+      "step": 5306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6555235218450428,
+      "learning_rate": 7.689277786133113e-06,
+      "loss": 0.7424,
+      "step": 5307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5800714698385339,
+      "learning_rate": 7.688403931301462e-06,
+      "loss": 0.858,
+      "step": 5308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2470508314349709,
+      "learning_rate": 7.687529960943107e-06,
+      "loss": 0.7906,
+      "step": 5309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7385511555007578,
+      "learning_rate": 7.686655875095603e-06,
+      "loss": 0.7404,
+      "step": 5310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7012457585923542,
+      "learning_rate": 7.685781673796515e-06,
+      "loss": 0.7625,
+      "step": 5311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7275470640692994,
+      "learning_rate": 7.684907357083404e-06,
+      "loss": 0.7152,
+      "step": 5312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5784377857223681,
+      "learning_rate": 7.684032924993845e-06,
+      "loss": 0.7377,
+      "step": 5313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1295410265820025,
+      "learning_rate": 7.683158377565415e-06,
+      "loss": 0.7303,
+      "step": 5314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.122312106467841,
+      "learning_rate": 7.682283714835691e-06,
+      "loss": 0.6126,
+      "step": 5315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.239980081613424,
+      "learning_rate": 7.681408936842266e-06,
+      "loss": 0.8053,
+      "step": 5316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7075773909135536,
+      "learning_rate": 7.680534043622725e-06,
+      "loss": 0.7085,
+      "step": 5317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0257156797338096,
+      "learning_rate": 7.679659035214666e-06,
+      "loss": 0.7015,
+      "step": 5318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.620911746473289,
+      "learning_rate": 7.678783911655691e-06,
+      "loss": 0.6444,
+      "step": 5319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5732190981871987,
+      "learning_rate": 7.677908672983404e-06,
+      "loss": 0.8071,
+      "step": 5320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9878792421891145,
+      "learning_rate": 7.677033319235418e-06,
+      "loss": 0.6202,
+      "step": 5321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8678884442873291,
+      "learning_rate": 7.676157850449348e-06,
+      "loss": 0.702,
+      "step": 5322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.73569303923237,
+      "learning_rate": 7.675282266662816e-06,
+      "loss": 0.7314,
+      "step": 5323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8085583778520082,
+      "learning_rate": 7.674406567913447e-06,
+      "loss": 0.7347,
+      "step": 5324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.054644332258479,
+      "learning_rate": 7.67353075423887e-06,
+      "loss": 0.8531,
+      "step": 5325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.570983649166806,
+      "learning_rate": 7.672654825676724e-06,
+      "loss": 0.7271,
+      "step": 5326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.797068824156457,
+      "learning_rate": 7.671778782264647e-06,
+      "loss": 0.731,
+      "step": 5327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5311578140415398,
+      "learning_rate": 7.670902624040286e-06,
+      "loss": 0.8108,
+      "step": 5328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8417381030541287,
+      "learning_rate": 7.67002635104129e-06,
+      "loss": 0.8075,
+      "step": 5329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2613197458889465,
+      "learning_rate": 7.669149963305314e-06,
+      "loss": 0.6448,
+      "step": 5330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.06278738670752,
+      "learning_rate": 7.668273460870022e-06,
+      "loss": 0.7002,
+      "step": 5331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5226736033771966,
+      "learning_rate": 7.667396843773074e-06,
+      "loss": 0.7299,
+      "step": 5332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.064010822934759,
+      "learning_rate": 7.666520112052145e-06,
+      "loss": 0.716,
+      "step": 5333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7369871692612573,
+      "learning_rate": 7.66564326574491e-06,
+      "loss": 0.7101,
+      "step": 5334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.610867463694607,
+      "learning_rate": 7.664766304889044e-06,
+      "loss": 0.8476,
+      "step": 5335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.558302515519443,
+      "learning_rate": 7.663889229522235e-06,
+      "loss": 0.6516,
+      "step": 5336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7112407863196624,
+      "learning_rate": 7.663012039682174e-06,
+      "loss": 0.7755,
+      "step": 5337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5469016498642076,
+      "learning_rate": 7.662134735406555e-06,
+      "loss": 0.7385,
+      "step": 5338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6057584919250916,
+      "learning_rate": 7.661257316733078e-06,
+      "loss": 0.7537,
+      "step": 5339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5295214181660566,
+      "learning_rate": 7.660379783699446e-06,
+      "loss": 0.7215,
+      "step": 5340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8550357715834191,
+      "learning_rate": 7.65950213634337e-06,
+      "loss": 0.8145,
+      "step": 5341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9453899981885407,
+      "learning_rate": 7.658624374702565e-06,
+      "loss": 0.7269,
+      "step": 5342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6904098197875723,
+      "learning_rate": 7.657746498814749e-06,
+      "loss": 0.7793,
+      "step": 5343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.309252319966777,
+      "learning_rate": 7.656868508717648e-06,
+      "loss": 0.6247,
+      "step": 5344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8214035975320682,
+      "learning_rate": 7.655990404448988e-06,
+      "loss": 0.9321,
+      "step": 5345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6490775069527295,
+      "learning_rate": 7.655112186046508e-06,
+      "loss": 0.6628,
+      "step": 5346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.037091424870431,
+      "learning_rate": 7.65423385354794e-06,
+      "loss": 0.7039,
+      "step": 5347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5444784884300249,
+      "learning_rate": 7.653355406991034e-06,
+      "loss": 0.8138,
+      "step": 5348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8192041810783037,
+      "learning_rate": 7.652476846413537e-06,
+      "loss": 0.7629,
+      "step": 5349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.781668595252135,
+      "learning_rate": 7.651598171853203e-06,
+      "loss": 0.7657,
+      "step": 5350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8412242203855376,
+      "learning_rate": 7.65071938334779e-06,
+      "loss": 0.8248,
+      "step": 5351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.530934132113123,
+      "learning_rate": 7.64984048093506e-06,
+      "loss": 0.7842,
+      "step": 5352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4845753754801483,
+      "learning_rate": 7.648961464652784e-06,
+      "loss": 0.6934,
+      "step": 5353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5574127405128686,
+      "learning_rate": 7.648082334538735e-06,
+      "loss": 0.7745,
+      "step": 5354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5403218924593074,
+      "learning_rate": 7.64720309063069e-06,
+      "loss": 0.7445,
+      "step": 5355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.994702585693322,
+      "learning_rate": 7.646323732966433e-06,
+      "loss": 0.8643,
+      "step": 5356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6900448289711845,
+      "learning_rate": 7.645444261583752e-06,
+      "loss": 0.831,
+      "step": 5357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9565837608155046,
+      "learning_rate": 7.644564676520439e-06,
+      "loss": 0.8339,
+      "step": 5358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7777531557403414,
+      "learning_rate": 7.64368497781429e-06,
+      "loss": 0.7524,
+      "step": 5359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0803232269247882,
+      "learning_rate": 7.642805165503112e-06,
+      "loss": 0.6701,
+      "step": 5360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.318178505079482,
+      "learning_rate": 7.641925239624708e-06,
+      "loss": 0.7038,
+      "step": 5361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6824193815629256,
+      "learning_rate": 7.641045200216896e-06,
+      "loss": 0.6931,
+      "step": 5362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.699336776947092,
+      "learning_rate": 7.640165047317488e-06,
+      "loss": 0.7264,
+      "step": 5363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5261642159740927,
+      "learning_rate": 7.639284780964307e-06,
+      "loss": 0.8001,
+      "step": 5364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5536091614371488,
+      "learning_rate": 7.638404401195182e-06,
+      "loss": 0.746,
+      "step": 5365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1104896261088677,
+      "learning_rate": 7.637523908047943e-06,
+      "loss": 0.8899,
+      "step": 5366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.754484513801628,
+      "learning_rate": 7.63664330156043e-06,
+      "loss": 0.7547,
+      "step": 5367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6859043417428707,
+      "learning_rate": 7.635762581770478e-06,
+      "loss": 0.7246,
+      "step": 5368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.862096161410753,
+      "learning_rate": 7.634881748715941e-06,
+      "loss": 0.7815,
+      "step": 5369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8091123172300803,
+      "learning_rate": 7.634000802434667e-06,
+      "loss": 0.7262,
+      "step": 5370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7107732599526155,
+      "learning_rate": 7.633119742964511e-06,
+      "loss": 0.6919,
+      "step": 5371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6294292478522205,
+      "learning_rate": 7.632238570343335e-06,
+      "loss": 0.8216,
+      "step": 5372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4994420020874153,
+      "learning_rate": 7.631357284609006e-06,
+      "loss": 0.795,
+      "step": 5373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1565761622226924,
+      "learning_rate": 7.630475885799395e-06,
+      "loss": 0.7772,
+      "step": 5374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5514963324297695,
+      "learning_rate": 7.629594373952375e-06,
+      "loss": 0.7103,
+      "step": 5375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5623996156499735,
+      "learning_rate": 7.628712749105831e-06,
+      "loss": 0.6806,
+      "step": 5376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6329695958409456,
+      "learning_rate": 7.627831011297644e-06,
+      "loss": 0.7852,
+      "step": 5377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7428941408849756,
+      "learning_rate": 7.626949160565706e-06,
+      "loss": 0.8227,
+      "step": 5378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.874811823276944,
+      "learning_rate": 7.626067196947913e-06,
+      "loss": 0.7176,
+      "step": 5379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6174757653674054,
+      "learning_rate": 7.625185120482162e-06,
+      "loss": 0.7886,
+      "step": 5380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5366087061786018,
+      "learning_rate": 7.624302931206362e-06,
+      "loss": 0.7309,
+      "step": 5381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7938923388735115,
+      "learning_rate": 7.623420629158418e-06,
+      "loss": 0.6851,
+      "step": 5382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.80381826947219,
+      "learning_rate": 7.622538214376248e-06,
+      "loss": 0.6836,
+      "step": 5383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6920549413019306,
+      "learning_rate": 7.621655686897771e-06,
+      "loss": 0.7421,
+      "step": 5384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6367227670197086,
+      "learning_rate": 7.62077304676091e-06,
+      "loss": 0.7168,
+      "step": 5385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5818100886932924,
+      "learning_rate": 7.619890294003593e-06,
+      "loss": 0.748,
+      "step": 5386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7403390228992166,
+      "learning_rate": 7.6190074286637575e-06,
+      "loss": 0.673,
+      "step": 5387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8888032593231228,
+      "learning_rate": 7.618124450779338e-06,
+      "loss": 0.7603,
+      "step": 5388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8268916042814818,
+      "learning_rate": 7.617241360388282e-06,
+      "loss": 0.7676,
+      "step": 5389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6413007299433091,
+      "learning_rate": 7.616358157528535e-06,
+      "loss": 0.6716,
+      "step": 5390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.763929161567694,
+      "learning_rate": 7.615474842238051e-06,
+      "loss": 0.7747,
+      "step": 5391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6653207356381003,
+      "learning_rate": 7.614591414554787e-06,
+      "loss": 0.8476,
+      "step": 5392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5838126383095321,
+      "learning_rate": 7.613707874516708e-06,
+      "loss": 0.6868,
+      "step": 5393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.175578887681155,
+      "learning_rate": 7.612824222161781e-06,
+      "loss": 0.744,
+      "step": 5394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.79502558583993,
+      "learning_rate": 7.611940457527976e-06,
+      "loss": 0.7454,
+      "step": 5395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8210089556608338,
+      "learning_rate": 7.6110565806532745e-06,
+      "loss": 0.745,
+      "step": 5396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.573865847557262,
+      "learning_rate": 7.610172591575656e-06,
+      "loss": 0.8029,
+      "step": 5397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8694797657520594,
+      "learning_rate": 7.609288490333109e-06,
+      "loss": 0.664,
+      "step": 5398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.219741332042404,
+      "learning_rate": 7.608404276963623e-06,
+      "loss": 0.7254,
+      "step": 5399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4845429538228894,
+      "learning_rate": 7.6075199515051955e-06,
+      "loss": 0.7962,
+      "step": 5400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6234005728713001,
+      "learning_rate": 7.60663551399583e-06,
+      "loss": 0.6966,
+      "step": 5401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5947505174300391,
+      "learning_rate": 7.605750964473531e-06,
+      "loss": 0.8464,
+      "step": 5402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3142273535449003,
+      "learning_rate": 7.604866302976309e-06,
+      "loss": 0.6552,
+      "step": 5403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8488003954917243,
+      "learning_rate": 7.60398152954218e-06,
+      "loss": 0.7753,
+      "step": 5404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5344230647337056,
+      "learning_rate": 7.6030966442091665e-06,
+      "loss": 0.791,
+      "step": 5405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7918534273803297,
+      "learning_rate": 7.602211647015292e-06,
+      "loss": 0.6673,
+      "step": 5406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.468567666049839,
+      "learning_rate": 7.601326537998589e-06,
+      "loss": 0.8194,
+      "step": 5407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6585158645986302,
+      "learning_rate": 7.60044131719709e-06,
+      "loss": 0.7121,
+      "step": 5408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7929368469192442,
+      "learning_rate": 7.599555984648836e-06,
+      "loss": 0.7968,
+      "step": 5409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7606369395240564,
+      "learning_rate": 7.598670540391872e-06,
+      "loss": 0.6449,
+      "step": 5410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7552835011050172,
+      "learning_rate": 7.597784984464248e-06,
+      "loss": 0.6563,
+      "step": 5411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5490884489624523,
+      "learning_rate": 7.596899316904018e-06,
+      "loss": 0.7942,
+      "step": 5412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0530025636546985,
+      "learning_rate": 7.596013537749238e-06,
+      "loss": 0.6502,
+      "step": 5413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6017260068848205,
+      "learning_rate": 7.595127647037976e-06,
+      "loss": 0.842,
+      "step": 5414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4935524109996932,
+      "learning_rate": 7.5942416448083e-06,
+      "loss": 0.666,
+      "step": 5415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.665113955040917,
+      "learning_rate": 7.593355531098284e-06,
+      "loss": 0.6779,
+      "step": 5416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.672063659440741,
+      "learning_rate": 7.592469305946003e-06,
+      "loss": 0.6899,
+      "step": 5417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9907003718838268,
+      "learning_rate": 7.5915829693895435e-06,
+      "loss": 0.6425,
+      "step": 5418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8395741715271832,
+      "learning_rate": 7.590696521466992e-06,
+      "loss": 0.8542,
+      "step": 5419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4767087445470743,
+      "learning_rate": 7.589809962216439e-06,
+      "loss": 0.7844,
+      "step": 5420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.597406313679826,
+      "learning_rate": 7.588923291675987e-06,
+      "loss": 0.7661,
+      "step": 5421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1248557970925055,
+      "learning_rate": 7.588036509883734e-06,
+      "loss": 0.8312,
+      "step": 5422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.728495985269159,
+      "learning_rate": 7.587149616877789e-06,
+      "loss": 0.7219,
+      "step": 5423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7377020258471154,
+      "learning_rate": 7.586262612696263e-06,
+      "loss": 0.6445,
+      "step": 5424
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7261345910482768,
+      "learning_rate": 7.585375497377271e-06,
+      "loss": 0.7978,
+      "step": 5425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0027172606044323,
+      "learning_rate": 7.584488270958937e-06,
+      "loss": 0.8257,
+      "step": 5426
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1666665532762286,
+      "learning_rate": 7.583600933479387e-06,
+      "loss": 0.8244,
+      "step": 5427
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8987452872601749,
+      "learning_rate": 7.582713484976749e-06,
+      "loss": 0.8853,
+      "step": 5428
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0660905111106935,
+      "learning_rate": 7.5818259254891614e-06,
+      "loss": 0.6664,
+      "step": 5429
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8207185137333264,
+      "learning_rate": 7.580938255054765e-06,
+      "loss": 0.7682,
+      "step": 5430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.897044889346319,
+      "learning_rate": 7.580050473711702e-06,
+      "loss": 0.7353,
+      "step": 5431
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8125489848063838,
+      "learning_rate": 7.579162581498125e-06,
+      "loss": 0.8227,
+      "step": 5432
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.180248050770491,
+      "learning_rate": 7.5782745784521875e-06,
+      "loss": 0.8059,
+      "step": 5433
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6884037476137004,
+      "learning_rate": 7.577386464612049e-06,
+      "loss": 0.6739,
+      "step": 5434
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.027027833951511,
+      "learning_rate": 7.576498240015875e-06,
+      "loss": 0.9165,
+      "step": 5435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.681118040630121,
+      "learning_rate": 7.5756099047018315e-06,
+      "loss": 0.6833,
+      "step": 5436
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8488563792736243,
+      "learning_rate": 7.574721458708096e-06,
+      "loss": 0.8088,
+      "step": 5437
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7173379968561222,
+      "learning_rate": 7.573832902072844e-06,
+      "loss": 0.8223,
+      "step": 5438
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8695554687379585,
+      "learning_rate": 7.572944234834261e-06,
+      "loss": 0.8211,
+      "step": 5439
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7948643219468714,
+      "learning_rate": 7.572055457030533e-06,
+      "loss": 0.7377,
+      "step": 5440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5802979607621037,
+      "learning_rate": 7.5711665686998545e-06,
+      "loss": 0.6631,
+      "step": 5441
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.634962385349325,
+      "learning_rate": 7.570277569880422e-06,
+      "loss": 0.8027,
+      "step": 5442
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.753168260259702,
+      "learning_rate": 7.569388460610437e-06,
+      "loss": 0.7087,
+      "step": 5443
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7071081332168467,
+      "learning_rate": 7.568499240928109e-06,
+      "loss": 0.7619,
+      "step": 5444
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6678465485795877,
+      "learning_rate": 7.5676099108716485e-06,
+      "loss": 0.7707,
+      "step": 5445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.037459236605317,
+      "learning_rate": 7.5667204704792706e-06,
+      "loss": 0.717,
+      "step": 5446
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7604137887077218,
+      "learning_rate": 7.565830919789198e-06,
+      "loss": 0.7869,
+      "step": 5447
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6968514138107083,
+      "learning_rate": 7.564941258839655e-06,
+      "loss": 0.7816,
+      "step": 5448
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5875929071153805,
+      "learning_rate": 7.5640514876688765e-06,
+      "loss": 0.8616,
+      "step": 5449
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6226925468633078,
+      "learning_rate": 7.563161606315095e-06,
+      "loss": 0.7295,
+      "step": 5450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.502177630421277,
+      "learning_rate": 7.5622716148165475e-06,
+      "loss": 0.6923,
+      "step": 5451
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.033245536522205,
+      "learning_rate": 7.561381513211486e-06,
+      "loss": 0.9279,
+      "step": 5452
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.093132156978106,
+      "learning_rate": 7.5604913015381535e-06,
+      "loss": 0.6386,
+      "step": 5453
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.626067800167246,
+      "learning_rate": 7.559600979834809e-06,
+      "loss": 0.74,
+      "step": 5454
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7825846759927253,
+      "learning_rate": 7.5587105481397095e-06,
+      "loss": 0.6687,
+      "step": 5455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9504600824362377,
+      "learning_rate": 7.557820006491118e-06,
+      "loss": 0.7567,
+      "step": 5456
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8592040456517553,
+      "learning_rate": 7.5569293549273046e-06,
+      "loss": 0.9186,
+      "step": 5457
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.637154101389795,
+      "learning_rate": 7.556038593486543e-06,
+      "loss": 0.7225,
+      "step": 5458
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0977720011866823,
+      "learning_rate": 7.555147722207111e-06,
+      "loss": 0.6635,
+      "step": 5459
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6613189000846649,
+      "learning_rate": 7.554256741127291e-06,
+      "loss": 0.7714,
+      "step": 5460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7770237447113661,
+      "learning_rate": 7.553365650285368e-06,
+      "loss": 0.7426,
+      "step": 5461
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.749174175581104,
+      "learning_rate": 7.5524744497196376e-06,
+      "loss": 0.7589,
+      "step": 5462
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.761346326554504,
+      "learning_rate": 7.5515831394683954e-06,
+      "loss": 0.6071,
+      "step": 5463
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.594442777510683,
+      "learning_rate": 7.550691719569944e-06,
+      "loss": 0.7571,
+      "step": 5464
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.459099086810015,
+      "learning_rate": 7.549800190062586e-06,
+      "loss": 0.6148,
+      "step": 5465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0109514171943235,
+      "learning_rate": 7.548908550984637e-06,
+      "loss": 0.696,
+      "step": 5466
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8310015325181703,
+      "learning_rate": 7.548016802374412e-06,
+      "loss": 0.8855,
+      "step": 5467
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9415092537725225,
+      "learning_rate": 7.547124944270229e-06,
+      "loss": 0.7358,
+      "step": 5468
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8238770363873615,
+      "learning_rate": 7.546232976710413e-06,
+      "loss": 0.7468,
+      "step": 5469
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.194332913502691,
+      "learning_rate": 7.545340899733297e-06,
+      "loss": 0.6805,
+      "step": 5470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5298130136339876,
+      "learning_rate": 7.544448713377213e-06,
+      "loss": 0.7152,
+      "step": 5471
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5825601249188161,
+      "learning_rate": 7.543556417680502e-06,
+      "loss": 0.7759,
+      "step": 5472
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6723475267809602,
+      "learning_rate": 7.542664012681507e-06,
+      "loss": 0.7954,
+      "step": 5473
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7037534111584598,
+      "learning_rate": 7.541771498418575e-06,
+      "loss": 0.8744,
+      "step": 5474
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.580587930420683,
+      "learning_rate": 7.540878874930062e-06,
+      "loss": 0.7405,
+      "step": 5475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8416150980706842,
+      "learning_rate": 7.539986142254324e-06,
+      "loss": 0.7709,
+      "step": 5476
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.845025742269121,
+      "learning_rate": 7.539093300429727e-06,
+      "loss": 0.7712,
+      "step": 5477
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.550064589249128,
+      "learning_rate": 7.538200349494633e-06,
+      "loss": 0.7782,
+      "step": 5478
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.86664976749035,
+      "learning_rate": 7.537307289487419e-06,
+      "loss": 0.7969,
+      "step": 5479
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3001067685022945,
+      "learning_rate": 7.536414120446458e-06,
+      "loss": 0.6712,
+      "step": 5480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.746577471456193,
+      "learning_rate": 7.535520842410136e-06,
+      "loss": 0.7748,
+      "step": 5481
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8646649850827735,
+      "learning_rate": 7.534627455416835e-06,
+      "loss": 0.7379,
+      "step": 5482
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9015499056093028,
+      "learning_rate": 7.533733959504947e-06,
+      "loss": 0.7106,
+      "step": 5483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6937230207719582,
+      "learning_rate": 7.532840354712868e-06,
+      "loss": 0.8083,
+      "step": 5484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.86340988293661,
+      "learning_rate": 7.531946641078998e-06,
+      "loss": 0.8419,
+      "step": 5485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7841215107238293,
+      "learning_rate": 7.531052818641743e-06,
+      "loss": 0.8519,
+      "step": 5486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9834455769514627,
+      "learning_rate": 7.530158887439509e-06,
+      "loss": 0.6826,
+      "step": 5487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4806341454200076,
+      "learning_rate": 7.529264847510715e-06,
+      "loss": 0.7113,
+      "step": 5488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7808376358761966,
+      "learning_rate": 7.5283706988937765e-06,
+      "loss": 0.7278,
+      "step": 5489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1774789067452078,
+      "learning_rate": 7.5274764416271185e-06,
+      "loss": 0.6341,
+      "step": 5490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4838266325668013,
+      "learning_rate": 7.526582075749169e-06,
+      "loss": 0.6921,
+      "step": 5491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6280741174110174,
+      "learning_rate": 7.525687601298361e-06,
+      "loss": 0.7827,
+      "step": 5492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8875420512071794,
+      "learning_rate": 7.5247930183131305e-06,
+      "loss": 0.7001,
+      "step": 5493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0739910236495294,
+      "learning_rate": 7.523898326831921e-06,
+      "loss": 0.6554,
+      "step": 5494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.72163961884774,
+      "learning_rate": 7.52300352689318e-06,
+      "loss": 0.7862,
+      "step": 5495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.796724718585902,
+      "learning_rate": 7.522108618535359e-06,
+      "loss": 0.704,
+      "step": 5496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.824292713530959,
+      "learning_rate": 7.521213601796913e-06,
+      "loss": 0.7664,
+      "step": 5497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7666243177065417,
+      "learning_rate": 7.520318476716303e-06,
+      "loss": 0.7752,
+      "step": 5498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1129591626717843,
+      "learning_rate": 7.5194232433319955e-06,
+      "loss": 0.7039,
+      "step": 5499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6789487166726422,
+      "learning_rate": 7.51852790168246e-06,
+      "loss": 0.8384,
+      "step": 5500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6353366150118198,
+      "learning_rate": 7.517632451806172e-06,
+      "loss": 0.8416,
+      "step": 5501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.482770695782654,
+      "learning_rate": 7.516736893741611e-06,
+      "loss": 0.7418,
+      "step": 5502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6071893298224142,
+      "learning_rate": 7.515841227527258e-06,
+      "loss": 0.7329,
+      "step": 5503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3938870984089407,
+      "learning_rate": 7.514945453201608e-06,
+      "loss": 0.6698,
+      "step": 5504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6374661264226396,
+      "learning_rate": 7.514049570803148e-06,
+      "loss": 0.8359,
+      "step": 5505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6793475829797602,
+      "learning_rate": 7.5131535803703805e-06,
+      "loss": 0.8965,
+      "step": 5506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2442677890444958,
+      "learning_rate": 7.512257481941805e-06,
+      "loss": 0.7171,
+      "step": 5507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5542517520295485,
+      "learning_rate": 7.511361275555931e-06,
+      "loss": 0.771,
+      "step": 5508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6747921943559763,
+      "learning_rate": 7.510464961251271e-06,
+      "loss": 0.6868,
+      "step": 5509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6998560158307594,
+      "learning_rate": 7.50956853906634e-06,
+      "loss": 0.7974,
+      "step": 5510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.05523530117,
+      "learning_rate": 7.508672009039659e-06,
+      "loss": 0.8123,
+      "step": 5511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2578088033284387,
+      "learning_rate": 7.507775371209754e-06,
+      "loss": 0.691,
+      "step": 5512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7712830638527386,
+      "learning_rate": 7.5068786256151585e-06,
+      "loss": 0.8433,
+      "step": 5513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7781004485794336,
+      "learning_rate": 7.505981772294404e-06,
+      "loss": 0.7659,
+      "step": 5514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9362277957445915,
+      "learning_rate": 7.505084811286032e-06,
+      "loss": 0.7308,
+      "step": 5515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6283497254464852,
+      "learning_rate": 7.5041877426285856e-06,
+      "loss": 0.6442,
+      "step": 5516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.735031318308035,
+      "learning_rate": 7.503290566360615e-06,
+      "loss": 0.6797,
+      "step": 5517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8211074724372878,
+      "learning_rate": 7.502393282520675e-06,
+      "loss": 0.6986,
+      "step": 5518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6119413411287,
+      "learning_rate": 7.501495891147322e-06,
+      "loss": 0.716,
+      "step": 5519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.439827010743472,
+      "learning_rate": 7.500598392279119e-06,
+      "loss": 0.7324,
+      "step": 5520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.639381827047401,
+      "learning_rate": 7.499700785954633e-06,
+      "loss": 0.7687,
+      "step": 5521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1066993886954677,
+      "learning_rate": 7.498803072212439e-06,
+      "loss": 0.6057,
+      "step": 5522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7832775154035858,
+      "learning_rate": 7.49790525109111e-06,
+      "loss": 0.6846,
+      "step": 5523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.644931929177211,
+      "learning_rate": 7.497007322629231e-06,
+      "loss": 0.6672,
+      "step": 5524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6661727425967683,
+      "learning_rate": 7.496109286865384e-06,
+      "loss": 0.7421,
+      "step": 5525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.737066725749994,
+      "learning_rate": 7.495211143838164e-06,
+      "loss": 0.6735,
+      "step": 5526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3390877169730406,
+      "learning_rate": 7.494312893586163e-06,
+      "loss": 0.8488,
+      "step": 5527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5677840781064307,
+      "learning_rate": 7.493414536147983e-06,
+      "loss": 0.8472,
+      "step": 5528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5854394700270285,
+      "learning_rate": 7.492516071562226e-06,
+      "loss": 0.757,
+      "step": 5529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.747560578338676,
+      "learning_rate": 7.491617499867502e-06,
+      "loss": 0.7598,
+      "step": 5530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7620442348014727,
+      "learning_rate": 7.4907188211024254e-06,
+      "loss": 0.7982,
+      "step": 5531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.059952724411816,
+      "learning_rate": 7.489820035305613e-06,
+      "loss": 0.6883,
+      "step": 5532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.622579152378716,
+      "learning_rate": 7.488921142515691e-06,
+      "loss": 0.7725,
+      "step": 5533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.705298905289747,
+      "learning_rate": 7.488022142771282e-06,
+      "loss": 0.6324,
+      "step": 5534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1706784534425694,
+      "learning_rate": 7.4871230361110214e-06,
+      "loss": 0.6771,
+      "step": 5535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4852584880815984,
+      "learning_rate": 7.486223822573545e-06,
+      "loss": 0.7298,
+      "step": 5536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9168764862504846,
+      "learning_rate": 7.485324502197494e-06,
+      "loss": 0.6428,
+      "step": 5537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7082123318649443,
+      "learning_rate": 7.484425075021513e-06,
+      "loss": 0.7456,
+      "step": 5538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6550265768898285,
+      "learning_rate": 7.483525541084253e-06,
+      "loss": 0.8182,
+      "step": 5539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.426252395706854,
+      "learning_rate": 7.482625900424372e-06,
+      "loss": 0.6924,
+      "step": 5540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5858641687264512,
+      "learning_rate": 7.481726153080524e-06,
+      "loss": 0.8216,
+      "step": 5541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7224954249705993,
+      "learning_rate": 7.480826299091378e-06,
+      "loss": 0.8057,
+      "step": 5542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9762776198486343,
+      "learning_rate": 7.4799263384956e-06,
+      "loss": 0.5981,
+      "step": 5543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0425093041998972,
+      "learning_rate": 7.479026271331864e-06,
+      "loss": 0.6065,
+      "step": 5544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.292968969850368,
+      "learning_rate": 7.478126097638848e-06,
+      "loss": 0.6971,
+      "step": 5545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.14033792351972,
+      "learning_rate": 7.477225817455236e-06,
+      "loss": 0.8239,
+      "step": 5546
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.874039471379416,
+      "learning_rate": 7.476325430819713e-06,
+      "loss": 0.8325,
+      "step": 5547
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5092672724563523,
+      "learning_rate": 7.47542493777097e-06,
+      "loss": 0.8002,
+      "step": 5548
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6146236167872563,
+      "learning_rate": 7.4745243383477055e-06,
+      "loss": 0.7259,
+      "step": 5549
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2857996464925792,
+      "learning_rate": 7.473623632588618e-06,
+      "loss": 0.7114,
+      "step": 5550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5705040552880578,
+      "learning_rate": 7.472722820532414e-06,
+      "loss": 0.6921,
+      "step": 5551
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6092755453468297,
+      "learning_rate": 7.471821902217804e-06,
+      "loss": 0.7263,
+      "step": 5552
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7050096316149435,
+      "learning_rate": 7.470920877683503e-06,
+      "loss": 0.6879,
+      "step": 5553
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6074203197621697,
+      "learning_rate": 7.470019746968226e-06,
+      "loss": 0.7062,
+      "step": 5554
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5787256385755493,
+      "learning_rate": 7.469118510110702e-06,
+      "loss": 0.813,
+      "step": 5555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6048449105255218,
+      "learning_rate": 7.468217167149656e-06,
+      "loss": 0.7611,
+      "step": 5556
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0903087301898726,
+      "learning_rate": 7.467315718123821e-06,
+      "loss": 0.6672,
+      "step": 5557
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0317278364366573,
+      "learning_rate": 7.466414163071934e-06,
+      "loss": 0.8827,
+      "step": 5558
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0753041620218906,
+      "learning_rate": 7.4655125020327376e-06,
+      "loss": 0.6592,
+      "step": 5559
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.039485529098868,
+      "learning_rate": 7.464610735044979e-06,
+      "loss": 0.6058,
+      "step": 5560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.638967890662945,
+      "learning_rate": 7.463708862147408e-06,
+      "loss": 0.7798,
+      "step": 5561
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1000018595732453,
+      "learning_rate": 7.462806883378779e-06,
+      "loss": 0.6994,
+      "step": 5562
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5586054847513504,
+      "learning_rate": 7.4619047987778536e-06,
+      "loss": 0.787,
+      "step": 5563
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7021713582470692,
+      "learning_rate": 7.461002608383396e-06,
+      "loss": 0.8088,
+      "step": 5564
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5897353353782,
+      "learning_rate": 7.460100312234176e-06,
+      "loss": 0.6697,
+      "step": 5565
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1463755536996056,
+      "learning_rate": 7.4591979103689675e-06,
+      "loss": 0.7225,
+      "step": 5566
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3511559381041789,
+      "learning_rate": 7.458295402826546e-06,
+      "loss": 0.6967,
+      "step": 5567
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6474341776555361,
+      "learning_rate": 7.4573927896456966e-06,
+      "loss": 0.7828,
+      "step": 5568
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8769545726744143,
+      "learning_rate": 7.456490070865206e-06,
+      "loss": 0.8017,
+      "step": 5569
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5713353239681376,
+      "learning_rate": 7.455587246523868e-06,
+      "loss": 0.7207,
+      "step": 5570
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8461442761547884,
+      "learning_rate": 7.454684316660475e-06,
+      "loss": 0.7606,
+      "step": 5571
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6292786247795334,
+      "learning_rate": 7.453781281313831e-06,
+      "loss": 0.689,
+      "step": 5572
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8215273627674515,
+      "learning_rate": 7.45287814052274e-06,
+      "loss": 0.8195,
+      "step": 5573
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5332459822616495,
+      "learning_rate": 7.4519748943260126e-06,
+      "loss": 0.6899,
+      "step": 5574
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1067691669318085,
+      "learning_rate": 7.4510715427624635e-06,
+      "loss": 0.7749,
+      "step": 5575
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7236120200876601,
+      "learning_rate": 7.450168085870911e-06,
+      "loss": 0.7599,
+      "step": 5576
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2963715696988505,
+      "learning_rate": 7.4492645236901795e-06,
+      "loss": 0.69,
+      "step": 5577
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3574524937535655,
+      "learning_rate": 7.4483608562590955e-06,
+      "loss": 0.7258,
+      "step": 5578
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6434857834578736,
+      "learning_rate": 7.447457083616494e-06,
+      "loss": 0.7341,
+      "step": 5579
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.529706555667706,
+      "learning_rate": 7.4465532058012105e-06,
+      "loss": 0.7734,
+      "step": 5580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7068721273167804,
+      "learning_rate": 7.445649222852086e-06,
+      "loss": 0.7499,
+      "step": 5581
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6603624769356113,
+      "learning_rate": 7.444745134807969e-06,
+      "loss": 0.7128,
+      "step": 5582
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.698543617869762,
+      "learning_rate": 7.443840941707709e-06,
+      "loss": 0.7149,
+      "step": 5583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.015014163836794,
+      "learning_rate": 7.44293664359016e-06,
+      "loss": 0.6594,
+      "step": 5584
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.565081158767379,
+      "learning_rate": 7.442032240494184e-06,
+      "loss": 0.6783,
+      "step": 5585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.262219951135763,
+      "learning_rate": 7.441127732458642e-06,
+      "loss": 0.6514,
+      "step": 5586
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0017995943552918,
+      "learning_rate": 7.440223119522405e-06,
+      "loss": 0.8291,
+      "step": 5587
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6480292468362177,
+      "learning_rate": 7.439318401724347e-06,
+      "loss": 0.9251,
+      "step": 5588
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6369731927117468,
+      "learning_rate": 7.438413579103344e-06,
+      "loss": 0.7749,
+      "step": 5589
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.613212174661293,
+      "learning_rate": 7.4375086516982775e-06,
+      "loss": 0.6299,
+      "step": 5590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7857024233437646,
+      "learning_rate": 7.436603619548038e-06,
+      "loss": 0.7912,
+      "step": 5591
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0799417105513112,
+      "learning_rate": 7.4356984826915135e-06,
+      "loss": 0.6869,
+      "step": 5592
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6668846032634477,
+      "learning_rate": 7.434793241167601e-06,
+      "loss": 0.7315,
+      "step": 5593
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.788797122876829,
+      "learning_rate": 7.433887895015199e-06,
+      "loss": 0.8338,
+      "step": 5594
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.784277419616746,
+      "learning_rate": 7.432982444273215e-06,
+      "loss": 0.7373,
+      "step": 5595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7612081140713571,
+      "learning_rate": 7.432076888980556e-06,
+      "loss": 0.7902,
+      "step": 5596
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0196091395576365,
+      "learning_rate": 7.431171229176136e-06,
+      "loss": 0.7348,
+      "step": 5597
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7322915787200237,
+      "learning_rate": 7.4302654648988755e-06,
+      "loss": 0.7666,
+      "step": 5598
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7606958650756643,
+      "learning_rate": 7.429359596187694e-06,
+      "loss": 0.7279,
+      "step": 5599
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0639895571791667,
+      "learning_rate": 7.428453623081522e-06,
+      "loss": 0.6941,
+      "step": 5600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7224195647299136,
+      "learning_rate": 7.427547545619288e-06,
+      "loss": 0.7512,
+      "step": 5601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6673988027226079,
+      "learning_rate": 7.4266413638399284e-06,
+      "loss": 0.6899,
+      "step": 5602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.649260290145493,
+      "learning_rate": 7.425735077782387e-06,
+      "loss": 0.8257,
+      "step": 5603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5370549696527194,
+      "learning_rate": 7.424828687485606e-06,
+      "loss": 0.6751,
+      "step": 5604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6876664213751886,
+      "learning_rate": 7.423922192988536e-06,
+      "loss": 0.8291,
+      "step": 5605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7007780588519534,
+      "learning_rate": 7.423015594330131e-06,
+      "loss": 0.6982,
+      "step": 5606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1971547681650627,
+      "learning_rate": 7.422108891549349e-06,
+      "loss": 0.6708,
+      "step": 5607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.242742963399482,
+      "learning_rate": 7.421202084685155e-06,
+      "loss": 0.6568,
+      "step": 5608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2621354873505815,
+      "learning_rate": 7.420295173776515e-06,
+      "loss": 0.6271,
+      "step": 5609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6545145870845028,
+      "learning_rate": 7.4193881588624e-06,
+      "loss": 0.7317,
+      "step": 5610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6751264744311667,
+      "learning_rate": 7.418481039981787e-06,
+      "loss": 0.7552,
+      "step": 5611
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5963209028803325,
+      "learning_rate": 7.41757381717366e-06,
+      "loss": 0.7692,
+      "step": 5612
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.688428509575311,
+      "learning_rate": 7.416666490476999e-06,
+      "loss": 0.6882,
+      "step": 5613
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6393366318795997,
+      "learning_rate": 7.415759059930799e-06,
+      "loss": 0.8154,
+      "step": 5614
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7621234554901188,
+      "learning_rate": 7.41485152557405e-06,
+      "loss": 0.7293,
+      "step": 5615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7949246022709016,
+      "learning_rate": 7.4139438874457536e-06,
+      "loss": 0.7487,
+      "step": 5616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5959488249749132,
+      "learning_rate": 7.413036145584913e-06,
+      "loss": 0.8427,
+      "step": 5617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.099315075383213,
+      "learning_rate": 7.412128300030534e-06,
+      "loss": 0.6608,
+      "step": 5618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7056865804447583,
+      "learning_rate": 7.411220350821631e-06,
+      "loss": 0.6974,
+      "step": 5619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5447356649859552,
+      "learning_rate": 7.410312297997219e-06,
+      "loss": 0.723,
+      "step": 5620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6851916344129183,
+      "learning_rate": 7.409404141596319e-06,
+      "loss": 0.7269,
+      "step": 5621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8714937426676026,
+      "learning_rate": 7.4084958816579575e-06,
+      "loss": 0.727,
+      "step": 5622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7556800218742166,
+      "learning_rate": 7.407587518221164e-06,
+      "loss": 0.7915,
+      "step": 5623
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.256959214545982,
+      "learning_rate": 7.406679051324972e-06,
+      "loss": 0.6627,
+      "step": 5624
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8332627537610842,
+      "learning_rate": 7.4057704810084215e-06,
+      "loss": 0.7919,
+      "step": 5625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9105825151806548,
+      "learning_rate": 7.404861807310556e-06,
+      "loss": 0.7386,
+      "step": 5626
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3479217179688388,
+      "learning_rate": 7.4039530302704236e-06,
+      "loss": 0.6667,
+      "step": 5627
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.642310451346506,
+      "learning_rate": 7.403044149927074e-06,
+      "loss": 0.8162,
+      "step": 5628
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.15124724375951,
+      "learning_rate": 7.402135166319567e-06,
+      "loss": 0.7311,
+      "step": 5629
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6454317093786897,
+      "learning_rate": 7.40122607948696e-06,
+      "loss": 0.8041,
+      "step": 5630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7372475810363008,
+      "learning_rate": 7.400316889468322e-06,
+      "loss": 0.8894,
+      "step": 5631
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7663384996301308,
+      "learning_rate": 7.399407596302721e-06,
+      "loss": 0.6486,
+      "step": 5632
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7856734095005362,
+      "learning_rate": 7.398498200029232e-06,
+      "loss": 0.7479,
+      "step": 5633
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6752272729340756,
+      "learning_rate": 7.397588700686933e-06,
+      "loss": 0.7428,
+      "step": 5634
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7750865824088,
+      "learning_rate": 7.396679098314908e-06,
+      "loss": 0.8269,
+      "step": 5635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8461275582060048,
+      "learning_rate": 7.395769392952246e-06,
+      "loss": 0.7612,
+      "step": 5636
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7587703619393957,
+      "learning_rate": 7.394859584638035e-06,
+      "loss": 0.7479,
+      "step": 5637
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3027756401039456,
+      "learning_rate": 7.393949673411375e-06,
+      "loss": 0.7014,
+      "step": 5638
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7944658162735858,
+      "learning_rate": 7.393039659311366e-06,
+      "loss": 0.8397,
+      "step": 5639
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5051522778524316,
+      "learning_rate": 7.392129542377114e-06,
+      "loss": 0.7715,
+      "step": 5640
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6243599844802037,
+      "learning_rate": 7.391219322647728e-06,
+      "loss": 0.6874,
+      "step": 5641
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7889564843349892,
+      "learning_rate": 7.390309000162321e-06,
+      "loss": 0.8376,
+      "step": 5642
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7788404134568676,
+      "learning_rate": 7.389398574960014e-06,
+      "loss": 0.6367,
+      "step": 5643
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0750163610508008,
+      "learning_rate": 7.388488047079927e-06,
+      "loss": 0.6326,
+      "step": 5644
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.78416551079991,
+      "learning_rate": 7.387577416561191e-06,
+      "loss": 0.7012,
+      "step": 5645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.583227010696303,
+      "learning_rate": 7.386666683442936e-06,
+      "loss": 0.6869,
+      "step": 5646
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6062053372648657,
+      "learning_rate": 7.3857558477642966e-06,
+      "loss": 0.7587,
+      "step": 5647
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7069433044892728,
+      "learning_rate": 7.3848449095644155e-06,
+      "loss": 0.7093,
+      "step": 5648
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9725755154939375,
+      "learning_rate": 7.383933868882438e-06,
+      "loss": 0.8232,
+      "step": 5649
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9991195703601109,
+      "learning_rate": 7.383022725757514e-06,
+      "loss": 0.6841,
+      "step": 5650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.972203677806557,
+      "learning_rate": 7.382111480228793e-06,
+      "loss": 0.7763,
+      "step": 5651
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7033080099546072,
+      "learning_rate": 7.381200132335438e-06,
+      "loss": 0.7137,
+      "step": 5652
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1706576426380844,
+      "learning_rate": 7.38028868211661e-06,
+      "loss": 0.5743,
+      "step": 5653
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.819640712714391,
+      "learning_rate": 7.379377129611478e-06,
+      "loss": 0.7754,
+      "step": 5654
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0683965013664216,
+      "learning_rate": 7.378465474859211e-06,
+      "loss": 0.5108,
+      "step": 5655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8784940978957103,
+      "learning_rate": 7.377553717898983e-06,
+      "loss": 0.737,
+      "step": 5656
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6760018314040548,
+      "learning_rate": 7.376641858769978e-06,
+      "loss": 0.6372,
+      "step": 5657
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0680184035976263,
+      "learning_rate": 7.375729897511379e-06,
+      "loss": 0.7484,
+      "step": 5658
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0303506219895358,
+      "learning_rate": 7.374817834162378e-06,
+      "loss": 0.7359,
+      "step": 5659
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6531517760220207,
+      "learning_rate": 7.3739056687621616e-06,
+      "loss": 0.7922,
+      "step": 5660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6504639796923402,
+      "learning_rate": 7.3729934013499335e-06,
+      "loss": 0.9082,
+      "step": 5661
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5118100525522642,
+      "learning_rate": 7.372081031964893e-06,
+      "loss": 0.7072,
+      "step": 5662
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5588670295714033,
+      "learning_rate": 7.37116856064625e-06,
+      "loss": 0.7179,
+      "step": 5663
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7121582935036792,
+      "learning_rate": 7.3702559874332125e-06,
+      "loss": 0.719,
+      "step": 5664
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7029610036308642,
+      "learning_rate": 7.369343312364994e-06,
+      "loss": 0.8592,
+      "step": 5665
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6329280925773577,
+      "learning_rate": 7.368430535480818e-06,
+      "loss": 0.6511,
+      "step": 5666
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3314702059696157,
+      "learning_rate": 7.367517656819907e-06,
+      "loss": 0.7746,
+      "step": 5667
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5331359041106045,
+      "learning_rate": 7.366604676421491e-06,
+      "loss": 0.8407,
+      "step": 5668
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8772755000612968,
+      "learning_rate": 7.3656915943247984e-06,
+      "loss": 0.835,
+      "step": 5669
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9448733009417702,
+      "learning_rate": 7.364778410569071e-06,
+      "loss": 0.796,
+      "step": 5670
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6367159588288953,
+      "learning_rate": 7.363865125193549e-06,
+      "loss": 0.7165,
+      "step": 5671
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.535549399704271,
+      "learning_rate": 7.362951738237477e-06,
+      "loss": 0.8019,
+      "step": 5672
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.664335897237656,
+      "learning_rate": 7.362038249740108e-06,
+      "loss": 0.8048,
+      "step": 5673
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.936594631870382,
+      "learning_rate": 7.3611246597406925e-06,
+      "loss": 0.7601,
+      "step": 5674
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4533457750429442,
+      "learning_rate": 7.360210968278494e-06,
+      "loss": 0.8185,
+      "step": 5675
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.886118374636111,
+      "learning_rate": 7.359297175392772e-06,
+      "loss": 0.7316,
+      "step": 5676
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2198090644836657,
+      "learning_rate": 7.358383281122797e-06,
+      "loss": 0.6678,
+      "step": 5677
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4528318649743026,
+      "learning_rate": 7.35746928550784e-06,
+      "loss": 0.6764,
+      "step": 5678
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.526527771006716,
+      "learning_rate": 7.356555188587178e-06,
+      "loss": 0.6603,
+      "step": 5679
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.007014334685616,
+      "learning_rate": 7.355640990400089e-06,
+      "loss": 0.6223,
+      "step": 5680
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6452482257227254,
+      "learning_rate": 7.354726690985862e-06,
+      "loss": 0.7103,
+      "step": 5681
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.630579986878214,
+      "learning_rate": 7.353812290383785e-06,
+      "loss": 0.7988,
+      "step": 5682
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.973074740777377,
+      "learning_rate": 7.352897788633151e-06,
+      "loss": 0.7195,
+      "step": 5683
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.314763078082875,
+      "learning_rate": 7.351983185773259e-06,
+      "loss": 0.5773,
+      "step": 5684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7309091519131325,
+      "learning_rate": 7.351068481843411e-06,
+      "loss": 0.8201,
+      "step": 5685
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7797447469431356,
+      "learning_rate": 7.3501536768829154e-06,
+      "loss": 0.799,
+      "step": 5686
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0597916106184093,
+      "learning_rate": 7.349238770931081e-06,
+      "loss": 0.6166,
+      "step": 5687
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2512450167349654,
+      "learning_rate": 7.348323764027225e-06,
+      "loss": 0.8525,
+      "step": 5688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9333282165652536,
+      "learning_rate": 7.347408656210666e-06,
+      "loss": 0.5434,
+      "step": 5689
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8807508672821152,
+      "learning_rate": 7.346493447520731e-06,
+      "loss": 0.8392,
+      "step": 5690
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8374527028400367,
+      "learning_rate": 7.345578137996745e-06,
+      "loss": 0.8437,
+      "step": 5691
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2026850320326552,
+      "learning_rate": 7.344662727678043e-06,
+      "loss": 0.7256,
+      "step": 5692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7938477662545718,
+      "learning_rate": 7.343747216603962e-06,
+      "loss": 0.8588,
+      "step": 5693
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9219940110291622,
+      "learning_rate": 7.342831604813844e-06,
+      "loss": 0.752,
+      "step": 5694
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7169657086740429,
+      "learning_rate": 7.341915892347033e-06,
+      "loss": 0.7395,
+      "step": 5695
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1171865117666138,
+      "learning_rate": 7.341000079242881e-06,
+      "loss": 0.5708,
+      "step": 5696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6792821596793206,
+      "learning_rate": 7.340084165540742e-06,
+      "loss": 0.9128,
+      "step": 5697
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.139228266348063,
+      "learning_rate": 7.339168151279974e-06,
+      "loss": 0.9064,
+      "step": 5698
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7408433103990215,
+      "learning_rate": 7.338252036499941e-06,
+      "loss": 0.777,
+      "step": 5699
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8313503236492192,
+      "learning_rate": 7.3373358212400104e-06,
+      "loss": 0.732,
+      "step": 5700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5276868274182602,
+      "learning_rate": 7.336419505539555e-06,
+      "loss": 0.7503,
+      "step": 5701
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.009879883056801,
+      "learning_rate": 7.33550308943795e-06,
+      "loss": 0.7754,
+      "step": 5702
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7238316702172112,
+      "learning_rate": 7.334586572974573e-06,
+      "loss": 0.7698,
+      "step": 5703
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8255868420964503,
+      "learning_rate": 7.333669956188815e-06,
+      "loss": 0.8305,
+      "step": 5704
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6510017572678066,
+      "learning_rate": 7.332753239120061e-06,
+      "loss": 0.8152,
+      "step": 5705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0635492750208524,
+      "learning_rate": 7.331836421807704e-06,
+      "loss": 0.8335,
+      "step": 5706
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2586956457855594,
+      "learning_rate": 7.330919504291143e-06,
+      "loss": 0.6206,
+      "step": 5707
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7712622558365125,
+      "learning_rate": 7.33000248660978e-06,
+      "loss": 0.7144,
+      "step": 5708
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1694267531125484,
+      "learning_rate": 7.3290853688030196e-06,
+      "loss": 0.7408,
+      "step": 5709
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.839495045062336,
+      "learning_rate": 7.328168150910275e-06,
+      "loss": 0.7588,
+      "step": 5710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4757908412818377,
+      "learning_rate": 7.32725083297096e-06,
+      "loss": 0.6579,
+      "step": 5711
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.483678093574956,
+      "learning_rate": 7.326333415024494e-06,
+      "loss": 0.7211,
+      "step": 5712
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0223295936865686,
+      "learning_rate": 7.325415897110299e-06,
+      "loss": 0.6854,
+      "step": 5713
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5277220814535382,
+      "learning_rate": 7.324498279267803e-06,
+      "loss": 0.6754,
+      "step": 5714
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9475028097886615,
+      "learning_rate": 7.323580561536442e-06,
+      "loss": 0.6879,
+      "step": 5715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5325071800500765,
+      "learning_rate": 7.3226627439556466e-06,
+      "loss": 0.7817,
+      "step": 5716
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4331424089342841,
+      "learning_rate": 7.321744826564862e-06,
+      "loss": 0.6781,
+      "step": 5717
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.060754684879502,
+      "learning_rate": 7.3208268094035295e-06,
+      "loss": 0.7617,
+      "step": 5718
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.730639369349756,
+      "learning_rate": 7.319908692511103e-06,
+      "loss": 0.7301,
+      "step": 5719
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6822753866170035,
+      "learning_rate": 7.318990475927033e-06,
+      "loss": 0.7697,
+      "step": 5720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6887919633393693,
+      "learning_rate": 7.318072159690775e-06,
+      "loss": 0.7564,
+      "step": 5721
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6116276385443598,
+      "learning_rate": 7.317153743841795e-06,
+      "loss": 0.8128,
+      "step": 5722
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1097670979549314,
+      "learning_rate": 7.316235228419557e-06,
+      "loss": 0.6409,
+      "step": 5723
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0711126885275224,
+      "learning_rate": 7.315316613463535e-06,
+      "loss": 0.7045,
+      "step": 5724
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.571120900304084,
+      "learning_rate": 7.314397899013199e-06,
+      "loss": 0.6935,
+      "step": 5725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9841460301426093,
+      "learning_rate": 7.313479085108033e-06,
+      "loss": 0.688,
+      "step": 5726
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0262980346053352,
+      "learning_rate": 7.312560171787518e-06,
+      "loss": 0.7404,
+      "step": 5727
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.571026074088997,
+      "learning_rate": 7.311641159091141e-06,
+      "loss": 0.7909,
+      "step": 5728
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6188193568995133,
+      "learning_rate": 7.310722047058396e-06,
+      "loss": 0.6398,
+      "step": 5729
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5707077119716972,
+      "learning_rate": 7.309802835728779e-06,
+      "loss": 0.8209,
+      "step": 5730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2856734656723456,
+      "learning_rate": 7.308883525141788e-06,
+      "loss": 0.6311,
+      "step": 5731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.777384070352202,
+      "learning_rate": 7.307964115336931e-06,
+      "loss": 0.6724,
+      "step": 5732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.586941460404108,
+      "learning_rate": 7.307044606353715e-06,
+      "loss": 0.7591,
+      "step": 5733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7263732524131719,
+      "learning_rate": 7.306124998231655e-06,
+      "loss": 0.6867,
+      "step": 5734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8602252723034431,
+      "learning_rate": 7.305205291010269e-06,
+      "loss": 0.7073,
+      "step": 5735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9691253104946462,
+      "learning_rate": 7.304285484729074e-06,
+      "loss": 0.7546,
+      "step": 5736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8089929312556097,
+      "learning_rate": 7.303365579427603e-06,
+      "loss": 0.7621,
+      "step": 5737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2172679101867652,
+      "learning_rate": 7.302445575145383e-06,
+      "loss": 0.7489,
+      "step": 5738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6128669877773127,
+      "learning_rate": 7.301525471921949e-06,
+      "loss": 0.7811,
+      "step": 5739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7343484744187443,
+      "learning_rate": 7.300605269796839e-06,
+      "loss": 0.7448,
+      "step": 5740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4584622621617385,
+      "learning_rate": 7.299684968809595e-06,
+      "loss": 0.7251,
+      "step": 5741
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7464287865665442,
+      "learning_rate": 7.298764568999768e-06,
+      "loss": 0.8279,
+      "step": 5742
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6885930762890289,
+      "learning_rate": 7.297844070406908e-06,
+      "loss": 0.7111,
+      "step": 5743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8374734658875271,
+      "learning_rate": 7.296923473070571e-06,
+      "loss": 0.8448,
+      "step": 5744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6704096355104339,
+      "learning_rate": 7.296002777030316e-06,
+      "loss": 0.7844,
+      "step": 5745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7288921685840883,
+      "learning_rate": 7.295081982325708e-06,
+      "loss": 0.8163,
+      "step": 5746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3278646790530533,
+      "learning_rate": 7.2941610889963164e-06,
+      "loss": 0.8313,
+      "step": 5747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.705778060037478,
+      "learning_rate": 7.293240097081714e-06,
+      "loss": 0.7272,
+      "step": 5748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5249471119742652,
+      "learning_rate": 7.292319006621477e-06,
+      "loss": 0.7482,
+      "step": 5749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8360337532241386,
+      "learning_rate": 7.291397817655186e-06,
+      "loss": 0.8122,
+      "step": 5750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5971035457128593,
+      "learning_rate": 7.2904765302224276e-06,
+      "loss": 0.6715,
+      "step": 5751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8088041960373111,
+      "learning_rate": 7.289555144362793e-06,
+      "loss": 0.7145,
+      "step": 5752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5460952199585316,
+      "learning_rate": 7.288633660115874e-06,
+      "loss": 0.7597,
+      "step": 5753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9422185055239745,
+      "learning_rate": 7.2877120775212685e-06,
+      "loss": 0.647,
+      "step": 5754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.628361797240938,
+      "learning_rate": 7.286790396618582e-06,
+      "loss": 0.6089,
+      "step": 5755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7777475986087654,
+      "learning_rate": 7.285868617447418e-06,
+      "loss": 0.7113,
+      "step": 5756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2953872557615247,
+      "learning_rate": 7.28494674004739e-06,
+      "loss": 0.7131,
+      "step": 5757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8288870476668926,
+      "learning_rate": 7.28402476445811e-06,
+      "loss": 0.8516,
+      "step": 5758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.701421640599944,
+      "learning_rate": 7.283102690719198e-06,
+      "loss": 0.6956,
+      "step": 5759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8876799592058886,
+      "learning_rate": 7.282180518870281e-06,
+      "loss": 0.7424,
+      "step": 5760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9712477368044772,
+      "learning_rate": 7.2812582489509844e-06,
+      "loss": 0.832,
+      "step": 5761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1925499361656584,
+      "learning_rate": 7.28033588100094e-06,
+      "loss": 0.6194,
+      "step": 5762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6965879009038602,
+      "learning_rate": 7.279413415059782e-06,
+      "loss": 0.7383,
+      "step": 5763
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6126636357084323,
+      "learning_rate": 7.278490851167155e-06,
+      "loss": 0.6757,
+      "step": 5764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.322389253155655,
+      "learning_rate": 7.277568189362701e-06,
+      "loss": 0.7553,
+      "step": 5765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9357333040395976,
+      "learning_rate": 7.276645429686071e-06,
+      "loss": 0.8453,
+      "step": 5766
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6572937145757103,
+      "learning_rate": 7.275722572176915e-06,
+      "loss": 0.841,
+      "step": 5767
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5245048292842136,
+      "learning_rate": 7.2747996168748915e-06,
+      "loss": 0.8651,
+      "step": 5768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7138903938548808,
+      "learning_rate": 7.2738765638196625e-06,
+      "loss": 0.8342,
+      "step": 5769
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7717382132873098,
+      "learning_rate": 7.272953413050894e-06,
+      "loss": 0.7658,
+      "step": 5770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3561047307640366,
+      "learning_rate": 7.272030164608256e-06,
+      "loss": 0.6801,
+      "step": 5771
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8489550660240286,
+      "learning_rate": 7.27110681853142e-06,
+      "loss": 0.7546,
+      "step": 5772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8713021072317804,
+      "learning_rate": 7.270183374860066e-06,
+      "loss": 0.6896,
+      "step": 5773
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1237556929155565,
+      "learning_rate": 7.269259833633877e-06,
+      "loss": 0.8204,
+      "step": 5774
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.416518660278377,
+      "learning_rate": 7.26833619489254e-06,
+      "loss": 0.6792,
+      "step": 5775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9931999532042122,
+      "learning_rate": 7.267412458675744e-06,
+      "loss": 0.5804,
+      "step": 5776
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7020266250202796,
+      "learning_rate": 7.266488625023184e-06,
+      "loss": 0.8885,
+      "step": 5777
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.698788980436472,
+      "learning_rate": 7.265564693974561e-06,
+      "loss": 0.7063,
+      "step": 5778
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8334966678494309,
+      "learning_rate": 7.264640665569577e-06,
+      "loss": 0.6982,
+      "step": 5779
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1787586779451127,
+      "learning_rate": 7.2637165398479405e-06,
+      "loss": 0.6494,
+      "step": 5780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9365330772470781,
+      "learning_rate": 7.262792316849362e-06,
+      "loss": 0.8942,
+      "step": 5781
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7669044391057005,
+      "learning_rate": 7.261867996613559e-06,
+      "loss": 0.7569,
+      "step": 5782
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6558335586278947,
+      "learning_rate": 7.26094357918025e-06,
+      "loss": 0.8192,
+      "step": 5783
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7924727716169473,
+      "learning_rate": 7.26001906458916e-06,
+      "loss": 0.8284,
+      "step": 5784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6469532227258665,
+      "learning_rate": 7.259094452880019e-06,
+      "loss": 0.8552,
+      "step": 5785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7511180158747977,
+      "learning_rate": 7.258169744092556e-06,
+      "loss": 0.9016,
+      "step": 5786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8739074084126934,
+      "learning_rate": 7.257244938266511e-06,
+      "loss": 0.6511,
+      "step": 5787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8336551526868594,
+      "learning_rate": 7.256320035441623e-06,
+      "loss": 0.8095,
+      "step": 5788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9368444473105557,
+      "learning_rate": 7.255395035657639e-06,
+      "loss": 0.7683,
+      "step": 5789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5688109132723875,
+      "learning_rate": 7.254469938954307e-06,
+      "loss": 0.6971,
+      "step": 5790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9072860727663632,
+      "learning_rate": 7.25354474537138e-06,
+      "loss": 0.7144,
+      "step": 5791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0666543671032258,
+      "learning_rate": 7.252619454948617e-06,
+      "loss": 0.5942,
+      "step": 5792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6895577708499774,
+      "learning_rate": 7.251694067725779e-06,
+      "loss": 0.7827,
+      "step": 5793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6217810364855096,
+      "learning_rate": 7.250768583742634e-06,
+      "loss": 0.7895,
+      "step": 5794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0385945995454113,
+      "learning_rate": 7.249843003038948e-06,
+      "loss": 0.7156,
+      "step": 5795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.891249671184688,
+      "learning_rate": 7.2489173256544975e-06,
+      "loss": 0.7487,
+      "step": 5796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.66266900736411,
+      "learning_rate": 7.247991551629061e-06,
+      "loss": 0.7645,
+      "step": 5797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.581284576590573,
+      "learning_rate": 7.247065681002421e-06,
+      "loss": 0.6748,
+      "step": 5798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7341016182612612,
+      "learning_rate": 7.246139713814365e-06,
+      "loss": 0.6623,
+      "step": 5799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.339769252311597,
+      "learning_rate": 7.245213650104683e-06,
+      "loss": 0.6219,
+      "step": 5800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.648442454566744,
+      "learning_rate": 7.2442874899131685e-06,
+      "loss": 0.8004,
+      "step": 5801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4961319004060034,
+      "learning_rate": 7.2433612332796245e-06,
+      "loss": 0.6732,
+      "step": 5802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.533472845234504,
+      "learning_rate": 7.242434880243851e-06,
+      "loss": 0.672,
+      "step": 5803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6494176832715273,
+      "learning_rate": 7.241508430845656e-06,
+      "loss": 0.7646,
+      "step": 5804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7168389928130914,
+      "learning_rate": 7.2405818851248535e-06,
+      "loss": 0.6578,
+      "step": 5805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.653977292741349,
+      "learning_rate": 7.239655243121255e-06,
+      "loss": 0.7168,
+      "step": 5806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3195988046701486,
+      "learning_rate": 7.238728504874683e-06,
+      "loss": 0.6863,
+      "step": 5807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4663873021088383,
+      "learning_rate": 7.237801670424963e-06,
+      "loss": 0.741,
+      "step": 5808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5316203886201154,
+      "learning_rate": 7.236874739811921e-06,
+      "loss": 0.7164,
+      "step": 5809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.25761718309745,
+      "learning_rate": 7.235947713075389e-06,
+      "loss": 0.7928,
+      "step": 5810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5153294822498313,
+      "learning_rate": 7.235020590255204e-06,
+      "loss": 0.8627,
+      "step": 5811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7157391730645328,
+      "learning_rate": 7.234093371391206e-06,
+      "loss": 0.7396,
+      "step": 5812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5361828413097496,
+      "learning_rate": 7.233166056523242e-06,
+      "loss": 0.7545,
+      "step": 5813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7072537862919395,
+      "learning_rate": 7.232238645691157e-06,
+      "loss": 0.6694,
+      "step": 5814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5991437939335416,
+      "learning_rate": 7.231311138934807e-06,
+      "loss": 0.6992,
+      "step": 5815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6007708670755532,
+      "learning_rate": 7.230383536294048e-06,
+      "loss": 0.6605,
+      "step": 5816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7145068810996615,
+      "learning_rate": 7.229455837808741e-06,
+      "loss": 0.6895,
+      "step": 5817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.154826877284134,
+      "learning_rate": 7.228528043518752e-06,
+      "loss": 0.8054,
+      "step": 5818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.145954706861147,
+      "learning_rate": 7.227600153463947e-06,
+      "loss": 0.7693,
+      "step": 5819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.678448176839132,
+      "learning_rate": 7.226672167684205e-06,
+      "loss": 0.8223,
+      "step": 5820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5481429271344553,
+      "learning_rate": 7.225744086219402e-06,
+      "loss": 0.7032,
+      "step": 5821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0496359412836287,
+      "learning_rate": 7.224815909109416e-06,
+      "loss": 0.6753,
+      "step": 5822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.802605701321641,
+      "learning_rate": 7.223887636394137e-06,
+      "loss": 0.7893,
+      "step": 5823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6169838535664798,
+      "learning_rate": 7.222959268113452e-06,
+      "loss": 0.79,
+      "step": 5824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6627425495923207,
+      "learning_rate": 7.222030804307257e-06,
+      "loss": 0.6909,
+      "step": 5825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6504914289522516,
+      "learning_rate": 7.22110224501545e-06,
+      "loss": 0.7425,
+      "step": 5826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1144614583991161,
+      "learning_rate": 7.220173590277933e-06,
+      "loss": 0.7578,
+      "step": 5827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5911130303996899,
+      "learning_rate": 7.219244840134612e-06,
+      "loss": 0.6963,
+      "step": 5828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7015537250658492,
+      "learning_rate": 7.218315994625397e-06,
+      "loss": 0.7711,
+      "step": 5829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7732972781941947,
+      "learning_rate": 7.217387053790204e-06,
+      "loss": 0.862,
+      "step": 5830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8515073045286077,
+      "learning_rate": 7.216458017668951e-06,
+      "loss": 0.6837,
+      "step": 5831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.214985361127954,
+      "learning_rate": 7.21552888630156e-06,
+      "loss": 0.6929,
+      "step": 5832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.540902816470525,
+      "learning_rate": 7.214599659727959e-06,
+      "loss": 0.6444,
+      "step": 5833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.775810735733069,
+      "learning_rate": 7.213670337988079e-06,
+      "loss": 0.6693,
+      "step": 5834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7085295997642593,
+      "learning_rate": 7.2127409211218526e-06,
+      "loss": 0.781,
+      "step": 5835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6708016145683857,
+      "learning_rate": 7.211811409169222e-06,
+      "loss": 0.8638,
+      "step": 5836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5208239932708616,
+      "learning_rate": 7.21088180217013e-06,
+      "loss": 0.7731,
+      "step": 5837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6278397425693338,
+      "learning_rate": 7.2099521001645225e-06,
+      "loss": 0.787,
+      "step": 5838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6116715863220323,
+      "learning_rate": 7.209022303192351e-06,
+      "loss": 0.6672,
+      "step": 5839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3072054934789632,
+      "learning_rate": 7.208092411293572e-06,
+      "loss": 0.6253,
+      "step": 5840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.165708061682108,
+      "learning_rate": 7.207162424508144e-06,
+      "loss": 0.7233,
+      "step": 5841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9807420110980318,
+      "learning_rate": 7.206232342876031e-06,
+      "loss": 0.6908,
+      "step": 5842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.875842350994876,
+      "learning_rate": 7.2053021664372e-06,
+      "loss": 0.7371,
+      "step": 5843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4437929442431108,
+      "learning_rate": 7.204371895231623e-06,
+      "loss": 0.7545,
+      "step": 5844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7332838881629398,
+      "learning_rate": 7.20344152929928e-06,
+      "loss": 0.8761,
+      "step": 5845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7430689900254857,
+      "learning_rate": 7.202511068680143e-06,
+      "loss": 0.7185,
+      "step": 5846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7573903932599995,
+      "learning_rate": 7.201580513414201e-06,
+      "loss": 0.7475,
+      "step": 5847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.801526721497375,
+      "learning_rate": 7.200649863541442e-06,
+      "loss": 0.636,
+      "step": 5848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5763346442617743,
+      "learning_rate": 7.199719119101858e-06,
+      "loss": 0.8169,
+      "step": 5849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8574629423817093,
+      "learning_rate": 7.198788280135443e-06,
+      "loss": 0.8705,
+      "step": 5850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7631373723023624,
+      "learning_rate": 7.197857346682199e-06,
+      "loss": 0.76,
+      "step": 5851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1129353073180097,
+      "learning_rate": 7.19692631878213e-06,
+      "loss": 0.7126,
+      "step": 5852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7303240652603964,
+      "learning_rate": 7.195995196475244e-06,
+      "loss": 0.8081,
+      "step": 5853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6445850886410953,
+      "learning_rate": 7.195063979801554e-06,
+      "loss": 0.8253,
+      "step": 5854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2664037074782684,
+      "learning_rate": 7.194132668801077e-06,
+      "loss": 0.6502,
+      "step": 5855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.754956952686219,
+      "learning_rate": 7.193201263513831e-06,
+      "loss": 0.7259,
+      "step": 5856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7234176542048223,
+      "learning_rate": 7.192269763979844e-06,
+      "loss": 0.7933,
+      "step": 5857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6532308606073196,
+      "learning_rate": 7.1913381702391405e-06,
+      "loss": 0.7619,
+      "step": 5858
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9094470269547126,
+      "learning_rate": 7.190406482331757e-06,
+      "loss": 0.7339,
+      "step": 5859
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8208947495189771,
+      "learning_rate": 7.189474700297728e-06,
+      "loss": 0.7441,
+      "step": 5860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.781695001290336,
+      "learning_rate": 7.188542824177096e-06,
+      "loss": 0.6857,
+      "step": 5861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0941868074296903,
+      "learning_rate": 7.187610854009903e-06,
+      "loss": 0.5253,
+      "step": 5862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.040572193747426,
+      "learning_rate": 7.186678789836201e-06,
+      "loss": 0.7518,
+      "step": 5863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7369257674167737,
+      "learning_rate": 7.18574663169604e-06,
+      "loss": 0.7435,
+      "step": 5864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4494965162991338,
+      "learning_rate": 7.18481437962948e-06,
+      "loss": 0.748,
+      "step": 5865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6041487787070106,
+      "learning_rate": 7.183882033676579e-06,
+      "loss": 0.614,
+      "step": 5866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3179193686730144,
+      "learning_rate": 7.182949593877404e-06,
+      "loss": 0.708,
+      "step": 5867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.654817509379598,
+      "learning_rate": 7.182017060272023e-06,
+      "loss": 0.745,
+      "step": 5868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6901384938857555,
+      "learning_rate": 7.1810844329005095e-06,
+      "loss": 0.6633,
+      "step": 5869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7312379672079294,
+      "learning_rate": 7.180151711802941e-06,
+      "loss": 0.8032,
+      "step": 5870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7162194105622481,
+      "learning_rate": 7.179218897019395e-06,
+      "loss": 0.7419,
+      "step": 5871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5849646518948488,
+      "learning_rate": 7.178285988589963e-06,
+      "loss": 0.6664,
+      "step": 5872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6282835872635306,
+      "learning_rate": 7.177352986554729e-06,
+      "loss": 0.6946,
+      "step": 5873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9127071942520575,
+      "learning_rate": 7.176419890953788e-06,
+      "loss": 0.7691,
+      "step": 5874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0838226013506578,
+      "learning_rate": 7.175486701827239e-06,
+      "loss": 0.6938,
+      "step": 5875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9615361807269596,
+      "learning_rate": 7.174553419215179e-06,
+      "loss": 0.7581,
+      "step": 5876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0757423119005223,
+      "learning_rate": 7.173620043157716e-06,
+      "loss": 0.58,
+      "step": 5877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1973161474983283,
+      "learning_rate": 7.17268657369496e-06,
+      "loss": 0.6448,
+      "step": 5878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5157916141284948,
+      "learning_rate": 7.171753010867023e-06,
+      "loss": 0.8114,
+      "step": 5879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9473373938648995,
+      "learning_rate": 7.1708193547140205e-06,
+      "loss": 0.8132,
+      "step": 5880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6875077902811102,
+      "learning_rate": 7.169885605276077e-06,
+      "loss": 0.6688,
+      "step": 5881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6802787131599055,
+      "learning_rate": 7.1689517625933165e-06,
+      "loss": 0.8051,
+      "step": 5882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0169384496150164,
+      "learning_rate": 7.168017826705869e-06,
+      "loss": 0.5887,
+      "step": 5883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6025277997808436,
+      "learning_rate": 7.167083797653866e-06,
+      "loss": 0.7428,
+      "step": 5884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7088674682948408,
+      "learning_rate": 7.166149675477445e-06,
+      "loss": 0.6686,
+      "step": 5885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5838006555645914,
+      "learning_rate": 7.16521546021675e-06,
+      "loss": 0.777,
+      "step": 5886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7277105858502924,
+      "learning_rate": 7.164281151911923e-06,
+      "loss": 0.7062,
+      "step": 5887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6111549623701715,
+      "learning_rate": 7.163346750603117e-06,
+      "loss": 0.7087,
+      "step": 5888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7360261611291683,
+      "learning_rate": 7.162412256330481e-06,
+      "loss": 0.6942,
+      "step": 5889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7496205082532417,
+      "learning_rate": 7.161477669134174e-06,
+      "loss": 0.6788,
+      "step": 5890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8155349956098217,
+      "learning_rate": 7.160542989054359e-06,
+      "loss": 0.7125,
+      "step": 5891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2527472645075328,
+      "learning_rate": 7.1596082161312005e-06,
+      "loss": 0.6758,
+      "step": 5892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8694272280478508,
+      "learning_rate": 7.158673350404866e-06,
+      "loss": 0.7564,
+      "step": 5893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.765656511521039,
+      "learning_rate": 7.157738391915531e-06,
+      "loss": 0.721,
+      "step": 5894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.781046498444235,
+      "learning_rate": 7.156803340703371e-06,
+      "loss": 0.885,
+      "step": 5895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7963376611890487,
+      "learning_rate": 7.155868196808569e-06,
+      "loss": 0.7643,
+      "step": 5896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.70431164077882,
+      "learning_rate": 7.154932960271308e-06,
+      "loss": 0.799,
+      "step": 5897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9990567794637804,
+      "learning_rate": 7.153997631131778e-06,
+      "loss": 0.7275,
+      "step": 5898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.150024623168937,
+      "learning_rate": 7.153062209430174e-06,
+      "loss": 0.7306,
+      "step": 5899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8279108936172908,
+      "learning_rate": 7.152126695206689e-06,
+      "loss": 0.7912,
+      "step": 5900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0996334928116596,
+      "learning_rate": 7.151191088501531e-06,
+      "loss": 0.6383,
+      "step": 5901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7860812587385175,
+      "learning_rate": 7.150255389354899e-06,
+      "loss": 0.7571,
+      "step": 5902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.991386432683699,
+      "learning_rate": 7.149319597807003e-06,
+      "loss": 0.8803,
+      "step": 5903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7574786801763498,
+      "learning_rate": 7.148383713898058e-06,
+      "loss": 0.733,
+      "step": 5904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6584989823939014,
+      "learning_rate": 7.14744773766828e-06,
+      "loss": 0.7501,
+      "step": 5905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6211875803907636,
+      "learning_rate": 7.146511669157892e-06,
+      "loss": 0.6843,
+      "step": 5906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8524179217887542,
+      "learning_rate": 7.145575508407114e-06,
+      "loss": 0.8265,
+      "step": 5907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4288264954111405,
+      "learning_rate": 7.14463925545618e-06,
+      "loss": 0.6957,
+      "step": 5908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.100110434154817,
+      "learning_rate": 7.143702910345318e-06,
+      "loss": 0.6485,
+      "step": 5909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.195744262135385,
+      "learning_rate": 7.142766473114771e-06,
+      "loss": 0.7227,
+      "step": 5910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8302273904752255,
+      "learning_rate": 7.141829943804777e-06,
+      "loss": 0.8086,
+      "step": 5911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4948747607801631,
+      "learning_rate": 7.1408933224555765e-06,
+      "loss": 0.7842,
+      "step": 5912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6172927261406678,
+      "learning_rate": 7.139956609107425e-06,
+      "loss": 0.8713,
+      "step": 5913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1807830608336298,
+      "learning_rate": 7.139019803800569e-06,
+      "loss": 0.6208,
+      "step": 5914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.108667876425122,
+      "learning_rate": 7.138082906575271e-06,
+      "loss": 0.9016,
+      "step": 5915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6835932629237713,
+      "learning_rate": 7.137145917471788e-06,
+      "loss": 0.8256,
+      "step": 5916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.628333019591827,
+      "learning_rate": 7.136208836530385e-06,
+      "loss": 0.7869,
+      "step": 5917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9535363587731713,
+      "learning_rate": 7.13527166379133e-06,
+      "loss": 0.7668,
+      "step": 5918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.478454831560235,
+      "learning_rate": 7.134334399294897e-06,
+      "loss": 0.8182,
+      "step": 5919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.967475393246498,
+      "learning_rate": 7.1333970430813615e-06,
+      "loss": 0.7391,
+      "step": 5920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7308648261152726,
+      "learning_rate": 7.132459595191003e-06,
+      "loss": 0.6751,
+      "step": 5921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2668319214685482,
+      "learning_rate": 7.131522055664109e-06,
+      "loss": 0.668,
+      "step": 5922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5543764051649271,
+      "learning_rate": 7.13058442454096e-06,
+      "loss": 0.6827,
+      "step": 5923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7554820631426662,
+      "learning_rate": 7.129646701861858e-06,
+      "loss": 0.8824,
+      "step": 5924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.269445421101455,
+      "learning_rate": 7.128708887667093e-06,
+      "loss": 0.663,
+      "step": 5925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8599178669682708,
+      "learning_rate": 7.127770981996966e-06,
+      "loss": 0.7466,
+      "step": 5926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.565517515056074,
+      "learning_rate": 7.1268329848917795e-06,
+      "loss": 0.7048,
+      "step": 5927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4234455867689007,
+      "learning_rate": 7.125894896391845e-06,
+      "loss": 0.6321,
+      "step": 5928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.148980742067566,
+      "learning_rate": 7.124956716537471e-06,
+      "loss": 0.5781,
+      "step": 5929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6083022844044303,
+      "learning_rate": 7.124018445368975e-06,
+      "loss": 0.7702,
+      "step": 5930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.713094809907016,
+      "learning_rate": 7.123080082926677e-06,
+      "loss": 0.6646,
+      "step": 5931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8594534082486323,
+      "learning_rate": 7.122141629250896e-06,
+      "loss": 0.7914,
+      "step": 5932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2527194690305048,
+      "learning_rate": 7.1212030843819656e-06,
+      "loss": 0.6816,
+      "step": 5933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8479456271681078,
+      "learning_rate": 7.120264448360214e-06,
+      "loss": 0.8024,
+      "step": 5934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.436772909817573,
+      "learning_rate": 7.1193257212259784e-06,
+      "loss": 0.7581,
+      "step": 5935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.055086587992074,
+      "learning_rate": 7.118386903019594e-06,
+      "loss": 0.7048,
+      "step": 5936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5907628747659677,
+      "learning_rate": 7.117447993781408e-06,
+      "loss": 0.8398,
+      "step": 5937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0825251081205867,
+      "learning_rate": 7.116508993551765e-06,
+      "loss": 0.6216,
+      "step": 5938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9026915575248236,
+      "learning_rate": 7.115569902371018e-06,
+      "loss": 0.9196,
+      "step": 5939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.035722146199551,
+      "learning_rate": 7.114630720279521e-06,
+      "loss": 0.8137,
+      "step": 5940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7419748134975936,
+      "learning_rate": 7.11369144731763e-06,
+      "loss": 0.7053,
+      "step": 5941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1860132413909206,
+      "learning_rate": 7.1127520835257125e-06,
+      "loss": 0.7746,
+      "step": 5942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9260050369859814,
+      "learning_rate": 7.111812628944132e-06,
+      "loss": 0.7319,
+      "step": 5943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8095726152195948,
+      "learning_rate": 7.110873083613259e-06,
+      "loss": 0.7671,
+      "step": 5944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.84258023086613,
+      "learning_rate": 7.109933447573469e-06,
+      "loss": 0.758,
+      "step": 5945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7721316006144692,
+      "learning_rate": 7.108993720865138e-06,
+      "loss": 0.8096,
+      "step": 5946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6539757381577618,
+      "learning_rate": 7.108053903528652e-06,
+      "loss": 0.6638,
+      "step": 5947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1039791891933635,
+      "learning_rate": 7.107113995604395e-06,
+      "loss": 0.6861,
+      "step": 5948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5580281142634744,
+      "learning_rate": 7.106173997132755e-06,
+      "loss": 0.6554,
+      "step": 5949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8628568977961013,
+      "learning_rate": 7.105233908154128e-06,
+      "loss": 0.7513,
+      "step": 5950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6095053071929075,
+      "learning_rate": 7.104293728708912e-06,
+      "loss": 0.62,
+      "step": 5951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7518135467965508,
+      "learning_rate": 7.103353458837509e-06,
+      "loss": 0.6854,
+      "step": 5952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9210342438841796,
+      "learning_rate": 7.102413098580322e-06,
+      "loss": 0.8122,
+      "step": 5953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9965124935404197,
+      "learning_rate": 7.101472647977761e-06,
+      "loss": 0.7815,
+      "step": 5954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6340200127121765,
+      "learning_rate": 7.100532107070242e-06,
+      "loss": 0.6922,
+      "step": 5955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8003873734089157,
+      "learning_rate": 7.099591475898178e-06,
+      "loss": 0.6824,
+      "step": 5956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.671224718475806,
+      "learning_rate": 7.098650754501994e-06,
+      "loss": 0.6668,
+      "step": 5957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2721983846945883,
+      "learning_rate": 7.0977099429221145e-06,
+      "loss": 0.6884,
+      "step": 5958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3182880296655406,
+      "learning_rate": 7.096769041198964e-06,
+      "loss": 0.7269,
+      "step": 5959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8683992708583892,
+      "learning_rate": 7.095828049372979e-06,
+      "loss": 0.811,
+      "step": 5960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.622706762721094,
+      "learning_rate": 7.0948869674845955e-06,
+      "loss": 0.8157,
+      "step": 5961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6647597319332927,
+      "learning_rate": 7.093945795574254e-06,
+      "loss": 0.7431,
+      "step": 5962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8409901018416779,
+      "learning_rate": 7.093004533682397e-06,
+      "loss": 0.7224,
+      "step": 5963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5198498514524803,
+      "learning_rate": 7.0920631818494745e-06,
+      "loss": 0.6885,
+      "step": 5964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6011124623669746,
+      "learning_rate": 7.091121740115936e-06,
+      "loss": 0.6534,
+      "step": 5965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8808026732996386,
+      "learning_rate": 7.090180208522241e-06,
+      "loss": 0.7525,
+      "step": 5966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1416667157817115,
+      "learning_rate": 7.089238587108849e-06,
+      "loss": 0.6558,
+      "step": 5967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7489075651868835,
+      "learning_rate": 7.08829687591622e-06,
+      "loss": 0.7199,
+      "step": 5968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7606346638999135,
+      "learning_rate": 7.087355074984823e-06,
+      "loss": 0.8044,
+      "step": 5969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.628052050648961,
+      "learning_rate": 7.086413184355129e-06,
+      "loss": 0.7366,
+      "step": 5970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6523859446558373,
+      "learning_rate": 7.085471204067616e-06,
+      "loss": 0.8568,
+      "step": 5971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7747188028011174,
+      "learning_rate": 7.084529134162759e-06,
+      "loss": 0.7182,
+      "step": 5972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9679068517065241,
+      "learning_rate": 7.083586974681042e-06,
+      "loss": 0.6871,
+      "step": 5973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9399745662920072,
+      "learning_rate": 7.082644725662954e-06,
+      "loss": 0.8004,
+      "step": 5974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7800160998496077,
+      "learning_rate": 7.0817023871489835e-06,
+      "loss": 0.8622,
+      "step": 5975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.750302052624048,
+      "learning_rate": 7.080759959179626e-06,
+      "loss": 0.6918,
+      "step": 5976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.576581602167211,
+      "learning_rate": 7.079817441795378e-06,
+      "loss": 0.7914,
+      "step": 5977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.083683195587729,
+      "learning_rate": 7.078874835036742e-06,
+      "loss": 0.7917,
+      "step": 5978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.683519188839088,
+      "learning_rate": 7.077932138944225e-06,
+      "loss": 0.6925,
+      "step": 5979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.636100020319709,
+      "learning_rate": 7.076989353558335e-06,
+      "loss": 0.8001,
+      "step": 5980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4690331965345091,
+      "learning_rate": 7.076046478919589e-06,
+      "loss": 0.7311,
+      "step": 5981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7873434126542767,
+      "learning_rate": 7.075103515068501e-06,
+      "loss": 0.7813,
+      "step": 5982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5609698563937082,
+      "learning_rate": 7.074160462045593e-06,
+      "loss": 0.6976,
+      "step": 5983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6473878273610385,
+      "learning_rate": 7.073217319891391e-06,
+      "loss": 0.6363,
+      "step": 5984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8526993747713432,
+      "learning_rate": 7.072274088646425e-06,
+      "loss": 0.6906,
+      "step": 5985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0855255246875624,
+      "learning_rate": 7.071330768351223e-06,
+      "loss": 0.7598,
+      "step": 5986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8575754759307137,
+      "learning_rate": 7.070387359046328e-06,
+      "loss": 0.8288,
+      "step": 5987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9103146622587581,
+      "learning_rate": 7.0694438607722746e-06,
+      "loss": 0.7048,
+      "step": 5988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.829693967835356,
+      "learning_rate": 7.068500273569612e-06,
+      "loss": 0.7629,
+      "step": 5989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.053003596669567,
+      "learning_rate": 7.067556597478884e-06,
+      "loss": 0.7389,
+      "step": 5990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7015061920001358,
+      "learning_rate": 7.066612832540644e-06,
+      "loss": 0.6505,
+      "step": 5991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1526112064878589,
+      "learning_rate": 7.065668978795449e-06,
+      "loss": 0.7284,
+      "step": 5992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8114145210064705,
+      "learning_rate": 7.064725036283857e-06,
+      "loss": 0.7675,
+      "step": 5993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4037291634075244,
+      "learning_rate": 7.063781005046433e-06,
+      "loss": 0.7579,
+      "step": 5994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6324515944634415,
+      "learning_rate": 7.0628368851237424e-06,
+      "loss": 0.7734,
+      "step": 5995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8054656899183739,
+      "learning_rate": 7.061892676556357e-06,
+      "loss": 0.7381,
+      "step": 5996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.810718087793561,
+      "learning_rate": 7.06094837938485e-06,
+      "loss": 0.7476,
+      "step": 5997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6881073810470522,
+      "learning_rate": 7.060003993649802e-06,
+      "loss": 0.8225,
+      "step": 5998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1403732373836264,
+      "learning_rate": 7.059059519391794e-06,
+      "loss": 0.7778,
+      "step": 5999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4507316793336293,
+      "learning_rate": 7.058114956651413e-06,
+      "loss": 0.7455,
+      "step": 6000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8179646900482853,
+      "learning_rate": 7.057170305469248e-06,
+      "loss": 0.7488,
+      "step": 6001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8382111374508192,
+      "learning_rate": 7.056225565885895e-06,
+      "loss": 0.7235,
+      "step": 6002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.387832735784366,
+      "learning_rate": 7.055280737941949e-06,
+      "loss": 0.7719,
+      "step": 6003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5084839015657914,
+      "learning_rate": 7.054335821678012e-06,
+      "loss": 0.7249,
+      "step": 6004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7047540602321143,
+      "learning_rate": 7.05339081713469e-06,
+      "loss": 0.7629,
+      "step": 6005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8593861308190083,
+      "learning_rate": 7.05244572435259e-06,
+      "loss": 0.7557,
+      "step": 6006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.700870618566003,
+      "learning_rate": 7.051500543372328e-06,
+      "loss": 0.7619,
+      "step": 6007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5484903494363829,
+      "learning_rate": 7.050555274234518e-06,
+      "loss": 0.8058,
+      "step": 6008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7087801563038492,
+      "learning_rate": 7.049609916979782e-06,
+      "loss": 0.6895,
+      "step": 6009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.919510930460796,
+      "learning_rate": 7.04866447164874e-06,
+      "loss": 0.5489,
+      "step": 6010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6701985350501054,
+      "learning_rate": 7.0477189382820266e-06,
+      "loss": 0.6548,
+      "step": 6011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.86678400971035,
+      "learning_rate": 7.046773316920269e-06,
+      "loss": 0.6561,
+      "step": 6012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7043506245893745,
+      "learning_rate": 7.045827607604103e-06,
+      "loss": 0.71,
+      "step": 6013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0518607775370479,
+      "learning_rate": 7.044881810374169e-06,
+      "loss": 0.7297,
+      "step": 6014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5061068277364094,
+      "learning_rate": 7.043935925271108e-06,
+      "loss": 0.7573,
+      "step": 6015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5953522072706148,
+      "learning_rate": 7.042989952335571e-06,
+      "loss": 0.7391,
+      "step": 6016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.647387866393971,
+      "learning_rate": 7.0420438916082024e-06,
+      "loss": 0.7952,
+      "step": 6017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6442330778978722,
+      "learning_rate": 7.041097743129662e-06,
+      "loss": 0.7037,
+      "step": 6018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6297216378838273,
+      "learning_rate": 7.040151506940605e-06,
+      "loss": 0.772,
+      "step": 6019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4569318840171714,
+      "learning_rate": 7.039205183081694e-06,
+      "loss": 0.7479,
+      "step": 6020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6581083038747502,
+      "learning_rate": 7.038258771593596e-06,
+      "loss": 0.6468,
+      "step": 6021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4772311593811147,
+      "learning_rate": 7.037312272516977e-06,
+      "loss": 0.6864,
+      "step": 6022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.158056700442059,
+      "learning_rate": 7.036365685892515e-06,
+      "loss": 0.764,
+      "step": 6023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6478504617248186,
+      "learning_rate": 7.035419011760882e-06,
+      "loss": 0.6964,
+      "step": 6024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2309370865284188,
+      "learning_rate": 7.0344722501627625e-06,
+      "loss": 0.6438,
+      "step": 6025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0463779123855868,
+      "learning_rate": 7.033525401138838e-06,
+      "loss": 0.7762,
+      "step": 6026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.972724594616515,
+      "learning_rate": 7.0325784647298e-06,
+      "loss": 0.7428,
+      "step": 6027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.789386021762879,
+      "learning_rate": 7.031631440976336e-06,
+      "loss": 0.8044,
+      "step": 6028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6076311577363585,
+      "learning_rate": 7.0306843299191465e-06,
+      "loss": 0.6854,
+      "step": 6029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.046199308724519,
+      "learning_rate": 7.0297371315989284e-06,
+      "loss": 0.8041,
+      "step": 6030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.869403176062257,
+      "learning_rate": 7.028789846056383e-06,
+      "loss": 0.6244,
+      "step": 6031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.792048405475697,
+      "learning_rate": 7.0278424733322225e-06,
+      "loss": 0.7988,
+      "step": 6032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8384425967364004,
+      "learning_rate": 7.026895013467153e-06,
+      "loss": 0.7906,
+      "step": 6033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4910585979393862,
+      "learning_rate": 7.0259474665018915e-06,
+      "loss": 0.7686,
+      "step": 6034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7710383519102224,
+      "learning_rate": 7.024999832477154e-06,
+      "loss": 0.7787,
+      "step": 6035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.511493612597553,
+      "learning_rate": 7.0240521114336655e-06,
+      "loss": 0.7904,
+      "step": 6036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7764283826235423,
+      "learning_rate": 7.023104303412149e-06,
+      "loss": 0.6699,
+      "step": 6037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5580203009846572,
+      "learning_rate": 7.022156408453336e-06,
+      "loss": 0.7219,
+      "step": 6038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5161423801153482,
+      "learning_rate": 7.0212084265979575e-06,
+      "loss": 0.7604,
+      "step": 6039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3167507041664948,
+      "learning_rate": 7.02026035788675e-06,
+      "loss": 0.6723,
+      "step": 6040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6554171746010957,
+      "learning_rate": 7.019312202360457e-06,
+      "loss": 0.7821,
+      "step": 6041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6403891731275186,
+      "learning_rate": 7.018363960059823e-06,
+      "loss": 0.7536,
+      "step": 6042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.686939369478169,
+      "learning_rate": 7.017415631025593e-06,
+      "loss": 0.741,
+      "step": 6043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6626621834374333,
+      "learning_rate": 7.016467215298519e-06,
+      "loss": 0.656,
+      "step": 6044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1619817907070695,
+      "learning_rate": 7.01551871291936e-06,
+      "loss": 0.8219,
+      "step": 6045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6677301343282294,
+      "learning_rate": 7.014570123928872e-06,
+      "loss": 0.7275,
+      "step": 6046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8195972973574057,
+      "learning_rate": 7.01362144836782e-06,
+      "loss": 0.872,
+      "step": 6047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5680752239334383,
+      "learning_rate": 7.012672686276969e-06,
+      "loss": 0.6892,
+      "step": 6048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9514219352010445,
+      "learning_rate": 7.011723837697091e-06,
+      "loss": 0.8621,
+      "step": 6049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7548862149397113,
+      "learning_rate": 7.010774902668958e-06,
+      "loss": 0.8399,
+      "step": 6050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5021065773622864,
+      "learning_rate": 7.009825881233351e-06,
+      "loss": 0.7906,
+      "step": 6051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8739812148202428,
+      "learning_rate": 7.008876773431048e-06,
+      "loss": 0.9075,
+      "step": 6052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5607119338858688,
+      "learning_rate": 7.007927579302837e-06,
+      "loss": 0.7548,
+      "step": 6053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9374107862766516,
+      "learning_rate": 7.0069782988895056e-06,
+      "loss": 0.6881,
+      "step": 6054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5642988990928663,
+      "learning_rate": 7.006028932231847e-06,
+      "loss": 0.7645,
+      "step": 6055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6985271010569165,
+      "learning_rate": 7.005079479370658e-06,
+      "loss": 0.7603,
+      "step": 6056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1747951780622112,
+      "learning_rate": 7.0041299403467365e-06,
+      "loss": 0.8601,
+      "step": 6057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.65440388354719,
+      "learning_rate": 7.003180315200891e-06,
+      "loss": 0.667,
+      "step": 6058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8887835271033995,
+      "learning_rate": 7.002230603973924e-06,
+      "loss": 0.6924,
+      "step": 6059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1330378347183057,
+      "learning_rate": 7.001280806706649e-06,
+      "loss": 0.6935,
+      "step": 6060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6983548042196102,
+      "learning_rate": 7.0003309234398805e-06,
+      "loss": 0.7435,
+      "step": 6061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5941885882704294,
+      "learning_rate": 6.999380954214438e-06,
+      "loss": 0.8593,
+      "step": 6062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6393264919420072,
+      "learning_rate": 6.9984308990711425e-06,
+      "loss": 0.809,
+      "step": 6063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8508013998637303,
+      "learning_rate": 6.9974807580508205e-06,
+      "loss": 0.8527,
+      "step": 6064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.680900543264549,
+      "learning_rate": 6.996530531194302e-06,
+      "loss": 0.7507,
+      "step": 6065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5191736765270216,
+      "learning_rate": 6.995580218542419e-06,
+      "loss": 0.6289,
+      "step": 6066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6253061127239463,
+      "learning_rate": 6.994629820136011e-06,
+      "loss": 0.7677,
+      "step": 6067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6278164237161405,
+      "learning_rate": 6.9936793360159175e-06,
+      "loss": 0.8595,
+      "step": 6068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7693636548002825,
+      "learning_rate": 6.992728766222982e-06,
+      "loss": 0.7424,
+      "step": 6069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.779088214170566,
+      "learning_rate": 6.991778110798053e-06,
+      "loss": 0.6542,
+      "step": 6070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9875582586061815,
+      "learning_rate": 6.990827369781983e-06,
+      "loss": 0.7751,
+      "step": 6071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.171136966271487,
+      "learning_rate": 6.9898765432156266e-06,
+      "loss": 0.7234,
+      "step": 6072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.093731653354785,
+      "learning_rate": 6.988925631139843e-06,
+      "loss": 0.8028,
+      "step": 6073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.689704949920941,
+      "learning_rate": 6.987974633595498e-06,
+      "loss": 0.778,
+      "step": 6074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8154691185885752,
+      "learning_rate": 6.987023550623453e-06,
+      "loss": 0.7805,
+      "step": 6075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6126384732954495,
+      "learning_rate": 6.9860723822645825e-06,
+      "loss": 0.7146,
+      "step": 6076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.760468026031143,
+      "learning_rate": 6.985121128559757e-06,
+      "loss": 0.8056,
+      "step": 6077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.792985104559129,
+      "learning_rate": 6.984169789549857e-06,
+      "loss": 0.8923,
+      "step": 6078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7920002835393223,
+      "learning_rate": 6.9832183652757625e-06,
+      "loss": 0.7902,
+      "step": 6079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9655688560242743,
+      "learning_rate": 6.982266855778356e-06,
+      "loss": 0.7763,
+      "step": 6080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7481681505070719,
+      "learning_rate": 6.981315261098529e-06,
+      "loss": 0.6879,
+      "step": 6081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7163245863163485,
+      "learning_rate": 6.980363581277172e-06,
+      "loss": 0.7808,
+      "step": 6082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.808579401056845,
+      "learning_rate": 6.979411816355183e-06,
+      "loss": 0.7982,
+      "step": 6083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8667299369986996,
+      "learning_rate": 6.978459966373458e-06,
+      "loss": 0.7775,
+      "step": 6084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6299369915621515,
+      "learning_rate": 6.9775080313729026e-06,
+      "loss": 0.7339,
+      "step": 6085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9412596139637242,
+      "learning_rate": 6.9765560113944235e-06,
+      "loss": 0.7029,
+      "step": 6086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5654073080216764,
+      "learning_rate": 6.9756039064789295e-06,
+      "loss": 0.7929,
+      "step": 6087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9534496801842092,
+      "learning_rate": 6.9746517166673375e-06,
+      "loss": 0.7946,
+      "step": 6088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6740797141280475,
+      "learning_rate": 6.973699442000561e-06,
+      "loss": 0.888,
+      "step": 6089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9916875459006949,
+      "learning_rate": 6.972747082519526e-06,
+      "loss": 0.8682,
+      "step": 6090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4702756238881594,
+      "learning_rate": 6.971794638265153e-06,
+      "loss": 0.7813,
+      "step": 6091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0037464330048147,
+      "learning_rate": 6.970842109278375e-06,
+      "loss": 0.6436,
+      "step": 6092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8323076577346347,
+      "learning_rate": 6.969889495600122e-06,
+      "loss": 0.8249,
+      "step": 6093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.735866493909092,
+      "learning_rate": 6.96893679727133e-06,
+      "loss": 0.8025,
+      "step": 6094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6621269030806871,
+      "learning_rate": 6.967984014332938e-06,
+      "loss": 0.8385,
+      "step": 6095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8180334718511468,
+      "learning_rate": 6.96703114682589e-06,
+      "loss": 0.7278,
+      "step": 6096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6013571737806576,
+      "learning_rate": 6.966078194791133e-06,
+      "loss": 0.7942,
+      "step": 6097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7306226044699071,
+      "learning_rate": 6.965125158269619e-06,
+      "loss": 0.7476,
+      "step": 6098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7117669345894004,
+      "learning_rate": 6.9641720373022996e-06,
+      "loss": 0.7283,
+      "step": 6099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.728178508818848,
+      "learning_rate": 6.963218831930132e-06,
+      "loss": 0.7364,
+      "step": 6100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7588792669672475,
+      "learning_rate": 6.96226554219408e-06,
+      "loss": 0.6718,
+      "step": 6101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1195972567046257,
+      "learning_rate": 6.961312168135107e-06,
+      "loss": 0.6711,
+      "step": 6102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2132777596098412,
+      "learning_rate": 6.960358709794182e-06,
+      "loss": 0.6564,
+      "step": 6103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.580746506422482,
+      "learning_rate": 6.959405167212278e-06,
+      "loss": 0.798,
+      "step": 6104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6410075404170523,
+      "learning_rate": 6.9584515404303685e-06,
+      "loss": 0.8699,
+      "step": 6105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0801719417251125,
+      "learning_rate": 6.9574978294894355e-06,
+      "loss": 0.7279,
+      "step": 6106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4907545060617018,
+      "learning_rate": 6.956544034430463e-06,
+      "loss": 0.686,
+      "step": 6107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7249468166834747,
+      "learning_rate": 6.955590155294434e-06,
+      "loss": 0.7802,
+      "step": 6108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.811423486015123,
+      "learning_rate": 6.954636192122339e-06,
+      "loss": 0.7124,
+      "step": 6109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7410779519432955,
+      "learning_rate": 6.953682144955176e-06,
+      "loss": 0.6845,
+      "step": 6110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.574494508598861,
+      "learning_rate": 6.952728013833941e-06,
+      "loss": 0.841,
+      "step": 6111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4764304022175865,
+      "learning_rate": 6.951773798799632e-06,
+      "loss": 0.7647,
+      "step": 6112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0152271565821196,
+      "learning_rate": 6.950819499893257e-06,
+      "loss": 0.8448,
+      "step": 6113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.737180431000323,
+      "learning_rate": 6.949865117155823e-06,
+      "loss": 0.7508,
+      "step": 6114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9381994834694276,
+      "learning_rate": 6.948910650628343e-06,
+      "loss": 0.7711,
+      "step": 6115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9476250171417038,
+      "learning_rate": 6.947956100351833e-06,
+      "loss": 0.7932,
+      "step": 6116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.192774283039405,
+      "learning_rate": 6.947001466367309e-06,
+      "loss": 0.6132,
+      "step": 6117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.112282083117691,
+      "learning_rate": 6.946046748715796e-06,
+      "loss": 0.7838,
+      "step": 6118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1495020131114382,
+      "learning_rate": 6.94509194743832e-06,
+      "loss": 0.7449,
+      "step": 6119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5460274707527326,
+      "learning_rate": 6.944137062575913e-06,
+      "loss": 0.6279,
+      "step": 6120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5865261247955198,
+      "learning_rate": 6.943182094169605e-06,
+      "loss": 0.6809,
+      "step": 6121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.832232762101104,
+      "learning_rate": 6.942227042260435e-06,
+      "loss": 0.6827,
+      "step": 6122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.082646619510683,
+      "learning_rate": 6.941271906889443e-06,
+      "loss": 0.6271,
+      "step": 6123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9578266287584474,
+      "learning_rate": 6.940316688097675e-06,
+      "loss": 0.8826,
+      "step": 6124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6635130716925923,
+      "learning_rate": 6.9393613859261755e-06,
+      "loss": 0.854,
+      "step": 6125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1147745511330218,
+      "learning_rate": 6.938406000416e-06,
+      "loss": 0.5462,
+      "step": 6126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.33474920963922,
+      "learning_rate": 6.9374505316082e-06,
+      "loss": 0.6214,
+      "step": 6127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7196845132846392,
+      "learning_rate": 6.936494979543836e-06,
+      "loss": 0.628,
+      "step": 6128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8069878789022282,
+      "learning_rate": 6.935539344263971e-06,
+      "loss": 0.7532,
+      "step": 6129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.481711359621969,
+      "learning_rate": 6.93458362580967e-06,
+      "loss": 0.741,
+      "step": 6130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.911195531549821,
+      "learning_rate": 6.9336278242220015e-06,
+      "loss": 0.8208,
+      "step": 6131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.158183058236986,
+      "learning_rate": 6.932671939542037e-06,
+      "loss": 0.6664,
+      "step": 6132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7414096459029322,
+      "learning_rate": 6.9317159718108575e-06,
+      "loss": 0.7682,
+      "step": 6133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6718729767748857,
+      "learning_rate": 6.93075992106954e-06,
+      "loss": 0.8136,
+      "step": 6134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6345958193198789,
+      "learning_rate": 6.929803787359169e-06,
+      "loss": 0.6996,
+      "step": 6135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.644435485055116,
+      "learning_rate": 6.928847570720829e-06,
+      "loss": 0.7962,
+      "step": 6136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6021995161706055,
+      "learning_rate": 6.927891271195616e-06,
+      "loss": 0.7569,
+      "step": 6137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1573185623364466,
+      "learning_rate": 6.926934888824619e-06,
+      "loss": 0.6635,
+      "step": 6138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.575976753868241,
+      "learning_rate": 6.925978423648941e-06,
+      "loss": 0.7639,
+      "step": 6139
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9068849692492063,
+      "learning_rate": 6.92502187570968e-06,
+      "loss": 0.7725,
+      "step": 6140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5304645249308746,
+      "learning_rate": 6.92406524504794e-06,
+      "loss": 0.7275,
+      "step": 6141
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8739463242988554,
+      "learning_rate": 6.923108531704832e-06,
+      "loss": 0.6024,
+      "step": 6142
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5745770704885849,
+      "learning_rate": 6.9221517357214674e-06,
+      "loss": 0.7558,
+      "step": 6143
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4856959231649007,
+      "learning_rate": 6.921194857138963e-06,
+      "loss": 0.7704,
+      "step": 6144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1368716278323996,
+      "learning_rate": 6.920237895998436e-06,
+      "loss": 0.7771,
+      "step": 6145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5427449900408607,
+      "learning_rate": 6.919280852341011e-06,
+      "loss": 0.7041,
+      "step": 6146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6440246808086627,
+      "learning_rate": 6.91832372620781e-06,
+      "loss": 0.6846,
+      "step": 6147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8632603681881077,
+      "learning_rate": 6.917366517639969e-06,
+      "loss": 0.7718,
+      "step": 6148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7373474206372135,
+      "learning_rate": 6.91640922667862e-06,
+      "loss": 0.8478,
+      "step": 6149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.677229787166522,
+      "learning_rate": 6.9154518533648966e-06,
+      "loss": 0.6941,
+      "step": 6150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5340069868555555,
+      "learning_rate": 6.91449439773994e-06,
+      "loss": 0.7675,
+      "step": 6151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8717817265800274,
+      "learning_rate": 6.913536859844896e-06,
+      "loss": 0.8141,
+      "step": 6152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6778846230687317,
+      "learning_rate": 6.912579239720913e-06,
+      "loss": 0.7394,
+      "step": 6153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.709157759293513,
+      "learning_rate": 6.911621537409139e-06,
+      "loss": 0.9178,
+      "step": 6154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.79305566118116,
+      "learning_rate": 6.9106637529507305e-06,
+      "loss": 0.7686,
+      "step": 6155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5822163355990668,
+      "learning_rate": 6.909705886386844e-06,
+      "loss": 0.8761,
+      "step": 6156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.858068445002292,
+      "learning_rate": 6.908747937758644e-06,
+      "loss": 0.6918,
+      "step": 6157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8656066064904775,
+      "learning_rate": 6.9077899071072954e-06,
+      "loss": 0.7579,
+      "step": 6158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7217688929923811,
+      "learning_rate": 6.906831794473963e-06,
+      "loss": 0.8201,
+      "step": 6159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.757063823343889,
+      "learning_rate": 6.9058735998998224e-06,
+      "loss": 0.6782,
+      "step": 6160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4997828241552227,
+      "learning_rate": 6.9049153234260495e-06,
+      "loss": 0.7712,
+      "step": 6161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.657671387879305,
+      "learning_rate": 6.903956965093821e-06,
+      "loss": 0.7671,
+      "step": 6162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.441779619200941,
+      "learning_rate": 6.902998524944322e-06,
+      "loss": 0.6839,
+      "step": 6163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4587822408708155,
+      "learning_rate": 6.9020400030187394e-06,
+      "loss": 0.6739,
+      "step": 6164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4474541209586185,
+      "learning_rate": 6.90108139935826e-06,
+      "loss": 0.6975,
+      "step": 6165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7245010986923066,
+      "learning_rate": 6.90012271400408e-06,
+      "loss": 0.6917,
+      "step": 6166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6334054180027817,
+      "learning_rate": 6.899163946997396e-06,
+      "loss": 0.6873,
+      "step": 6167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8218494266942964,
+      "learning_rate": 6.898205098379407e-06,
+      "loss": 0.762,
+      "step": 6168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.066184269071502,
+      "learning_rate": 6.897246168191317e-06,
+      "loss": 0.749,
+      "step": 6169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6929573448827049,
+      "learning_rate": 6.8962871564743325e-06,
+      "loss": 0.6985,
+      "step": 6170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6491546732026425,
+      "learning_rate": 6.895328063269667e-06,
+      "loss": 0.7143,
+      "step": 6171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5912374346901705,
+      "learning_rate": 6.894368888618534e-06,
+      "loss": 0.7228,
+      "step": 6172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8947315541778564,
+      "learning_rate": 6.8934096325621505e-06,
+      "loss": 0.7081,
+      "step": 6173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0518338688015252,
+      "learning_rate": 6.892450295141737e-06,
+      "loss": 0.6836,
+      "step": 6174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3071047035535655,
+      "learning_rate": 6.891490876398521e-06,
+      "loss": 0.692,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6062379117809151,
+      "learning_rate": 6.890531376373729e-06,
+      "loss": 0.7227,
+      "step": 6176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4799746481609886,
+      "learning_rate": 6.889571795108594e-06,
+      "loss": 0.6727,
+      "step": 6177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8170713904300222,
+      "learning_rate": 6.88861213264435e-06,
+      "loss": 0.8398,
+      "step": 6178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.588713301249202,
+      "learning_rate": 6.887652389022236e-06,
+      "loss": 0.8016,
+      "step": 6179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9540324123656174,
+      "learning_rate": 6.886692564283496e-06,
+      "loss": 0.7492,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8651102263951922,
+      "learning_rate": 6.885732658469374e-06,
+      "loss": 0.7339,
+      "step": 6181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.746488847078647,
+      "learning_rate": 6.88477267162112e-06,
+      "loss": 0.6882,
+      "step": 6182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1349619859532403,
+      "learning_rate": 6.883812603779985e-06,
+      "loss": 0.6856,
+      "step": 6183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6471558251642322,
+      "learning_rate": 6.88285245498723e-06,
+      "loss": 0.848,
+      "step": 6184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.259763942123785,
+      "learning_rate": 6.88189222528411e-06,
+      "loss": 0.7686,
+      "step": 6185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8095108196273917,
+      "learning_rate": 6.8809319147118905e-06,
+      "loss": 0.7895,
+      "step": 6186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7974236371755388,
+      "learning_rate": 6.879971523311837e-06,
+      "loss": 0.7301,
+      "step": 6187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4002545087411458,
+      "learning_rate": 6.8790110511252195e-06,
+      "loss": 0.723,
+      "step": 6188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.884471200774913,
+      "learning_rate": 6.878050498193314e-06,
+      "loss": 0.7197,
+      "step": 6189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.509554255222643,
+      "learning_rate": 6.877089864557395e-06,
+      "loss": 0.83,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6371669557490116,
+      "learning_rate": 6.876129150258744e-06,
+      "loss": 0.7899,
+      "step": 6191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7412566353633898,
+      "learning_rate": 6.875168355338645e-06,
+      "loss": 0.6973,
+      "step": 6192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7777165805869224,
+      "learning_rate": 6.874207479838386e-06,
+      "loss": 0.5882,
+      "step": 6193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2983526990663885,
+      "learning_rate": 6.873246523799256e-06,
+      "loss": 0.6489,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4185641187533258,
+      "learning_rate": 6.872285487262555e-06,
+      "loss": 0.7608,
+      "step": 6195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5091723934782244,
+      "learning_rate": 6.871324370269573e-06,
+      "loss": 0.6476,
+      "step": 6196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0653974291783594,
+      "learning_rate": 6.870363172861616e-06,
+      "loss": 0.6857,
+      "step": 6197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7914904050239824,
+      "learning_rate": 6.869401895079988e-06,
+      "loss": 0.7716,
+      "step": 6198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.871919022163109,
+      "learning_rate": 6.868440536965997e-06,
+      "loss": 0.696,
+      "step": 6199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6522830150793346,
+      "learning_rate": 6.867479098560956e-06,
+      "loss": 0.793,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2335160261720495,
+      "learning_rate": 6.866517579906178e-06,
+      "loss": 0.6797,
+      "step": 6201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5804389309290074,
+      "learning_rate": 6.865555981042983e-06,
+      "loss": 0.8102,
+      "step": 6202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9567598585953635,
+      "learning_rate": 6.864594302012691e-06,
+      "loss": 0.6307,
+      "step": 6203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9626416841652283,
+      "learning_rate": 6.863632542856632e-06,
+      "loss": 0.7616,
+      "step": 6204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.333980745980387,
+      "learning_rate": 6.8626707036161314e-06,
+      "loss": 0.5792,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7215886905372486,
+      "learning_rate": 6.86170878433252e-06,
+      "loss": 0.7395,
+      "step": 6206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3059560735055786,
+      "learning_rate": 6.860746785047139e-06,
+      "loss": 0.7836,
+      "step": 6207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.857205274427664,
+      "learning_rate": 6.859784705801323e-06,
+      "loss": 0.8089,
+      "step": 6208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3639781982811594,
+      "learning_rate": 6.858822546636417e-06,
+      "loss": 0.7337,
+      "step": 6209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7533149573814975,
+      "learning_rate": 6.857860307593764e-06,
+      "loss": 0.6374,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6317550072990676,
+      "learning_rate": 6.856897988714718e-06,
+      "loss": 0.6797,
+      "step": 6211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.563988839604194,
+      "learning_rate": 6.855935590040629e-06,
+      "loss": 0.755,
+      "step": 6212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6685197403124037,
+      "learning_rate": 6.854973111612855e-06,
+      "loss": 0.7465,
+      "step": 6213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9291768212488858,
+      "learning_rate": 6.854010553472757e-06,
+      "loss": 0.7431,
+      "step": 6214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8651043090733022,
+      "learning_rate": 6.8530479156616925e-06,
+      "loss": 0.8224,
+      "step": 6215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.692047512186583,
+      "learning_rate": 6.852085198221035e-06,
+      "loss": 0.6797,
+      "step": 6216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.564702419220275,
+      "learning_rate": 6.851122401192149e-06,
+      "loss": 0.6366,
+      "step": 6217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.642672258021254,
+      "learning_rate": 6.850159524616413e-06,
+      "loss": 0.651,
+      "step": 6218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.525754290516905,
+      "learning_rate": 6.849196568535201e-06,
+      "loss": 0.7072,
+      "step": 6219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7830529549522638,
+      "learning_rate": 6.848233532989894e-06,
+      "loss": 0.738,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6453630429254569,
+      "learning_rate": 6.847270418021876e-06,
+      "loss": 0.6289,
+      "step": 6221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8606101198650202,
+      "learning_rate": 6.8463072236725345e-06,
+      "loss": 0.6732,
+      "step": 6222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7884075690556744,
+      "learning_rate": 6.845343949983258e-06,
+      "loss": 0.7002,
+      "step": 6223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1565866923801023,
+      "learning_rate": 6.8443805969954445e-06,
+      "loss": 0.6552,
+      "step": 6224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6074570297977473,
+      "learning_rate": 6.8434171647504885e-06,
+      "loss": 0.7672,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0985349211126967,
+      "learning_rate": 6.84245365328979e-06,
+      "loss": 0.7836,
+      "step": 6226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4795865348496977,
+      "learning_rate": 6.841490062654757e-06,
+      "loss": 0.8458,
+      "step": 6227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.218066968730941,
+      "learning_rate": 6.840526392886793e-06,
+      "loss": 0.7339,
+      "step": 6228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5297743475999892,
+      "learning_rate": 6.839562644027311e-06,
+      "loss": 0.7356,
+      "step": 6229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.686757599681423,
+      "learning_rate": 6.838598816117725e-06,
+      "loss": 0.8485,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7628294558524442,
+      "learning_rate": 6.837634909199453e-06,
+      "loss": 0.7502,
+      "step": 6231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.874521793404496,
+      "learning_rate": 6.836670923313917e-06,
+      "loss": 0.7243,
+      "step": 6232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6009373685143635,
+      "learning_rate": 6.835706858502542e-06,
+      "loss": 0.6958,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.111799007729763,
+      "learning_rate": 6.834742714806754e-06,
+      "loss": 0.6691,
+      "step": 6234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.811702973053789,
+      "learning_rate": 6.833778492267984e-06,
+      "loss": 0.7218,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1906758362198533,
+      "learning_rate": 6.832814190927669e-06,
+      "loss": 0.631,
+      "step": 6236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.815665810837433,
+      "learning_rate": 6.831849810827247e-06,
+      "loss": 0.7414,
+      "step": 6237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.780832878091985,
+      "learning_rate": 6.830885352008159e-06,
+      "loss": 0.7383,
+      "step": 6238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5916577982135096,
+      "learning_rate": 6.8299208145118475e-06,
+      "loss": 0.7463,
+      "step": 6239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2523783855673396,
+      "learning_rate": 6.828956198379766e-06,
+      "loss": 0.7722,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.775844159071708,
+      "learning_rate": 6.827991503653363e-06,
+      "loss": 0.7763,
+      "step": 6241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5861358852617267,
+      "learning_rate": 6.827026730374093e-06,
+      "loss": 0.6865,
+      "step": 6242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.983791634858843,
+      "learning_rate": 6.826061878583416e-06,
+      "loss": 0.8009,
+      "step": 6243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8902327446288645,
+      "learning_rate": 6.825096948322791e-06,
+      "loss": 0.7974,
+      "step": 6244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7025337556246376,
+      "learning_rate": 6.824131939633688e-06,
+      "loss": 0.6928,
+      "step": 6245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6181288096363518,
+      "learning_rate": 6.823166852557572e-06,
+      "loss": 0.7732,
+      "step": 6246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5441156414915045,
+      "learning_rate": 6.822201687135916e-06,
+      "loss": 0.6992,
+      "step": 6247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5459267682915991,
+      "learning_rate": 6.821236443410195e-06,
+      "loss": 0.753,
+      "step": 6248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7673899214149513,
+      "learning_rate": 6.820271121421889e-06,
+      "loss": 0.7696,
+      "step": 6249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9914119462069422,
+      "learning_rate": 6.819305721212479e-06,
+      "loss": 0.7738,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.525323434096852,
+      "learning_rate": 6.818340242823449e-06,
+      "loss": 0.6961,
+      "step": 6251
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.108950677487899,
+      "learning_rate": 6.8173746862962904e-06,
+      "loss": 0.7698,
+      "step": 6252
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.745347129433455,
+      "learning_rate": 6.816409051672492e-06,
+      "loss": 0.8609,
+      "step": 6253
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.73897658337265,
+      "learning_rate": 6.815443338993554e-06,
+      "loss": 0.6794,
+      "step": 6254
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8837822360109657,
+      "learning_rate": 6.81447754830097e-06,
+      "loss": 0.7768,
+      "step": 6255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5378804013482215,
+      "learning_rate": 6.813511679636247e-06,
+      "loss": 0.6505,
+      "step": 6256
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.587781448436845,
+      "learning_rate": 6.8125457330408865e-06,
+      "loss": 0.753,
+      "step": 6257
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8604671003966606,
+      "learning_rate": 6.8115797085564e-06,
+      "loss": 0.7873,
+      "step": 6258
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6479246450087797,
+      "learning_rate": 6.810613606224299e-06,
+      "loss": 0.6741,
+      "step": 6259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9283655758711826,
+      "learning_rate": 6.809647426086098e-06,
+      "loss": 0.8723,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9655011177089372,
+      "learning_rate": 6.808681168183317e-06,
+      "loss": 0.8171,
+      "step": 6261
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.704553148274822,
+      "learning_rate": 6.807714832557479e-06,
+      "loss": 0.7761,
+      "step": 6262
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0756072207210294,
+      "learning_rate": 6.806748419250107e-06,
+      "loss": 0.6395,
+      "step": 6263
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7251017490966662,
+      "learning_rate": 6.805781928302732e-06,
+      "loss": 0.7126,
+      "step": 6264
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.07608878500952,
+      "learning_rate": 6.804815359756887e-06,
+      "loss": 0.7692,
+      "step": 6265
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7215236926234316,
+      "learning_rate": 6.803848713654105e-06,
+      "loss": 0.7944,
+      "step": 6266
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6429075790433054,
+      "learning_rate": 6.8028819900359276e-06,
+      "loss": 0.7744,
+      "step": 6267
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8356938920570027,
+      "learning_rate": 6.801915188943895e-06,
+      "loss": 0.7484,
+      "step": 6268
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6082983754166191,
+      "learning_rate": 6.800948310419554e-06,
+      "loss": 0.8244,
+      "step": 6269
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1897476091827717,
+      "learning_rate": 6.799981354504455e-06,
+      "loss": 0.7058,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6707607749849993,
+      "learning_rate": 6.799014321240145e-06,
+      "loss": 0.7525,
+      "step": 6271
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9091039845376587,
+      "learning_rate": 6.798047210668185e-06,
+      "loss": 0.856,
+      "step": 6272
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6350941183180943,
+      "learning_rate": 6.797080022830131e-06,
+      "loss": 0.5842,
+      "step": 6273
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.392380486283044,
+      "learning_rate": 6.796112757767547e-06,
+      "loss": 0.7174,
+      "step": 6274
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.578564868356414,
+      "learning_rate": 6.795145415521997e-06,
+      "loss": 0.7845,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2156283682363032,
+      "learning_rate": 6.79417799613505e-06,
+      "loss": 0.6369,
+      "step": 6276
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8770901692240531,
+      "learning_rate": 6.793210499648281e-06,
+      "loss": 0.7328,
+      "step": 6277
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9594875525680397,
+      "learning_rate": 6.792242926103261e-06,
+      "loss": 0.6849,
+      "step": 6278
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6590318793793364,
+      "learning_rate": 6.7912752755415716e-06,
+      "loss": 0.6971,
+      "step": 6279
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.062788045843333,
+      "learning_rate": 6.790307548004794e-06,
+      "loss": 0.7208,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5023709946354822,
+      "learning_rate": 6.7893397435345144e-06,
+      "loss": 0.7743,
+      "step": 6281
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6650070191271316,
+      "learning_rate": 6.788371862172319e-06,
+      "loss": 0.7814,
+      "step": 6282
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.859055099451998,
+      "learning_rate": 6.787403903959804e-06,
+      "loss": 0.8321,
+      "step": 6283
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6777487656178154,
+      "learning_rate": 6.786435868938561e-06,
+      "loss": 0.82,
+      "step": 6284
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6040343863626276,
+      "learning_rate": 6.785467757150193e-06,
+      "loss": 0.6895,
+      "step": 6285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7031216622772587,
+      "learning_rate": 6.7844995686362955e-06,
+      "loss": 0.6993,
+      "step": 6286
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1938127081893084,
+      "learning_rate": 6.7835313034384786e-06,
+      "loss": 0.7307,
+      "step": 6287
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8548399974293932,
+      "learning_rate": 6.782562961598349e-06,
+      "loss": 0.8071,
+      "step": 6288
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4603753479385055,
+      "learning_rate": 6.78159454315752e-06,
+      "loss": 0.724,
+      "step": 6289
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8158573349711153,
+      "learning_rate": 6.780626048157606e-06,
+      "loss": 0.7351,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1546720076323096,
+      "learning_rate": 6.779657476640222e-06,
+      "loss": 0.7037,
+      "step": 6291
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7288267047400834,
+      "learning_rate": 6.778688828646996e-06,
+      "loss": 0.9569,
+      "step": 6292
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6461210740124765,
+      "learning_rate": 6.777720104219548e-06,
+      "loss": 0.6503,
+      "step": 6293
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9331974405106525,
+      "learning_rate": 6.776751303399509e-06,
+      "loss": 0.7901,
+      "step": 6294
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0562966307286925,
+      "learning_rate": 6.775782426228509e-06,
+      "loss": 0.8115,
+      "step": 6295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.505609416468651,
+      "learning_rate": 6.774813472748183e-06,
+      "loss": 0.7564,
+      "step": 6296
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8113857195823944,
+      "learning_rate": 6.773844443000169e-06,
+      "loss": 0.7251,
+      "step": 6297
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.642606002512174,
+      "learning_rate": 6.77287533702611e-06,
+      "loss": 0.7537,
+      "step": 6298
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1708077545098825,
+      "learning_rate": 6.771906154867649e-06,
+      "loss": 0.6848,
+      "step": 6299
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4346896190707097,
+      "learning_rate": 6.770936896566434e-06,
+      "loss": 0.6524,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8304872814357303,
+      "learning_rate": 6.769967562164116e-06,
+      "loss": 0.7452,
+      "step": 6301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6282991866395609,
+      "learning_rate": 6.768998151702351e-06,
+      "loss": 0.7233,
+      "step": 6302
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.693774252977571,
+      "learning_rate": 6.768028665222795e-06,
+      "loss": 0.6988,
+      "step": 6303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0441488731510207,
+      "learning_rate": 6.767059102767109e-06,
+      "loss": 0.5977,
+      "step": 6304
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.649321287837944,
+      "learning_rate": 6.766089464376958e-06,
+      "loss": 0.7439,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2637931864714713,
+      "learning_rate": 6.76511975009401e-06,
+      "loss": 0.665,
+      "step": 6306
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5016798106592526,
+      "learning_rate": 6.7641499599599355e-06,
+      "loss": 0.7155,
+      "step": 6307
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9563006075573148,
+      "learning_rate": 6.763180094016409e-06,
+      "loss": 0.7347,
+      "step": 6308
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.591991272564632,
+      "learning_rate": 6.7622101523051045e-06,
+      "loss": 0.7777,
+      "step": 6309
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3895130556537523,
+      "learning_rate": 6.761240134867706e-06,
+      "loss": 0.6576,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7974011565039834,
+      "learning_rate": 6.760270041745898e-06,
+      "loss": 0.6911,
+      "step": 6311
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1415363768029858,
+      "learning_rate": 6.759299872981364e-06,
+      "loss": 0.5859,
+      "step": 6312
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.498141954737428,
+      "learning_rate": 6.758329628615796e-06,
+      "loss": 0.6893,
+      "step": 6313
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3503205174546125,
+      "learning_rate": 6.757359308690889e-06,
+      "loss": 0.647,
+      "step": 6314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6004535855386957,
+      "learning_rate": 6.756388913248339e-06,
+      "loss": 0.7433,
+      "step": 6315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7052833568679715,
+      "learning_rate": 6.755418442329846e-06,
+      "loss": 0.7312,
+      "step": 6316
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7684589653525966,
+      "learning_rate": 6.754447895977113e-06,
+      "loss": 0.8159,
+      "step": 6317
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6467130545788153,
+      "learning_rate": 6.753477274231846e-06,
+      "loss": 0.7186,
+      "step": 6318
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.690629310565174,
+      "learning_rate": 6.7525065771357546e-06,
+      "loss": 0.6001,
+      "step": 6319
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0828371204489495,
+      "learning_rate": 6.751535804730553e-06,
+      "loss": 0.9159,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.205007889040121,
+      "learning_rate": 6.750564957057958e-06,
+      "loss": 0.6685,
+      "step": 6321
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0854595852780276,
+      "learning_rate": 6.749594034159687e-06,
+      "loss": 0.8209,
+      "step": 6322
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.579645929686028,
+      "learning_rate": 6.748623036077464e-06,
+      "loss": 0.7571,
+      "step": 6323
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.624020277507788,
+      "learning_rate": 6.7476519628530145e-06,
+      "loss": 0.8811,
+      "step": 6324
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.570782468414025,
+      "learning_rate": 6.7466808145280685e-06,
+      "loss": 0.7785,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3439130370448593,
+      "learning_rate": 6.745709591144358e-06,
+      "loss": 0.7362,
+      "step": 6326
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8983105853288655,
+      "learning_rate": 6.744738292743617e-06,
+      "loss": 0.6791,
+      "step": 6327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0499119558486654,
+      "learning_rate": 6.743766919367588e-06,
+      "loss": 0.7602,
+      "step": 6328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.755532473928082,
+      "learning_rate": 6.742795471058009e-06,
+      "loss": 0.7208,
+      "step": 6329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4439885790957359,
+      "learning_rate": 6.741823947856627e-06,
+      "loss": 0.6959,
+      "step": 6330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7225991850348101,
+      "learning_rate": 6.740852349805193e-06,
+      "loss": 0.6604,
+      "step": 6331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8776026669520072,
+      "learning_rate": 6.739880676945455e-06,
+      "loss": 0.8139,
+      "step": 6332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5395637003887075,
+      "learning_rate": 6.73890892931917e-06,
+      "loss": 0.7443,
+      "step": 6333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9631156225725042,
+      "learning_rate": 6.737937106968094e-06,
+      "loss": 0.9541,
+      "step": 6334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9668586925907572,
+      "learning_rate": 6.736965209933992e-06,
+      "loss": 0.8153,
+      "step": 6335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.041033857687528,
+      "learning_rate": 6.735993238258624e-06,
+      "loss": 0.8473,
+      "step": 6336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7885172633873165,
+      "learning_rate": 6.735021191983763e-06,
+      "loss": 0.8008,
+      "step": 6337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6552713828673773,
+      "learning_rate": 6.734049071151175e-06,
+      "loss": 0.7669,
+      "step": 6338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7199340293314682,
+      "learning_rate": 6.7330768758026374e-06,
+      "loss": 0.7724,
+      "step": 6339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5982165182467258,
+      "learning_rate": 6.732104605979928e-06,
+      "loss": 0.6311,
+      "step": 6340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9819746836677916,
+      "learning_rate": 6.7311322617248246e-06,
+      "loss": 0.7888,
+      "step": 6341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.748295045995632,
+      "learning_rate": 6.730159843079113e-06,
+      "loss": 0.7646,
+      "step": 6342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.961781022299082,
+      "learning_rate": 6.729187350084579e-06,
+      "loss": 0.791,
+      "step": 6343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3704878068619248,
+      "learning_rate": 6.728214782783013e-06,
+      "loss": 0.7176,
+      "step": 6344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5904393045339842,
+      "learning_rate": 6.72724214121621e-06,
+      "loss": 0.6925,
+      "step": 6345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.642689573608109,
+      "learning_rate": 6.726269425425967e-06,
+      "loss": 0.8312,
+      "step": 6346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5699516678271577,
+      "learning_rate": 6.725296635454079e-06,
+      "loss": 0.6657,
+      "step": 6347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5670462221308168,
+      "learning_rate": 6.724323771342354e-06,
+      "loss": 0.7705,
+      "step": 6348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.763456561535477,
+      "learning_rate": 6.723350833132596e-06,
+      "loss": 0.8162,
+      "step": 6349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7709125109646924,
+      "learning_rate": 6.722377820866614e-06,
+      "loss": 0.6651,
+      "step": 6350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.02972680060789,
+      "learning_rate": 6.721404734586221e-06,
+      "loss": 0.8191,
+      "step": 6351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4815278508115801,
+      "learning_rate": 6.7204315743332325e-06,
+      "loss": 0.8339,
+      "step": 6352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4718323432187628,
+      "learning_rate": 6.71945834014947e-06,
+      "loss": 0.7546,
+      "step": 6353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7039950489877154,
+      "learning_rate": 6.7184850320767505e-06,
+      "loss": 0.6283,
+      "step": 6354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.014375486511842,
+      "learning_rate": 6.717511650156903e-06,
+      "loss": 0.5999,
+      "step": 6355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4928800199862977,
+      "learning_rate": 6.716538194431754e-06,
+      "loss": 0.752,
+      "step": 6356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5344777053510517,
+      "learning_rate": 6.715564664943138e-06,
+      "loss": 0.6953,
+      "step": 6357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6321061550790577,
+      "learning_rate": 6.714591061732886e-06,
+      "loss": 0.8388,
+      "step": 6358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6173767983793998,
+      "learning_rate": 6.7136173848428375e-06,
+      "loss": 0.7918,
+      "step": 6359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.579257756631194,
+      "learning_rate": 6.712643634314834e-06,
+      "loss": 0.6191,
+      "step": 6360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8830036212527146,
+      "learning_rate": 6.711669810190717e-06,
+      "loss": 0.9265,
+      "step": 6361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7568645672145033,
+      "learning_rate": 6.710695912512339e-06,
+      "loss": 0.7626,
+      "step": 6362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.882429820971791,
+      "learning_rate": 6.7097219413215474e-06,
+      "loss": 0.7392,
+      "step": 6363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.002480091867793,
+      "learning_rate": 6.708747896660196e-06,
+      "loss": 0.6845,
+      "step": 6364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7900409819382763,
+      "learning_rate": 6.70777377857014e-06,
+      "loss": 0.7885,
+      "step": 6365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.478506721338207,
+      "learning_rate": 6.706799587093243e-06,
+      "loss": 0.892,
+      "step": 6366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.612913560408393,
+      "learning_rate": 6.705825322271367e-06,
+      "loss": 0.6705,
+      "step": 6367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.135327638398179,
+      "learning_rate": 6.7048509841463775e-06,
+      "loss": 0.6058,
+      "step": 6368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5519325595577134,
+      "learning_rate": 6.703876572760144e-06,
+      "loss": 0.7244,
+      "step": 6369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6955939610344306,
+      "learning_rate": 6.702902088154539e-06,
+      "loss": 0.8657,
+      "step": 6370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.541778000298396,
+      "learning_rate": 6.701927530371439e-06,
+      "loss": 0.625,
+      "step": 6371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.727650847867235,
+      "learning_rate": 6.700952899452723e-06,
+      "loss": 0.7123,
+      "step": 6372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7164408916175997,
+      "learning_rate": 6.699978195440273e-06,
+      "loss": 0.7619,
+      "step": 6373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8048418356043419,
+      "learning_rate": 6.6990034183759726e-06,
+      "loss": 0.6822,
+      "step": 6374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.783880341053786,
+      "learning_rate": 6.698028568301712e-06,
+      "loss": 0.7876,
+      "step": 6375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.128523475081713,
+      "learning_rate": 6.697053645259382e-06,
+      "loss": 0.7392,
+      "step": 6376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3925137261784446,
+      "learning_rate": 6.696078649290878e-06,
+      "loss": 0.7344,
+      "step": 6377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.566368051240005,
+      "learning_rate": 6.695103580438097e-06,
+      "loss": 0.7432,
+      "step": 6378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.71367514694518,
+      "learning_rate": 6.694128438742939e-06,
+      "loss": 0.7106,
+      "step": 6379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0728441706037057,
+      "learning_rate": 6.693153224247309e-06,
+      "loss": 0.6982,
+      "step": 6380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5590871617374624,
+      "learning_rate": 6.692177936993114e-06,
+      "loss": 0.828,
+      "step": 6381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6265048570989866,
+      "learning_rate": 6.691202577022267e-06,
+      "loss": 0.7235,
+      "step": 6382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7721086057485633,
+      "learning_rate": 6.690227144376676e-06,
+      "loss": 0.956,
+      "step": 6383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4205557287210775,
+      "learning_rate": 6.689251639098261e-06,
+      "loss": 0.745,
+      "step": 6384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.80465731124628,
+      "learning_rate": 6.68827606122894e-06,
+      "loss": 0.9852,
+      "step": 6385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6210615376455317,
+      "learning_rate": 6.687300410810638e-06,
+      "loss": 0.7632,
+      "step": 6386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.333556237953274,
+      "learning_rate": 6.686324687885281e-06,
+      "loss": 0.7962,
+      "step": 6387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.519641428682599,
+      "learning_rate": 6.685348892494793e-06,
+      "loss": 0.7008,
+      "step": 6388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8194922526676531,
+      "learning_rate": 6.684373024681112e-06,
+      "loss": 0.7604,
+      "step": 6389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.674647377517605,
+      "learning_rate": 6.683397084486169e-06,
+      "loss": 0.7192,
+      "step": 6390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5602316962568306,
+      "learning_rate": 6.682421071951907e-06,
+      "loss": 0.6834,
+      "step": 6391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0789041305812328,
+      "learning_rate": 6.681444987120262e-06,
+      "loss": 0.6655,
+      "step": 6392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6812377875018671,
+      "learning_rate": 6.680468830033182e-06,
+      "loss": 0.8692,
+      "step": 6393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6418251690156602,
+      "learning_rate": 6.679492600732614e-06,
+      "loss": 0.8418,
+      "step": 6394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8001002506730504,
+      "learning_rate": 6.678516299260507e-06,
+      "loss": 0.7681,
+      "step": 6395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1853915386521403,
+      "learning_rate": 6.677539925658819e-06,
+      "loss": 0.7634,
+      "step": 6396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.518750259879728,
+      "learning_rate": 6.676563479969502e-06,
+      "loss": 0.7411,
+      "step": 6397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7437794585060928,
+      "learning_rate": 6.67558696223452e-06,
+      "loss": 0.7661,
+      "step": 6398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6115649615527667,
+      "learning_rate": 6.674610372495832e-06,
+      "loss": 0.7427,
+      "step": 6399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6641105009275123,
+      "learning_rate": 6.673633710795409e-06,
+      "loss": 0.8165,
+      "step": 6400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7633177054712215,
+      "learning_rate": 6.672656977175218e-06,
+      "loss": 0.735,
+      "step": 6401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6989840331609651,
+      "learning_rate": 6.671680171677232e-06,
+      "loss": 0.6474,
+      "step": 6402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8238501630016117,
+      "learning_rate": 6.670703294343423e-06,
+      "loss": 0.7513,
+      "step": 6403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.03430186161196,
+      "learning_rate": 6.669726345215776e-06,
+      "loss": 0.7773,
+      "step": 6404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5980888265364184,
+      "learning_rate": 6.668749324336268e-06,
+      "loss": 0.7214,
+      "step": 6405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.545889725597307,
+      "learning_rate": 6.667772231746885e-06,
+      "loss": 0.81,
+      "step": 6406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0432800048828454,
+      "learning_rate": 6.666795067489615e-06,
+      "loss": 0.8398,
+      "step": 6407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8426753491753876,
+      "learning_rate": 6.665817831606447e-06,
+      "loss": 0.8044,
+      "step": 6408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6282540331998283,
+      "learning_rate": 6.66484052413938e-06,
+      "loss": 0.7359,
+      "step": 6409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8669952278409307,
+      "learning_rate": 6.663863145130406e-06,
+      "loss": 0.7834,
+      "step": 6410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5690189964374035,
+      "learning_rate": 6.662885694621527e-06,
+      "loss": 0.6737,
+      "step": 6411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8580570447609668,
+      "learning_rate": 6.661908172654746e-06,
+      "loss": 0.7037,
+      "step": 6412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7062358215702262,
+      "learning_rate": 6.6609305792720705e-06,
+      "loss": 0.8278,
+      "step": 6413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7264437582682608,
+      "learning_rate": 6.659952914515508e-06,
+      "loss": 0.6624,
+      "step": 6414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.18052153640933,
+      "learning_rate": 6.658975178427073e-06,
+      "loss": 0.6853,
+      "step": 6415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4331764708437718,
+      "learning_rate": 6.65799737104878e-06,
+      "loss": 0.715,
+      "step": 6416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0434301213085484,
+      "learning_rate": 6.657019492422645e-06,
+      "loss": 0.9087,
+      "step": 6417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8737598666103874,
+      "learning_rate": 6.656041542590694e-06,
+      "loss": 0.6461,
+      "step": 6418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4927392159899775,
+      "learning_rate": 6.65506352159495e-06,
+      "loss": 0.645,
+      "step": 6419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1370616628439683,
+      "learning_rate": 6.65408542947744e-06,
+      "loss": 0.6474,
+      "step": 6420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9418752590627462,
+      "learning_rate": 6.653107266280194e-06,
+      "loss": 0.764,
+      "step": 6421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5596479475286122,
+      "learning_rate": 6.652129032045249e-06,
+      "loss": 0.72,
+      "step": 6422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8256040701424356,
+      "learning_rate": 6.651150726814639e-06,
+      "loss": 0.7228,
+      "step": 6423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7359401723409276,
+      "learning_rate": 6.650172350630406e-06,
+      "loss": 0.7016,
+      "step": 6424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.644391935122887,
+      "learning_rate": 6.649193903534591e-06,
+      "loss": 0.6885,
+      "step": 6425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4543673801188919,
+      "learning_rate": 6.6482153855692395e-06,
+      "loss": 0.695,
+      "step": 6426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7779930522250904,
+      "learning_rate": 6.647236796776405e-06,
+      "loss": 0.845,
+      "step": 6427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4366476611872603,
+      "learning_rate": 6.646258137198136e-06,
+      "loss": 0.7203,
+      "step": 6428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7609982117251037,
+      "learning_rate": 6.645279406876488e-06,
+      "loss": 0.7488,
+      "step": 6429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7003555588001045,
+      "learning_rate": 6.644300605853518e-06,
+      "loss": 0.6653,
+      "step": 6430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0537215754322322,
+      "learning_rate": 6.643321734171291e-06,
+      "loss": 0.6944,
+      "step": 6431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0603166683765375,
+      "learning_rate": 6.642342791871868e-06,
+      "loss": 0.6997,
+      "step": 6432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0679537832522281,
+      "learning_rate": 6.64136377899732e-06,
+      "loss": 0.6052,
+      "step": 6433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0822136547173111,
+      "learning_rate": 6.640384695589714e-06,
+      "loss": 0.7523,
+      "step": 6434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6205831249425204,
+      "learning_rate": 6.639405541691122e-06,
+      "loss": 0.7483,
+      "step": 6435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6392360924094203,
+      "learning_rate": 6.638426317343624e-06,
+      "loss": 0.7153,
+      "step": 6436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4910574384990878,
+      "learning_rate": 6.6374470225892985e-06,
+      "loss": 0.6936,
+      "step": 6437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3957535201578666,
+      "learning_rate": 6.6364676574702284e-06,
+      "loss": 0.6452,
+      "step": 6438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8416180213579019,
+      "learning_rate": 6.635488222028497e-06,
+      "loss": 0.8374,
+      "step": 6439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4354704194186334,
+      "learning_rate": 6.6345087163061935e-06,
+      "loss": 0.6686,
+      "step": 6440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6466594513345068,
+      "learning_rate": 6.633529140345412e-06,
+      "loss": 0.773,
+      "step": 6441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6458961712012103,
+      "learning_rate": 6.632549494188245e-06,
+      "loss": 0.7315,
+      "step": 6442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9203847678593942,
+      "learning_rate": 6.631569777876791e-06,
+      "loss": 0.7102,
+      "step": 6443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6453241605332616,
+      "learning_rate": 6.630589991453148e-06,
+      "loss": 0.7951,
+      "step": 6444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9450053727209802,
+      "learning_rate": 6.629610134959423e-06,
+      "loss": 0.8283,
+      "step": 6445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7215506233574074,
+      "learning_rate": 6.628630208437719e-06,
+      "loss": 0.7472,
+      "step": 6446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1308802749443085,
+      "learning_rate": 6.627650211930152e-06,
+      "loss": 0.7342,
+      "step": 6447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7490526805979132,
+      "learning_rate": 6.626670145478826e-06,
+      "loss": 0.8463,
+      "step": 6448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5474190692863456,
+      "learning_rate": 6.6256900091258644e-06,
+      "loss": 0.7354,
+      "step": 6449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.807370419779466,
+      "learning_rate": 6.62470980291338e-06,
+      "loss": 0.7543,
+      "step": 6450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.211877563937216,
+      "learning_rate": 6.623729526883498e-06,
+      "loss": 0.6412,
+      "step": 6451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8267119083322494,
+      "learning_rate": 6.622749181078344e-06,
+      "loss": 0.8348,
+      "step": 6452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6671398653498963,
+      "learning_rate": 6.621768765540041e-06,
+      "loss": 0.7347,
+      "step": 6453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6244830516024569,
+      "learning_rate": 6.620788280310722e-06,
+      "loss": 0.8301,
+      "step": 6454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5509810296200799,
+      "learning_rate": 6.619807725432522e-06,
+      "loss": 0.6862,
+      "step": 6455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3658258766597189,
+      "learning_rate": 6.618827100947577e-06,
+      "loss": 0.7776,
+      "step": 6456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5792990126324031,
+      "learning_rate": 6.6178464068980256e-06,
+      "loss": 0.6749,
+      "step": 6457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0264998763780944,
+      "learning_rate": 6.616865643326011e-06,
+      "loss": 0.8795,
+      "step": 6458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.726907434549827,
+      "learning_rate": 6.615884810273678e-06,
+      "loss": 0.8,
+      "step": 6459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8386545771545892,
+      "learning_rate": 6.614903907783177e-06,
+      "loss": 0.8656,
+      "step": 6460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7568356714427458,
+      "learning_rate": 6.613922935896659e-06,
+      "loss": 0.888,
+      "step": 6461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7104568007594112,
+      "learning_rate": 6.612941894656277e-06,
+      "loss": 0.7811,
+      "step": 6462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0486099955626553,
+      "learning_rate": 6.61196078410419e-06,
+      "loss": 0.644,
+      "step": 6463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8919610664393214,
+      "learning_rate": 6.610979604282557e-06,
+      "loss": 0.7855,
+      "step": 6464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9024880684366152,
+      "learning_rate": 6.609998355233543e-06,
+      "loss": 0.7089,
+      "step": 6465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8116543069437643,
+      "learning_rate": 6.609017036999314e-06,
+      "loss": 0.8904,
+      "step": 6466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7063947354960542,
+      "learning_rate": 6.608035649622041e-06,
+      "loss": 0.7534,
+      "step": 6467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9673294393370522,
+      "learning_rate": 6.607054193143894e-06,
+      "loss": 0.7791,
+      "step": 6468
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.870911287561503,
+      "learning_rate": 6.606072667607048e-06,
+      "loss": 0.6461,
+      "step": 6469
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.751270477210211,
+      "learning_rate": 6.605091073053685e-06,
+      "loss": 0.7926,
+      "step": 6470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.503209152136767,
+      "learning_rate": 6.604109409525982e-06,
+      "loss": 0.6779,
+      "step": 6471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9633321558957206,
+      "learning_rate": 6.603127677066126e-06,
+      "loss": 0.8257,
+      "step": 6472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5686193193597262,
+      "learning_rate": 6.602145875716301e-06,
+      "loss": 0.7882,
+      "step": 6473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8821895561438216,
+      "learning_rate": 6.601164005518702e-06,
+      "loss": 0.7399,
+      "step": 6474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5245066792228272,
+      "learning_rate": 6.600182066515519e-06,
+      "loss": 0.6635,
+      "step": 6475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.945782201876192,
+      "learning_rate": 6.599200058748949e-06,
+      "loss": 0.7981,
+      "step": 6476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8508141310033557,
+      "learning_rate": 6.59821798226119e-06,
+      "loss": 0.81,
+      "step": 6477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7432799891183584,
+      "learning_rate": 6.597235837094446e-06,
+      "loss": 0.7132,
+      "step": 6478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7131975930489471,
+      "learning_rate": 6.59625362329092e-06,
+      "loss": 0.7629,
+      "step": 6479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7126706164324768,
+      "learning_rate": 6.595271340892823e-06,
+      "loss": 0.7157,
+      "step": 6480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.813053475514341,
+      "learning_rate": 6.594288989942361e-06,
+      "loss": 0.6749,
+      "step": 6481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2863748279177172,
+      "learning_rate": 6.593306570481751e-06,
+      "loss": 0.6177,
+      "step": 6482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.773721164376934,
+      "learning_rate": 6.592324082553209e-06,
+      "loss": 0.7529,
+      "step": 6483
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6155995098129692,
+      "learning_rate": 6.591341526198955e-06,
+      "loss": 0.6831,
+      "step": 6484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5711363544443169,
+      "learning_rate": 6.590358901461213e-06,
+      "loss": 0.7564,
+      "step": 6485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5579279526434844,
+      "learning_rate": 6.589376208382206e-06,
+      "loss": 0.7435,
+      "step": 6486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6931507630384433,
+      "learning_rate": 6.588393447004163e-06,
+      "loss": 0.7657,
+      "step": 6487
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8517625178069779,
+      "learning_rate": 6.587410617369316e-06,
+      "loss": 0.8232,
+      "step": 6488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8434389918583847,
+      "learning_rate": 6.586427719519901e-06,
+      "loss": 0.6739,
+      "step": 6489
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.684487481111826,
+      "learning_rate": 6.585444753498153e-06,
+      "loss": 0.6908,
+      "step": 6490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.672153774883491,
+      "learning_rate": 6.584461719346312e-06,
+      "loss": 0.7286,
+      "step": 6491
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7490899131158508,
+      "learning_rate": 6.583478617106623e-06,
+      "loss": 0.7591,
+      "step": 6492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8322665458612721,
+      "learning_rate": 6.582495446821331e-06,
+      "loss": 0.8546,
+      "step": 6493
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4876880610927379,
+      "learning_rate": 6.581512208532685e-06,
+      "loss": 0.7979,
+      "step": 6494
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9105940360665468,
+      "learning_rate": 6.580528902282938e-06,
+      "loss": 0.8455,
+      "step": 6495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7702449132514928,
+      "learning_rate": 6.579545528114344e-06,
+      "loss": 0.6637,
+      "step": 6496
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2401729568290387,
+      "learning_rate": 6.57856208606916e-06,
+      "loss": 0.6685,
+      "step": 6497
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1898658505089457,
+      "learning_rate": 6.577578576189648e-06,
+      "loss": 0.6682,
+      "step": 6498
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.796786133107013,
+      "learning_rate": 6.576594998518071e-06,
+      "loss": 0.6995,
+      "step": 6499
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4746993912498445,
+      "learning_rate": 6.575611353096695e-06,
+      "loss": 0.717,
+      "step": 6500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.823421083768313,
+      "learning_rate": 6.574627639967791e-06,
+      "loss": 0.6989,
+      "step": 6501
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7419273340155945,
+      "learning_rate": 6.57364385917363e-06,
+      "loss": 0.6894,
+      "step": 6502
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.593534909396428,
+      "learning_rate": 6.572660010756489e-06,
+      "loss": 0.8439,
+      "step": 6503
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.792691358242924,
+      "learning_rate": 6.5716760947586425e-06,
+      "loss": 0.7195,
+      "step": 6504
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.055892529921534,
+      "learning_rate": 6.570692111222375e-06,
+      "loss": 0.7096,
+      "step": 6505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6087208914863314,
+      "learning_rate": 6.56970806018997e-06,
+      "loss": 0.7024,
+      "step": 6506
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1165387691386264,
+      "learning_rate": 6.568723941703712e-06,
+      "loss": 0.7635,
+      "step": 6507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7017999127867365,
+      "learning_rate": 6.567739755805895e-06,
+      "loss": 0.6915,
+      "step": 6508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0703631203233042,
+      "learning_rate": 6.566755502538806e-06,
+      "loss": 0.7576,
+      "step": 6509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5724752906572297,
+      "learning_rate": 6.565771181944747e-06,
+      "loss": 0.6573,
+      "step": 6510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9428149305840983,
+      "learning_rate": 6.5647867940660105e-06,
+      "loss": 0.7483,
+      "step": 6511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.709884252400947,
+      "learning_rate": 6.563802338944902e-06,
+      "loss": 0.7269,
+      "step": 6512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7301017751443073,
+      "learning_rate": 6.562817816623724e-06,
+      "loss": 0.7268,
+      "step": 6513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7242557253763666,
+      "learning_rate": 6.561833227144784e-06,
+      "loss": 0.8027,
+      "step": 6514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8979095854396295,
+      "learning_rate": 6.560848570550391e-06,
+      "loss": 0.8048,
+      "step": 6515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.637635256131958,
+      "learning_rate": 6.559863846882859e-06,
+      "loss": 0.7013,
+      "step": 6516
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7162177714774176,
+      "learning_rate": 6.558879056184505e-06,
+      "loss": 0.7142,
+      "step": 6517
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8339519065039094,
+      "learning_rate": 6.557894198497644e-06,
+      "loss": 0.7675,
+      "step": 6518
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4941949538558632,
+      "learning_rate": 6.556909273864601e-06,
+      "loss": 0.6256,
+      "step": 6519
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5452252940824345,
+      "learning_rate": 6.555924282327698e-06,
+      "loss": 0.6867,
+      "step": 6520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5942745559873372,
+      "learning_rate": 6.554939223929264e-06,
+      "loss": 0.7317,
+      "step": 6521
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2078628885676013,
+      "learning_rate": 6.553954098711629e-06,
+      "loss": 0.6928,
+      "step": 6522
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7863188696078725,
+      "learning_rate": 6.5529689067171255e-06,
+      "loss": 0.8089,
+      "step": 6523
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9146211239930604,
+      "learning_rate": 6.551983647988089e-06,
+      "loss": 0.8684,
+      "step": 6524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7266575799757622,
+      "learning_rate": 6.550998322566857e-06,
+      "loss": 0.7798,
+      "step": 6525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6701942528945959,
+      "learning_rate": 6.550012930495775e-06,
+      "loss": 0.7335,
+      "step": 6526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5412012396757206,
+      "learning_rate": 6.549027471817186e-06,
+      "loss": 0.7708,
+      "step": 6527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5318697313688285,
+      "learning_rate": 6.5480419465734355e-06,
+      "loss": 0.8576,
+      "step": 6528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.627286607355866,
+      "learning_rate": 6.547056354806874e-06,
+      "loss": 0.754,
+      "step": 6529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.663919444338735,
+      "learning_rate": 6.546070696559856e-06,
+      "loss": 0.7578,
+      "step": 6530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8110147586775853,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.9035,
+      "step": 6531
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2322695655132094,
+      "learning_rate": 6.544099180793877e-06,
+      "loss": 0.6269,
+      "step": 6532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7832638373971745,
+      "learning_rate": 6.543113323359635e-06,
+      "loss": 0.6653,
+      "step": 6533
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.976449057339713,
+      "learning_rate": 6.542127399614376e-06,
+      "loss": 0.7043,
+      "step": 6534
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1885273431228742,
+      "learning_rate": 6.54114140960047e-06,
+      "loss": 0.7767,
+      "step": 6535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6241957076426834,
+      "learning_rate": 6.540155353360284e-06,
+      "loss": 0.693,
+      "step": 6536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7390077765098304,
+      "learning_rate": 6.539169230936193e-06,
+      "loss": 0.689,
+      "step": 6537
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6240877590757892,
+      "learning_rate": 6.5381830423705714e-06,
+      "loss": 0.7307,
+      "step": 6538
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.818228606156132,
+      "learning_rate": 6.5371967877058e-06,
+      "loss": 0.7052,
+      "step": 6539
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.721225920927979,
+      "learning_rate": 6.536210466984259e-06,
+      "loss": 0.7497,
+      "step": 6540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6149071171199592,
+      "learning_rate": 6.535224080248332e-06,
+      "loss": 0.6888,
+      "step": 6541
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9897973633902541,
+      "learning_rate": 6.534237627540408e-06,
+      "loss": 0.7414,
+      "step": 6542
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.110563502795732,
+      "learning_rate": 6.533251108902875e-06,
+      "loss": 0.7779,
+      "step": 6543
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7408882057008364,
+      "learning_rate": 6.532264524378128e-06,
+      "loss": 0.6336,
+      "step": 6544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.578505912581697,
+      "learning_rate": 6.531277874008562e-06,
+      "loss": 0.7989,
+      "step": 6545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6395329644661811,
+      "learning_rate": 6.5302911578365755e-06,
+      "loss": 0.8261,
+      "step": 6546
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.918500488780078,
+      "learning_rate": 6.529304375904568e-06,
+      "loss": 0.7048,
+      "step": 6547
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6677382560299983,
+      "learning_rate": 6.528317528254948e-06,
+      "loss": 0.637,
+      "step": 6548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7213214014978595,
+      "learning_rate": 6.52733061493012e-06,
+      "loss": 0.7143,
+      "step": 6549
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5437075582371975,
+      "learning_rate": 6.526343635972492e-06,
+      "loss": 0.8237,
+      "step": 6550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7204658556922672,
+      "learning_rate": 6.525356591424479e-06,
+      "loss": 0.8076,
+      "step": 6551
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8149395336812788,
+      "learning_rate": 6.5243694813284975e-06,
+      "loss": 0.6433,
+      "step": 6552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.014843601247199,
+      "learning_rate": 6.523382305726964e-06,
+      "loss": 0.7787,
+      "step": 6553
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7384047756853,
+      "learning_rate": 6.522395064662299e-06,
+      "loss": 0.645,
+      "step": 6554
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.637419035228776,
+      "learning_rate": 6.521407758176929e-06,
+      "loss": 0.5896,
+      "step": 6555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1415663097529998,
+      "learning_rate": 6.520420386313277e-06,
+      "loss": 0.7355,
+      "step": 6556
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6720457717654738,
+      "learning_rate": 6.5194329491137775e-06,
+      "loss": 0.7203,
+      "step": 6557
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.752659723860157,
+      "learning_rate": 6.518445446620858e-06,
+      "loss": 0.7761,
+      "step": 6558
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6297016127343815,
+      "learning_rate": 6.517457878876958e-06,
+      "loss": 0.6669,
+      "step": 6559
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7582746361818986,
+      "learning_rate": 6.516470245924511e-06,
+      "loss": 0.7949,
+      "step": 6560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.591736464117771,
+      "learning_rate": 6.515482547805962e-06,
+      "loss": 0.7354,
+      "step": 6561
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.833809836734877,
+      "learning_rate": 6.514494784563753e-06,
+      "loss": 0.7081,
+      "step": 6562
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8582620717213558,
+      "learning_rate": 6.5135069562403285e-06,
+      "loss": 0.687,
+      "step": 6563
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.895331149791442,
+      "learning_rate": 6.512519062878142e-06,
+      "loss": 0.9095,
+      "step": 6564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.464851129821143,
+      "learning_rate": 6.511531104519639e-06,
+      "loss": 0.8414,
+      "step": 6565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0343558367999115,
+      "learning_rate": 6.510543081207281e-06,
+      "loss": 0.5654,
+      "step": 6566
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6204590682317763,
+      "learning_rate": 6.509554992983521e-06,
+      "loss": 0.7279,
+      "step": 6567
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6649361527510247,
+      "learning_rate": 6.508566839890823e-06,
+      "loss": 0.6913,
+      "step": 6568
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6458932297891729,
+      "learning_rate": 6.507578621971646e-06,
+      "loss": 0.8172,
+      "step": 6569
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5216920505749836,
+      "learning_rate": 6.506590339268459e-06,
+      "loss": 0.7659,
+      "step": 6570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6669641276461888,
+      "learning_rate": 6.50560199182373e-06,
+      "loss": 0.8148,
+      "step": 6571
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.339068605507518,
+      "learning_rate": 6.504613579679929e-06,
+      "loss": 0.7475,
+      "step": 6572
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6627253267205004,
+      "learning_rate": 6.503625102879534e-06,
+      "loss": 0.7659,
+      "step": 6573
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9168285232112743,
+      "learning_rate": 6.502636561465018e-06,
+      "loss": 0.8469,
+      "step": 6574
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.088920797281858,
+      "learning_rate": 6.501647955478863e-06,
+      "loss": 0.7715,
+      "step": 6575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6184997849652376,
+      "learning_rate": 6.500659284963549e-06,
+      "loss": 0.8197,
+      "step": 6576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1424340438829599,
+      "learning_rate": 6.499670549961565e-06,
+      "loss": 0.6255,
+      "step": 6577
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6341970482037123,
+      "learning_rate": 6.498681750515398e-06,
+      "loss": 0.6884,
+      "step": 6578
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.751574365035639,
+      "learning_rate": 6.497692886667537e-06,
+      "loss": 0.7392,
+      "step": 6579
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3534345520044675,
+      "learning_rate": 6.496703958460479e-06,
+      "loss": 0.7996,
+      "step": 6580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.557469192915983,
+      "learning_rate": 6.495714965936717e-06,
+      "loss": 0.8076,
+      "step": 6581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7010379253243932,
+      "learning_rate": 6.494725909138751e-06,
+      "loss": 0.7032,
+      "step": 6582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.976148028341218,
+      "learning_rate": 6.493736788109086e-06,
+      "loss": 0.8748,
+      "step": 6583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6625521644658514,
+      "learning_rate": 6.492747602890223e-06,
+      "loss": 0.704,
+      "step": 6584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.524293111643658,
+      "learning_rate": 6.491758353524672e-06,
+      "loss": 0.763,
+      "step": 6585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.728330103091565,
+      "learning_rate": 6.490769040054942e-06,
+      "loss": 0.7916,
+      "step": 6586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.720259364522505,
+      "learning_rate": 6.489779662523545e-06,
+      "loss": 0.5921,
+      "step": 6587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2086841605505894,
+      "learning_rate": 6.4887902209730004e-06,
+      "loss": 0.5933,
+      "step": 6588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7025471447054872,
+      "learning_rate": 6.487800715445822e-06,
+      "loss": 0.8109,
+      "step": 6589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.778811941923827,
+      "learning_rate": 6.486811145984534e-06,
+      "loss": 0.7685,
+      "step": 6590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5275022962293392,
+      "learning_rate": 6.48582151263166e-06,
+      "loss": 0.7192,
+      "step": 6591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7401920245296563,
+      "learning_rate": 6.484831815429726e-06,
+      "loss": 0.7777,
+      "step": 6592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8073157019489619,
+      "learning_rate": 6.483842054421263e-06,
+      "loss": 0.7367,
+      "step": 6593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4602155935109333,
+      "learning_rate": 6.4828522296488014e-06,
+      "loss": 0.7128,
+      "step": 6594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.005251789152734,
+      "learning_rate": 6.481862341154878e-06,
+      "loss": 0.6688,
+      "step": 6595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4476466514306525,
+      "learning_rate": 6.480872388982028e-06,
+      "loss": 0.6995,
+      "step": 6596
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8019021747960517,
+      "learning_rate": 6.479882373172795e-06,
+      "loss": 0.8237,
+      "step": 6597
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4333363175757756,
+      "learning_rate": 6.47889229376972e-06,
+      "loss": 0.7882,
+      "step": 6598
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.771069309449368,
+      "learning_rate": 6.477902150815347e-06,
+      "loss": 0.7384,
+      "step": 6599
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7364692278441483,
+      "learning_rate": 6.476911944352231e-06,
+      "loss": 0.6019,
+      "step": 6600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7541655951686785,
+      "learning_rate": 6.475921674422917e-06,
+      "loss": 0.7839,
+      "step": 6601
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7616871966757786,
+      "learning_rate": 6.474931341069963e-06,
+      "loss": 0.8312,
+      "step": 6602
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.941467611801955,
+      "learning_rate": 6.473940944335923e-06,
+      "loss": 0.7933,
+      "step": 6603
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8243895113533093,
+      "learning_rate": 6.472950484263359e-06,
+      "loss": 0.7571,
+      "step": 6604
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0565976244920714,
+      "learning_rate": 6.471959960894832e-06,
+      "loss": 0.6686,
+      "step": 6605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6880003489235484,
+      "learning_rate": 6.470969374272908e-06,
+      "loss": 0.818,
+      "step": 6606
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0232794813148782,
+      "learning_rate": 6.469978724440153e-06,
+      "loss": 0.6409,
+      "step": 6607
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6872274764904382,
+      "learning_rate": 6.4689880114391375e-06,
+      "loss": 0.6577,
+      "step": 6608
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6979097131207974,
+      "learning_rate": 6.467997235312437e-06,
+      "loss": 0.6719,
+      "step": 6609
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0714615961077336,
+      "learning_rate": 6.4670063961026265e-06,
+      "loss": 0.702,
+      "step": 6610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8307173972572681,
+      "learning_rate": 6.466015493852283e-06,
+      "loss": 0.7398,
+      "step": 6611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2749272976621686,
+      "learning_rate": 6.465024528603987e-06,
+      "loss": 0.7678,
+      "step": 6612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2870214273106875,
+      "learning_rate": 6.464033500400327e-06,
+      "loss": 0.7958,
+      "step": 6613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6897001435771026,
+      "learning_rate": 6.463042409283885e-06,
+      "loss": 0.7769,
+      "step": 6614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5481601830019422,
+      "learning_rate": 6.462051255297255e-06,
+      "loss": 0.7612,
+      "step": 6615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6234127735828112,
+      "learning_rate": 6.461060038483025e-06,
+      "loss": 0.8659,
+      "step": 6616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6123566364473056,
+      "learning_rate": 6.460068758883789e-06,
+      "loss": 0.7481,
+      "step": 6617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.570648606595238,
+      "learning_rate": 6.4590774165421485e-06,
+      "loss": 0.6806,
+      "step": 6618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8527678954139686,
+      "learning_rate": 6.458086011500703e-06,
+      "loss": 0.72,
+      "step": 6619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.765472873363615,
+      "learning_rate": 6.457094543802053e-06,
+      "loss": 0.8584,
+      "step": 6620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.619812012349948,
+      "learning_rate": 6.456103013488804e-06,
+      "loss": 0.7327,
+      "step": 6621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6663806289116665,
+      "learning_rate": 6.455111420603568e-06,
+      "loss": 0.7542,
+      "step": 6622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5593839829405203,
+      "learning_rate": 6.454119765188951e-06,
+      "loss": 0.6946,
+      "step": 6623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6155279126905098,
+      "learning_rate": 6.453128047287573e-06,
+      "loss": 0.7297,
+      "step": 6624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.726745411512166,
+      "learning_rate": 6.452136266942044e-06,
+      "loss": 0.7724,
+      "step": 6625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.684711701233185,
+      "learning_rate": 6.451144424194985e-06,
+      "loss": 0.7252,
+      "step": 6626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8334462121464927,
+      "learning_rate": 6.450152519089019e-06,
+      "loss": 0.7324,
+      "step": 6627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.698071701444226,
+      "learning_rate": 6.449160551666769e-06,
+      "loss": 0.7278,
+      "step": 6628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.755719983908942,
+      "learning_rate": 6.448168521970865e-06,
+      "loss": 0.8148,
+      "step": 6629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7218532032757325,
+      "learning_rate": 6.447176430043931e-06,
+      "loss": 0.8943,
+      "step": 6630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.727334752596468,
+      "learning_rate": 6.446184275928604e-06,
+      "loss": 0.682,
+      "step": 6631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7592112906378004,
+      "learning_rate": 6.445192059667518e-06,
+      "loss": 0.6328,
+      "step": 6632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.948564165720879,
+      "learning_rate": 6.444199781303311e-06,
+      "loss": 0.8568,
+      "step": 6633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8041268965391044,
+      "learning_rate": 6.443207440878624e-06,
+      "loss": 0.8213,
+      "step": 6634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7556507420662344,
+      "learning_rate": 6.442215038436097e-06,
+      "loss": 0.7707,
+      "step": 6635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.669143101141087,
+      "learning_rate": 6.441222574018378e-06,
+      "loss": 0.7799,
+      "step": 6636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7938969917100156,
+      "learning_rate": 6.440230047668116e-06,
+      "loss": 0.7515,
+      "step": 6637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6533293913381153,
+      "learning_rate": 6.4392374594279614e-06,
+      "loss": 0.8371,
+      "step": 6638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8255197778009107,
+      "learning_rate": 6.438244809340568e-06,
+      "loss": 0.6624,
+      "step": 6639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2304912007703894,
+      "learning_rate": 6.4372520974485916e-06,
+      "loss": 0.7268,
+      "step": 6640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6110304011727188,
+      "learning_rate": 6.436259323794691e-06,
+      "loss": 0.7787,
+      "step": 6641
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.914785707775749,
+      "learning_rate": 6.43526648842153e-06,
+      "loss": 0.6964,
+      "step": 6642
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5756985974036706,
+      "learning_rate": 6.434273591371771e-06,
+      "loss": 0.6772,
+      "step": 6643
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5262641930562022,
+      "learning_rate": 6.43328063268808e-06,
+      "loss": 0.5589,
+      "step": 6644
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7557857481354857,
+      "learning_rate": 6.432287612413131e-06,
+      "loss": 0.721,
+      "step": 6645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.243233775987144,
+      "learning_rate": 6.431294530589589e-06,
+      "loss": 0.6064,
+      "step": 6646
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6776143021881698,
+      "learning_rate": 6.4303013872601375e-06,
+      "loss": 0.7272,
+      "step": 6647
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6235357049598116,
+      "learning_rate": 6.4293081824674485e-06,
+      "loss": 0.7194,
+      "step": 6648
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7965789518179065,
+      "learning_rate": 6.428314916254203e-06,
+      "loss": 0.6729,
+      "step": 6649
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8788410709194916,
+      "learning_rate": 6.427321588663085e-06,
+      "loss": 0.6352,
+      "step": 6650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.844838484396105,
+      "learning_rate": 6.42632819973678e-06,
+      "loss": 0.7001,
+      "step": 6651
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.669272043614643,
+      "learning_rate": 6.425334749517975e-06,
+      "loss": 0.7442,
+      "step": 6652
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0196225724904617,
+      "learning_rate": 6.424341238049362e-06,
+      "loss": 0.6117,
+      "step": 6653
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7158456365625023,
+      "learning_rate": 6.423347665373633e-06,
+      "loss": 0.6571,
+      "step": 6654
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0382953729967617,
+      "learning_rate": 6.422354031533485e-06,
+      "loss": 0.8138,
+      "step": 6655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7185149108861006,
+      "learning_rate": 6.421360336571618e-06,
+      "loss": 0.7808,
+      "step": 6656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.002199387711227,
+      "learning_rate": 6.42036658053073e-06,
+      "loss": 0.6758,
+      "step": 6657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0484528773718496,
+      "learning_rate": 6.419372763453529e-06,
+      "loss": 0.7274,
+      "step": 6658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2437843068482914,
+      "learning_rate": 6.418378885382716e-06,
+      "loss": 0.824,
+      "step": 6659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6917843034143167,
+      "learning_rate": 6.417384946361005e-06,
+      "loss": 0.6384,
+      "step": 6660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4189004221708434,
+      "learning_rate": 6.416390946431107e-06,
+      "loss": 0.6966,
+      "step": 6661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.908628194373596,
+      "learning_rate": 6.4153968856357355e-06,
+      "loss": 0.6937,
+      "step": 6662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.723372499086419,
+      "learning_rate": 6.414402764017607e-06,
+      "loss": 0.686,
+      "step": 6663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3993146145996926,
+      "learning_rate": 6.41340858161944e-06,
+      "loss": 0.7527,
+      "step": 6664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8259502430164076,
+      "learning_rate": 6.412414338483961e-06,
+      "loss": 0.8223,
+      "step": 6665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6247657051377682,
+      "learning_rate": 6.411420034653891e-06,
+      "loss": 0.7887,
+      "step": 6666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.535158551405786,
+      "learning_rate": 6.410425670171959e-06,
+      "loss": 0.8091,
+      "step": 6667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5642708972303738,
+      "learning_rate": 6.409431245080894e-06,
+      "loss": 0.7338,
+      "step": 6668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3016764886186063,
+      "learning_rate": 6.408436759423431e-06,
+      "loss": 0.6378,
+      "step": 6669
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3941059964789693,
+      "learning_rate": 6.4074422132423035e-06,
+      "loss": 0.6446,
+      "step": 6670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6162362172401477,
+      "learning_rate": 6.406447606580248e-06,
+      "loss": 0.6635,
+      "step": 6671
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1274713685353706,
+      "learning_rate": 6.405452939480008e-06,
+      "loss": 0.8317,
+      "step": 6672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.670476731434724,
+      "learning_rate": 6.404458211984324e-06,
+      "loss": 0.7577,
+      "step": 6673
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6563821809588475,
+      "learning_rate": 6.403463424135943e-06,
+      "loss": 0.7344,
+      "step": 6674
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1931346474084803,
+      "learning_rate": 6.402468575977614e-06,
+      "loss": 0.663,
+      "step": 6675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8696583098585071,
+      "learning_rate": 6.401473667552085e-06,
+      "loss": 0.7375,
+      "step": 6676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8118335703844222,
+      "learning_rate": 6.400478698902112e-06,
+      "loss": 0.7698,
+      "step": 6677
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6740970552845003,
+      "learning_rate": 6.399483670070451e-06,
+      "loss": 0.8502,
+      "step": 6678
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2039471132724942,
+      "learning_rate": 6.398488581099859e-06,
+      "loss": 0.6594,
+      "step": 6679
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1439984143515105,
+      "learning_rate": 6.3974934320330996e-06,
+      "loss": 0.6988,
+      "step": 6680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9732651544994277,
+      "learning_rate": 6.396498222912933e-06,
+      "loss": 0.921,
+      "step": 6681
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.746736428734097,
+      "learning_rate": 6.395502953782129e-06,
+      "loss": 0.8202,
+      "step": 6682
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6563726943599486,
+      "learning_rate": 6.394507624683455e-06,
+      "loss": 0.7981,
+      "step": 6683
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6664822091753366,
+      "learning_rate": 6.393512235659681e-06,
+      "loss": 0.7967,
+      "step": 6684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8188673105651718,
+      "learning_rate": 6.392516786753586e-06,
+      "loss": 0.6179,
+      "step": 6685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9013572212178194,
+      "learning_rate": 6.391521278007939e-06,
+      "loss": 0.7733,
+      "step": 6686
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.813987920737666,
+      "learning_rate": 6.390525709465527e-06,
+      "loss": 0.7908,
+      "step": 6687
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6772077513891854,
+      "learning_rate": 6.389530081169126e-06,
+      "loss": 0.7335,
+      "step": 6688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.182886061062386,
+      "learning_rate": 6.388534393161525e-06,
+      "loss": 0.6824,
+      "step": 6689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.306604209658895,
+      "learning_rate": 6.38753864548551e-06,
+      "loss": 0.7156,
+      "step": 6690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6160701230593473,
+      "learning_rate": 6.386542838183865e-06,
+      "loss": 0.7353,
+      "step": 6691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5012812452344386,
+      "learning_rate": 6.385546971299389e-06,
+      "loss": 0.797,
+      "step": 6692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0339304185235476,
+      "learning_rate": 6.384551044874871e-06,
+      "loss": 0.7841,
+      "step": 6693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9433502990445706,
+      "learning_rate": 6.383555058953115e-06,
+      "loss": 0.6614,
+      "step": 6694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.237323197083011,
+      "learning_rate": 6.382559013576914e-06,
+      "loss": 0.7733,
+      "step": 6695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8875752540328572,
+      "learning_rate": 6.381562908789075e-06,
+      "loss": 0.8683,
+      "step": 6696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.49575625265192,
+      "learning_rate": 6.380566744632399e-06,
+      "loss": 0.779,
+      "step": 6697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.70738050655598,
+      "learning_rate": 6.379570521149696e-06,
+      "loss": 0.6876,
+      "step": 6698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5505088117820272,
+      "learning_rate": 6.378574238383776e-06,
+      "loss": 0.8159,
+      "step": 6699
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6236119796972102,
+      "learning_rate": 6.377577896377451e-06,
+      "loss": 0.7791,
+      "step": 6700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7708932721910038,
+      "learning_rate": 6.3765814951735346e-06,
+      "loss": 0.7141,
+      "step": 6701
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0839626352611003,
+      "learning_rate": 6.375585034814845e-06,
+      "loss": 0.744,
+      "step": 6702
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4615353346544178,
+      "learning_rate": 6.374588515344204e-06,
+      "loss": 0.7467,
+      "step": 6703
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8824725550828352,
+      "learning_rate": 6.373591936804433e-06,
+      "loss": 0.8665,
+      "step": 6704
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7395578193951162,
+      "learning_rate": 6.372595299238357e-06,
+      "loss": 0.7521,
+      "step": 6705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.762768438533126,
+      "learning_rate": 6.3715986026888046e-06,
+      "loss": 0.6098,
+      "step": 6706
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9138577099493903,
+      "learning_rate": 6.370601847198606e-06,
+      "loss": 0.8284,
+      "step": 6707
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5389884149728712,
+      "learning_rate": 6.3696050328105915e-06,
+      "loss": 0.7573,
+      "step": 6708
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4177974579759844,
+      "learning_rate": 6.3686081595676e-06,
+      "loss": 0.7792,
+      "step": 6709
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5837585420866747,
+      "learning_rate": 6.367611227512468e-06,
+      "loss": 0.7992,
+      "step": 6710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5898445032976818,
+      "learning_rate": 6.366614236688034e-06,
+      "loss": 0.6525,
+      "step": 6711
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1592752838802616,
+      "learning_rate": 6.365617187137144e-06,
+      "loss": 0.7772,
+      "step": 6712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.742108015534675,
+      "learning_rate": 6.3646200789026426e-06,
+      "loss": 0.7748,
+      "step": 6713
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.147447218687791,
+      "learning_rate": 6.3636229120273766e-06,
+      "loss": 0.7307,
+      "step": 6714
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6294847445653025,
+      "learning_rate": 6.362625686554196e-06,
+      "loss": 0.7494,
+      "step": 6715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6246815519036757,
+      "learning_rate": 6.361628402525956e-06,
+      "loss": 0.7001,
+      "step": 6716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9269371982926045,
+      "learning_rate": 6.360631059985512e-06,
+      "loss": 0.7049,
+      "step": 6717
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5261459565335027,
+      "learning_rate": 6.35963365897572e-06,
+      "loss": 0.8163,
+      "step": 6718
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5712631355104223,
+      "learning_rate": 6.3586361995394415e-06,
+      "loss": 0.792,
+      "step": 6719
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8933370614502347,
+      "learning_rate": 6.35763868171954e-06,
+      "loss": 0.7432,
+      "step": 6720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7258646228378838,
+      "learning_rate": 6.3566411055588806e-06,
+      "loss": 0.7777,
+      "step": 6721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.909175514758122,
+      "learning_rate": 6.355643471100332e-06,
+      "loss": 0.8104,
+      "step": 6722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0865322758056712,
+      "learning_rate": 6.354645778386763e-06,
+      "loss": 0.5854,
+      "step": 6723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.629815598784555,
+      "learning_rate": 6.353648027461048e-06,
+      "loss": 0.7382,
+      "step": 6724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5819983017358723,
+      "learning_rate": 6.3526502183660646e-06,
+      "loss": 0.7258,
+      "step": 6725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7732299998323655,
+      "learning_rate": 6.351652351144687e-06,
+      "loss": 0.6953,
+      "step": 6726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.50975352930708,
+      "learning_rate": 6.350654425839799e-06,
+      "loss": 0.824,
+      "step": 6727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7425664627223285,
+      "learning_rate": 6.349656442494281e-06,
+      "loss": 0.6932,
+      "step": 6728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6329893418272958,
+      "learning_rate": 6.348658401151018e-06,
+      "loss": 0.7314,
+      "step": 6729
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7719719750635814,
+      "learning_rate": 6.347660301852903e-06,
+      "loss": 0.801,
+      "step": 6730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7521223614843207,
+      "learning_rate": 6.346662144642823e-06,
+      "loss": 0.696,
+      "step": 6731
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1104585712214605,
+      "learning_rate": 6.345663929563672e-06,
+      "loss": 0.5426,
+      "step": 6732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6816226689126852,
+      "learning_rate": 6.344665656658342e-06,
+      "loss": 0.7841,
+      "step": 6733
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7809335044905112,
+      "learning_rate": 6.343667325969736e-06,
+      "loss": 0.7752,
+      "step": 6734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5116801974668619,
+      "learning_rate": 6.342668937540752e-06,
+      "loss": 0.6882,
+      "step": 6735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5890443080454038,
+      "learning_rate": 6.341670491414294e-06,
+      "loss": 0.6542,
+      "step": 6736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7930162484671683,
+      "learning_rate": 6.3406719876332675e-06,
+      "loss": 0.7541,
+      "step": 6737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5316035480991577,
+      "learning_rate": 6.339673426240578e-06,
+      "loss": 0.7313,
+      "step": 6738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5131027559737895,
+      "learning_rate": 6.3386748072791395e-06,
+      "loss": 0.6093,
+      "step": 6739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7665495765012926,
+      "learning_rate": 6.337676130791861e-06,
+      "loss": 0.659,
+      "step": 6740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4951431018280386,
+      "learning_rate": 6.336677396821663e-06,
+      "loss": 0.74,
+      "step": 6741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6395519676033745,
+      "learning_rate": 6.3356786054114585e-06,
+      "loss": 0.6825,
+      "step": 6742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9048287837381153,
+      "learning_rate": 6.3346797566041695e-06,
+      "loss": 0.7454,
+      "step": 6743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.629689716723928,
+      "learning_rate": 6.33368085044272e-06,
+      "loss": 0.6409,
+      "step": 6744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.154523940546719,
+      "learning_rate": 6.332681886970032e-06,
+      "loss": 0.5869,
+      "step": 6745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7917679197988745,
+      "learning_rate": 6.331682866229039e-06,
+      "loss": 0.8812,
+      "step": 6746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8197249151349006,
+      "learning_rate": 6.3306837882626635e-06,
+      "loss": 0.7466,
+      "step": 6747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7079339541007743,
+      "learning_rate": 6.3296846531138445e-06,
+      "loss": 0.7191,
+      "step": 6748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7136523913406398,
+      "learning_rate": 6.328685460825512e-06,
+      "loss": 0.8743,
+      "step": 6749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5834572848229551,
+      "learning_rate": 6.327686211440609e-06,
+      "loss": 0.7003,
+      "step": 6750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.55534501517448,
+      "learning_rate": 6.32668690500207e-06,
+      "loss": 0.7991,
+      "step": 6751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1303174604566668,
+      "learning_rate": 6.325687541552843e-06,
+      "loss": 0.6455,
+      "step": 6752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7701212507807407,
+      "learning_rate": 6.324688121135868e-06,
+      "loss": 0.7952,
+      "step": 6753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.596169943659126,
+      "learning_rate": 6.323688643794094e-06,
+      "loss": 0.7162,
+      "step": 6754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.905002999693364,
+      "learning_rate": 6.322689109570472e-06,
+      "loss": 0.7582,
+      "step": 6755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7126023775407933,
+      "learning_rate": 6.3216895185079515e-06,
+      "loss": 0.6437,
+      "step": 6756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.098516602097217,
+      "learning_rate": 6.3206898706494915e-06,
+      "loss": 0.765,
+      "step": 6757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1253036958858367,
+      "learning_rate": 6.3196901660380436e-06,
+      "loss": 0.9063,
+      "step": 6758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1041370550747631,
+      "learning_rate": 6.318690404716572e-06,
+      "loss": 0.6923,
+      "step": 6759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7779792755114159,
+      "learning_rate": 6.317690586728036e-06,
+      "loss": 0.7784,
+      "step": 6760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3630597068384998,
+      "learning_rate": 6.316690712115401e-06,
+      "loss": 0.7733,
+      "step": 6761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6729050199007283,
+      "learning_rate": 6.315690780921634e-06,
+      "loss": 0.651,
+      "step": 6762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0303857665160365,
+      "learning_rate": 6.314690793189703e-06,
+      "loss": 0.6423,
+      "step": 6763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9686593790017923,
+      "learning_rate": 6.313690748962582e-06,
+      "loss": 0.8244,
+      "step": 6764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9163101583151383,
+      "learning_rate": 6.312690648283243e-06,
+      "loss": 0.7491,
+      "step": 6765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.643918602833632,
+      "learning_rate": 6.311690491194664e-06,
+      "loss": 0.9076,
+      "step": 6766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.784054412710899,
+      "learning_rate": 6.310690277739821e-06,
+      "loss": 0.7479,
+      "step": 6767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7488925491207095,
+      "learning_rate": 6.3096900079617e-06,
+      "loss": 0.8147,
+      "step": 6768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7368475073917442,
+      "learning_rate": 6.3086896819032814e-06,
+      "loss": 0.7092,
+      "step": 6769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9016314583064422,
+      "learning_rate": 6.307689299607555e-06,
+      "loss": 0.769,
+      "step": 6770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9720194548896823,
+      "learning_rate": 6.306688861117505e-06,
+      "loss": 0.8385,
+      "step": 6771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8168540584279944,
+      "learning_rate": 6.305688366476124e-06,
+      "loss": 0.7707,
+      "step": 6772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7770915819175477,
+      "learning_rate": 6.304687815726406e-06,
+      "loss": 0.7409,
+      "step": 6773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8075218217954763,
+      "learning_rate": 6.303687208911348e-06,
+      "loss": 0.8525,
+      "step": 6774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6669631970627687,
+      "learning_rate": 6.302686546073946e-06,
+      "loss": 0.6605,
+      "step": 6775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6122043208999863,
+      "learning_rate": 6.301685827257202e-06,
+      "loss": 0.7242,
+      "step": 6776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5553961940167718,
+      "learning_rate": 6.300685052504119e-06,
+      "loss": 0.6372,
+      "step": 6777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0219783135904064,
+      "learning_rate": 6.299684221857705e-06,
+      "loss": 0.6907,
+      "step": 6778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6496289771260666,
+      "learning_rate": 6.298683335360962e-06,
+      "loss": 0.712,
+      "step": 6779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0550567891649676,
+      "learning_rate": 6.297682393056906e-06,
+      "loss": 0.736,
+      "step": 6780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6678502148924004,
+      "learning_rate": 6.296681394988546e-06,
+      "loss": 0.7079,
+      "step": 6781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8120235078951135,
+      "learning_rate": 6.2956803411989e-06,
+      "loss": 0.7147,
+      "step": 6782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9582047212559033,
+      "learning_rate": 6.294679231730983e-06,
+      "loss": 0.718,
+      "step": 6783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6633214229973114,
+      "learning_rate": 6.293678066627816e-06,
+      "loss": 0.7691,
+      "step": 6784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.908060131189833,
+      "learning_rate": 6.292676845932422e-06,
+      "loss": 0.7412,
+      "step": 6785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4674217507780303,
+      "learning_rate": 6.291675569687824e-06,
+      "loss": 0.699,
+      "step": 6786
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5096678063260418,
+      "learning_rate": 6.290674237937051e-06,
+      "loss": 0.7667,
+      "step": 6787
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.991369408798374,
+      "learning_rate": 6.289672850723132e-06,
+      "loss": 0.7416,
+      "step": 6788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.564079269369797,
+      "learning_rate": 6.288671408089098e-06,
+      "loss": 0.7366,
+      "step": 6789
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.595035483508209,
+      "learning_rate": 6.2876699100779815e-06,
+      "loss": 0.7446,
+      "step": 6790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9262363202072015,
+      "learning_rate": 6.2866683567328235e-06,
+      "loss": 0.7714,
+      "step": 6791
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9764761919473401,
+      "learning_rate": 6.285666748096659e-06,
+      "loss": 0.7349,
+      "step": 6792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7381306744278164,
+      "learning_rate": 6.284665084212533e-06,
+      "loss": 0.8203,
+      "step": 6793
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.890846573496289,
+      "learning_rate": 6.283663365123486e-06,
+      "loss": 0.7284,
+      "step": 6794
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7769671618175011,
+      "learning_rate": 6.282661590872565e-06,
+      "loss": 0.7435,
+      "step": 6795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5549006369724332,
+      "learning_rate": 6.281659761502819e-06,
+      "loss": 0.6891,
+      "step": 6796
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5367716829035072,
+      "learning_rate": 6.2806578770573e-06,
+      "loss": 0.6896,
+      "step": 6797
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.580904481175134,
+      "learning_rate": 6.279655937579058e-06,
+      "loss": 0.7585,
+      "step": 6798
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1827823473904502,
+      "learning_rate": 6.278653943111152e-06,
+      "loss": 0.8034,
+      "step": 6799
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5100039309307425,
+      "learning_rate": 6.277651893696637e-06,
+      "loss": 0.7178,
+      "step": 6800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6411696932013005,
+      "learning_rate": 6.276649789378575e-06,
+      "loss": 0.7475,
+      "step": 6801
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7794755897994807,
+      "learning_rate": 6.275647630200029e-06,
+      "loss": 0.7417,
+      "step": 6802
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9986334571153598,
+      "learning_rate": 6.274645416204062e-06,
+      "loss": 0.7763,
+      "step": 6803
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8365200791317875,
+      "learning_rate": 6.273643147433743e-06,
+      "loss": 0.6387,
+      "step": 6804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.571662610673014,
+      "learning_rate": 6.272640823932141e-06,
+      "loss": 0.6117,
+      "step": 6805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.531570815412428,
+      "learning_rate": 6.271638445742332e-06,
+      "loss": 0.8025,
+      "step": 6806
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.222859270692711,
+      "learning_rate": 6.270636012907383e-06,
+      "loss": 0.7762,
+      "step": 6807
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9058583191592204,
+      "learning_rate": 6.269633525470376e-06,
+      "loss": 0.7522,
+      "step": 6808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6580038733236517,
+      "learning_rate": 6.268630983474388e-06,
+      "loss": 0.8327,
+      "step": 6809
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6970040958279644,
+      "learning_rate": 6.267628386962502e-06,
+      "loss": 0.7202,
+      "step": 6810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.450351526437985,
+      "learning_rate": 6.266625735977802e-06,
+      "loss": 0.7956,
+      "step": 6811
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7379959268374099,
+      "learning_rate": 6.265623030563372e-06,
+      "loss": 0.8319,
+      "step": 6812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.413316474157818,
+      "learning_rate": 6.264620270762302e-06,
+      "loss": 0.7105,
+      "step": 6813
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9991404652716109,
+      "learning_rate": 6.263617456617681e-06,
+      "loss": 0.6135,
+      "step": 6814
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7637516423086474,
+      "learning_rate": 6.262614588172607e-06,
+      "loss": 0.9054,
+      "step": 6815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5751868431823122,
+      "learning_rate": 6.261611665470168e-06,
+      "loss": 0.6615,
+      "step": 6816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4561770332332558,
+      "learning_rate": 6.260608688553469e-06,
+      "loss": 0.666,
+      "step": 6817
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3932911603098481,
+      "learning_rate": 6.259605657465607e-06,
+      "loss": 0.6935,
+      "step": 6818
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8862823487473817,
+      "learning_rate": 6.258602572249683e-06,
+      "loss": 0.7815,
+      "step": 6819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2468280593886072,
+      "learning_rate": 6.2575994329488045e-06,
+      "loss": 0.6821,
+      "step": 6820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7947845610207938,
+      "learning_rate": 6.2565962396060774e-06,
+      "loss": 0.7029,
+      "step": 6821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8800523151253272,
+      "learning_rate": 6.255592992264612e-06,
+      "loss": 0.8438,
+      "step": 6822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6040272212773854,
+      "learning_rate": 6.254589690967517e-06,
+      "loss": 0.6493,
+      "step": 6823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8796553357917298,
+      "learning_rate": 6.2535863357579105e-06,
+      "loss": 0.6003,
+      "step": 6824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.683080231839839,
+      "learning_rate": 6.252582926678908e-06,
+      "loss": 0.8097,
+      "step": 6825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.777177060207291,
+      "learning_rate": 6.2515794637736275e-06,
+      "loss": 0.7418,
+      "step": 6826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6530406841366427,
+      "learning_rate": 6.250575947085189e-06,
+      "loss": 0.7409,
+      "step": 6827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6355570545287772,
+      "learning_rate": 6.2495723766567165e-06,
+      "loss": 0.7657,
+      "step": 6828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5025111727622726,
+      "learning_rate": 6.248568752531337e-06,
+      "loss": 0.6551,
+      "step": 6829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2665084742496053,
+      "learning_rate": 6.247565074752177e-06,
+      "loss": 0.591,
+      "step": 6830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2565429973173945,
+      "learning_rate": 6.246561343362368e-06,
+      "loss": 0.7398,
+      "step": 6831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8032282251097131,
+      "learning_rate": 6.24555755840504e-06,
+      "loss": 0.7498,
+      "step": 6832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.70890921600025,
+      "learning_rate": 6.244553719923332e-06,
+      "loss": 0.6651,
+      "step": 6833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7894457585093713,
+      "learning_rate": 6.243549827960378e-06,
+      "loss": 0.6915,
+      "step": 6834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7847383706350757,
+      "learning_rate": 6.242545882559318e-06,
+      "loss": 0.7983,
+      "step": 6835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1372964282177072,
+      "learning_rate": 6.241541883763294e-06,
+      "loss": 0.7017,
+      "step": 6836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2769251457321724,
+      "learning_rate": 6.240537831615451e-06,
+      "loss": 0.6551,
+      "step": 6837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.593715535531433,
+      "learning_rate": 6.239533726158934e-06,
+      "loss": 0.821,
+      "step": 6838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9751215411761331,
+      "learning_rate": 6.238529567436892e-06,
+      "loss": 0.751,
+      "step": 6839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8002208531526658,
+      "learning_rate": 6.2375253554924766e-06,
+      "loss": 0.7401,
+      "step": 6840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0733060217980634,
+      "learning_rate": 6.23652109036884e-06,
+      "loss": 0.7484,
+      "step": 6841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.077468956780807,
+      "learning_rate": 6.235516772109138e-06,
+      "loss": 0.6892,
+      "step": 6842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5427847838245927,
+      "learning_rate": 6.23451240075653e-06,
+      "loss": 0.8221,
+      "step": 6843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4987754128656294,
+      "learning_rate": 6.233507976354174e-06,
+      "loss": 0.6731,
+      "step": 6844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6042526065364244,
+      "learning_rate": 6.2325034989452335e-06,
+      "loss": 0.7321,
+      "step": 6845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.340344816451215,
+      "learning_rate": 6.231498968572872e-06,
+      "loss": 0.6279,
+      "step": 6846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6278737135009118,
+      "learning_rate": 6.230494385280257e-06,
+      "loss": 0.7627,
+      "step": 6847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6843267561211626,
+      "learning_rate": 6.229489749110559e-06,
+      "loss": 0.7363,
+      "step": 6848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8063691118771656,
+      "learning_rate": 6.228485060106948e-06,
+      "loss": 0.8089,
+      "step": 6849
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6647111595743087,
+      "learning_rate": 6.227480318312597e-06,
+      "loss": 0.7598,
+      "step": 6850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6783171978076465,
+      "learning_rate": 6.226475523770683e-06,
+      "loss": 0.6399,
+      "step": 6851
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5634447356917105,
+      "learning_rate": 6.225470676524385e-06,
+      "loss": 0.7489,
+      "step": 6852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8367759053146218,
+      "learning_rate": 6.224465776616883e-06,
+      "loss": 0.8375,
+      "step": 6853
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5081245195988138,
+      "learning_rate": 6.223460824091358e-06,
+      "loss": 0.7129,
+      "step": 6854
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5374025961437816,
+      "learning_rate": 6.222455818990998e-06,
+      "loss": 0.8471,
+      "step": 6855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.792103427353209,
+      "learning_rate": 6.221450761358988e-06,
+      "loss": 0.8139,
+      "step": 6856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.721792709951181,
+      "learning_rate": 6.220445651238519e-06,
+      "loss": 0.6878,
+      "step": 6857
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8081066222716466,
+      "learning_rate": 6.2194404886727835e-06,
+      "loss": 0.6711,
+      "step": 6858
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6650591105133667,
+      "learning_rate": 6.218435273704973e-06,
+      "loss": 0.7278,
+      "step": 6859
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.620585749478713,
+      "learning_rate": 6.217430006378285e-06,
+      "loss": 0.7499,
+      "step": 6860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.563731389200141,
+      "learning_rate": 6.216424686735917e-06,
+      "loss": 0.7055,
+      "step": 6861
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9122137855771237,
+      "learning_rate": 6.215419314821074e-06,
+      "loss": 0.589,
+      "step": 6862
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6830672815220713,
+      "learning_rate": 6.214413890676956e-06,
+      "loss": 0.8653,
+      "step": 6863
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2329155246650674,
+      "learning_rate": 6.213408414346765e-06,
+      "loss": 0.6645,
+      "step": 6864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.71431575211973,
+      "learning_rate": 6.212402885873715e-06,
+      "loss": 0.7318,
+      "step": 6865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.955389175527135,
+      "learning_rate": 6.211397305301011e-06,
+      "loss": 0.7676,
+      "step": 6866
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7803332626353419,
+      "learning_rate": 6.210391672671869e-06,
+      "loss": 0.6982,
+      "step": 6867
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7158566581033092,
+      "learning_rate": 6.2093859880295e-06,
+      "loss": 0.8374,
+      "step": 6868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7087007021949914,
+      "learning_rate": 6.208380251417122e-06,
+      "loss": 0.717,
+      "step": 6869
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1009896154686172,
+      "learning_rate": 6.207374462877953e-06,
+      "loss": 0.6253,
+      "step": 6870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6764464961408103,
+      "learning_rate": 6.2063686224552144e-06,
+      "loss": 0.7616,
+      "step": 6871
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9131052114527232,
+      "learning_rate": 6.205362730192131e-06,
+      "loss": 0.6751,
+      "step": 6872
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6780449605218442,
+      "learning_rate": 6.2043567861319245e-06,
+      "loss": 0.7637,
+      "step": 6873
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6921824610191654,
+      "learning_rate": 6.203350790317825e-06,
+      "loss": 0.8908,
+      "step": 6874
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8393763591437633,
+      "learning_rate": 6.2023447427930615e-06,
+      "loss": 0.7909,
+      "step": 6875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0504935163567883,
+      "learning_rate": 6.201338643600869e-06,
+      "loss": 0.6362,
+      "step": 6876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6277658367944992,
+      "learning_rate": 6.2003324927844765e-06,
+      "loss": 0.8965,
+      "step": 6877
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9330723789643585,
+      "learning_rate": 6.199326290387126e-06,
+      "loss": 0.8139,
+      "step": 6878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.994861452718062,
+      "learning_rate": 6.198320036452051e-06,
+      "loss": 0.7303,
+      "step": 6879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1445396456588444,
+      "learning_rate": 6.197313731022497e-06,
+      "loss": 0.615,
+      "step": 6880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6449721717186503,
+      "learning_rate": 6.196307374141707e-06,
+      "loss": 0.7919,
+      "step": 6881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6503227221597068,
+      "learning_rate": 6.195300965852923e-06,
+      "loss": 0.8651,
+      "step": 6882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.43275361520037,
+      "learning_rate": 6.194294506199394e-06,
+      "loss": 0.8102,
+      "step": 6883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9556160885734861,
+      "learning_rate": 6.193287995224371e-06,
+      "loss": 0.7272,
+      "step": 6884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6943328090591516,
+      "learning_rate": 6.192281432971104e-06,
+      "loss": 0.7867,
+      "step": 6885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5234249327222154,
+      "learning_rate": 6.19127481948285e-06,
+      "loss": 0.7714,
+      "step": 6886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7504585106727986,
+      "learning_rate": 6.1902681548028646e-06,
+      "loss": 0.6881,
+      "step": 6887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6464322132747813,
+      "learning_rate": 6.189261438974403e-06,
+      "loss": 0.7312,
+      "step": 6888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5880243221121686,
+      "learning_rate": 6.18825467204073e-06,
+      "loss": 0.6926,
+      "step": 6889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6576300330339848,
+      "learning_rate": 6.187247854045109e-06,
+      "loss": 0.7746,
+      "step": 6890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5437765675764537,
+      "learning_rate": 6.186240985030802e-06,
+      "loss": 0.6497,
+      "step": 6891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6236537675261806,
+      "learning_rate": 6.1852340650410784e-06,
+      "loss": 0.6899,
+      "step": 6892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5212484574252996,
+      "learning_rate": 6.184227094119207e-06,
+      "loss": 0.7125,
+      "step": 6893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5552077901081591,
+      "learning_rate": 6.183220072308459e-06,
+      "loss": 0.7456,
+      "step": 6894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.485753864170484,
+      "learning_rate": 6.1822129996521105e-06,
+      "loss": 0.7921,
+      "step": 6895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5886566148204146,
+      "learning_rate": 6.181205876193437e-06,
+      "loss": 0.7068,
+      "step": 6896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9055695066627427,
+      "learning_rate": 6.180198701975715e-06,
+      "loss": 0.7126,
+      "step": 6897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6855486922503173,
+      "learning_rate": 6.179191477042227e-06,
+      "loss": 0.8144,
+      "step": 6898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2842355286595186,
+      "learning_rate": 6.178184201436256e-06,
+      "loss": 0.6337,
+      "step": 6899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.102781561338476,
+      "learning_rate": 6.177176875201086e-06,
+      "loss": 0.5567,
+      "step": 6900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6923358836416278,
+      "learning_rate": 6.176169498380001e-06,
+      "loss": 0.6752,
+      "step": 6901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6070703499259604,
+      "learning_rate": 6.175162071016295e-06,
+      "loss": 0.6932,
+      "step": 6902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6284325288306642,
+      "learning_rate": 6.174154593153258e-06,
+      "loss": 0.7096,
+      "step": 6903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0902850064376324,
+      "learning_rate": 6.173147064834183e-06,
+      "loss": 0.6579,
+      "step": 6904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8641127738890546,
+      "learning_rate": 6.1721394861023665e-06,
+      "loss": 0.6629,
+      "step": 6905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4471942404891904,
+      "learning_rate": 6.171131857001104e-06,
+      "loss": 0.6891,
+      "step": 6906
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4882286134906786,
+      "learning_rate": 6.170124177573698e-06,
+      "loss": 0.668,
+      "step": 6907
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6106658287895397,
+      "learning_rate": 6.169116447863451e-06,
+      "loss": 0.7851,
+      "step": 6908
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7006169709306174,
+      "learning_rate": 6.168108667913666e-06,
+      "loss": 0.8801,
+      "step": 6909
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7037322756580444,
+      "learning_rate": 6.16710083776765e-06,
+      "loss": 0.842,
+      "step": 6910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.770629089832967,
+      "learning_rate": 6.1660929574687115e-06,
+      "loss": 0.7769,
+      "step": 6911
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6778231377208264,
+      "learning_rate": 6.165085027060162e-06,
+      "loss": 0.7307,
+      "step": 6912
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5386935399499928,
+      "learning_rate": 6.164077046585314e-06,
+      "loss": 0.6394,
+      "step": 6913
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8439012251290294,
+      "learning_rate": 6.163069016087483e-06,
+      "loss": 0.754,
+      "step": 6914
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5309761919852802,
+      "learning_rate": 6.162060935609985e-06,
+      "loss": 0.7586,
+      "step": 6915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8486752680303762,
+      "learning_rate": 6.161052805196141e-06,
+      "loss": 0.727,
+      "step": 6916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4695029804101605,
+      "learning_rate": 6.160044624889272e-06,
+      "loss": 0.7506,
+      "step": 6917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3207683951646287,
+      "learning_rate": 6.1590363947327025e-06,
+      "loss": 0.6929,
+      "step": 6918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9165796099872539,
+      "learning_rate": 6.158028114769758e-06,
+      "loss": 0.8659,
+      "step": 6919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5765773859986703,
+      "learning_rate": 6.157019785043765e-06,
+      "loss": 0.7573,
+      "step": 6920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.196734987235353,
+      "learning_rate": 6.156011405598057e-06,
+      "loss": 0.827,
+      "step": 6921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6795213197843297,
+      "learning_rate": 6.155002976475962e-06,
+      "loss": 0.7608,
+      "step": 6922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5396106101825948,
+      "learning_rate": 6.15399449772082e-06,
+      "loss": 0.706,
+      "step": 6923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7176807268251575,
+      "learning_rate": 6.152985969375962e-06,
+      "loss": 0.7055,
+      "step": 6924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.450749403496666,
+      "learning_rate": 6.151977391484729e-06,
+      "loss": 0.7135,
+      "step": 6925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.856069947033487,
+      "learning_rate": 6.150968764090461e-06,
+      "loss": 0.6591,
+      "step": 6926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5140581451455355,
+      "learning_rate": 6.149960087236505e-06,
+      "loss": 0.7297,
+      "step": 6927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8739945300696654,
+      "learning_rate": 6.1489513609662025e-06,
+      "loss": 0.893,
+      "step": 6928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4853356899601118,
+      "learning_rate": 6.147942585322898e-06,
+      "loss": 0.7718,
+      "step": 6929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6205865385415992,
+      "learning_rate": 6.146933760349947e-06,
+      "loss": 0.7072,
+      "step": 6930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6236356237679481,
+      "learning_rate": 6.145924886090695e-06,
+      "loss": 0.7707,
+      "step": 6931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4862622255208862,
+      "learning_rate": 6.144915962588503e-06,
+      "loss": 0.6368,
+      "step": 6932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5142072122100607,
+      "learning_rate": 6.143906989886718e-06,
+      "loss": 0.65,
+      "step": 6933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6166146202455607,
+      "learning_rate": 6.142897968028704e-06,
+      "loss": 0.7659,
+      "step": 6934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5969538757301955,
+      "learning_rate": 6.141888897057817e-06,
+      "loss": 0.7715,
+      "step": 6935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5699459364017496,
+      "learning_rate": 6.140879777017423e-06,
+      "loss": 0.6468,
+      "step": 6936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6735526231345943,
+      "learning_rate": 6.139870607950885e-06,
+      "loss": 0.7816,
+      "step": 6937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.61925797062294,
+      "learning_rate": 6.138861389901566e-06,
+      "loss": 0.7017,
+      "step": 6938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1151882773947017,
+      "learning_rate": 6.137852122912839e-06,
+      "loss": 0.7151,
+      "step": 6939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7408037229722315,
+      "learning_rate": 6.136842807028069e-06,
+      "loss": 0.7986,
+      "step": 6940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.2027877750934675,
+      "learning_rate": 6.135833442290634e-06,
+      "loss": 0.8789,
+      "step": 6941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6956925687523925,
+      "learning_rate": 6.134824028743906e-06,
+      "loss": 0.7587,
+      "step": 6942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.54293468370692,
+      "learning_rate": 6.133814566431262e-06,
+      "loss": 0.7854,
+      "step": 6943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.991889229759348,
+      "learning_rate": 6.1328050553960804e-06,
+      "loss": 0.6926,
+      "step": 6944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.476206487077849,
+      "learning_rate": 6.131795495681743e-06,
+      "loss": 0.6961,
+      "step": 6945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.963983262300856,
+      "learning_rate": 6.130785887331633e-06,
+      "loss": 0.8436,
+      "step": 6946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7786660464486195,
+      "learning_rate": 6.129776230389135e-06,
+      "loss": 0.6787,
+      "step": 6947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8958831625387018,
+      "learning_rate": 6.128766524897636e-06,
+      "loss": 0.8314,
+      "step": 6948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7902162799206593,
+      "learning_rate": 6.1277567709005245e-06,
+      "loss": 0.7361,
+      "step": 6949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5271393359710883,
+      "learning_rate": 6.126746968441195e-06,
+      "loss": 0.7805,
+      "step": 6950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6855521377455827,
+      "learning_rate": 6.1257371175630375e-06,
+      "loss": 0.7615,
+      "step": 6951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7319594091698998,
+      "learning_rate": 6.124727218309449e-06,
+      "loss": 0.7558,
+      "step": 6952
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0769900067051097,
+      "learning_rate": 6.123717270723827e-06,
+      "loss": 0.7037,
+      "step": 6953
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0445551672983595,
+      "learning_rate": 6.122707274849572e-06,
+      "loss": 0.7631,
+      "step": 6954
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7159773236705895,
+      "learning_rate": 6.121697230730084e-06,
+      "loss": 0.7964,
+      "step": 6955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6103408417728389,
+      "learning_rate": 6.120687138408768e-06,
+      "loss": 0.7742,
+      "step": 6956
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1761186744512226,
+      "learning_rate": 6.11967699792903e-06,
+      "loss": 0.7465,
+      "step": 6957
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9379858505623122,
+      "learning_rate": 6.118666809334277e-06,
+      "loss": 0.7706,
+      "step": 6958
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.790362142753267,
+      "learning_rate": 6.117656572667921e-06,
+      "loss": 0.77,
+      "step": 6959
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5241037056536066,
+      "learning_rate": 6.116646287973374e-06,
+      "loss": 0.7434,
+      "step": 6960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7142099508330604,
+      "learning_rate": 6.115635955294047e-06,
+      "loss": 0.7897,
+      "step": 6961
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7445466491834005,
+      "learning_rate": 6.114625574673359e-06,
+      "loss": 0.7292,
+      "step": 6962
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8429875862451301,
+      "learning_rate": 6.1136151461547266e-06,
+      "loss": 0.8248,
+      "step": 6963
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1437217061454266,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.6912,
+      "step": 6964
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1671140984112793,
+      "learning_rate": 6.111594145597319e-06,
+      "loss": 0.7215,
+      "step": 6965
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5234004991429702,
+      "learning_rate": 6.1105835736453874e-06,
+      "loss": 0.7145,
+      "step": 6966
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.068031756596755,
+      "learning_rate": 6.109572953969206e-06,
+      "loss": 0.6477,
+      "step": 6967
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.712153631482105,
+      "learning_rate": 6.108562286612204e-06,
+      "loss": 0.7432,
+      "step": 6968
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7753337297187044,
+      "learning_rate": 6.107551571617813e-06,
+      "loss": 0.7908,
+      "step": 6969
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1863794551441862,
+      "learning_rate": 6.106540809029463e-06,
+      "loss": 0.621,
+      "step": 6970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.79241384729579,
+      "learning_rate": 6.105529998890589e-06,
+      "loss": 0.704,
+      "step": 6971
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6634996139023233,
+      "learning_rate": 6.104519141244631e-06,
+      "loss": 0.671,
+      "step": 6972
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4636937890052526,
+      "learning_rate": 6.103508236135025e-06,
+      "loss": 0.7365,
+      "step": 6973
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8588709224037288,
+      "learning_rate": 6.1024972836052135e-06,
+      "loss": 0.8292,
+      "step": 6974
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8756369818847203,
+      "learning_rate": 6.1014862836986364e-06,
+      "loss": 0.7218,
+      "step": 6975
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6413622825742904,
+      "learning_rate": 6.10047523645874e-06,
+      "loss": 0.7165,
+      "step": 6976
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1657807454436768,
+      "learning_rate": 6.099464141928973e-06,
+      "loss": 0.6361,
+      "step": 6977
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7817351208998848,
+      "learning_rate": 6.098453000152784e-06,
+      "loss": 0.7602,
+      "step": 6978
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6276095284916112,
+      "learning_rate": 6.0974418111736235e-06,
+      "loss": 0.8104,
+      "step": 6979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1533969390698668,
+      "learning_rate": 6.096430575034942e-06,
+      "loss": 0.7331,
+      "step": 6980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.18588159699368,
+      "learning_rate": 6.095419291780198e-06,
+      "loss": 0.7611,
+      "step": 6981
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5675098420662068,
+      "learning_rate": 6.0944079614528475e-06,
+      "loss": 0.7283,
+      "step": 6982
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7609196475772033,
+      "learning_rate": 6.0933965840963494e-06,
+      "loss": 0.8036,
+      "step": 6983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.116639381746495,
+      "learning_rate": 6.092385159754165e-06,
+      "loss": 0.7572,
+      "step": 6984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5302887561894918,
+      "learning_rate": 6.091373688469757e-06,
+      "loss": 0.7856,
+      "step": 6985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6427851628681491,
+      "learning_rate": 6.090362170286591e-06,
+      "loss": 0.7663,
+      "step": 6986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7989444991624588,
+      "learning_rate": 6.089350605248134e-06,
+      "loss": 0.5954,
+      "step": 6987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.868556718251715,
+      "learning_rate": 6.088338993397857e-06,
+      "loss": 0.709,
+      "step": 6988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.496221393746604,
+      "learning_rate": 6.0873273347792275e-06,
+      "loss": 0.6536,
+      "step": 6989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2415551421027229,
+      "learning_rate": 6.086315629435722e-06,
+      "loss": 0.6606,
+      "step": 6990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.592140227181187,
+      "learning_rate": 6.085303877410816e-06,
+      "loss": 0.8019,
+      "step": 6991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5956373506432178,
+      "learning_rate": 6.084292078747982e-06,
+      "loss": 0.6976,
+      "step": 6992
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7561374697256547,
+      "learning_rate": 6.0832802334907044e-06,
+      "loss": 0.85,
+      "step": 6993
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4712096671694572,
+      "learning_rate": 6.0822683416824625e-06,
+      "loss": 0.7385,
+      "step": 6994
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.463588425957546,
+      "learning_rate": 6.08125640336674e-06,
+      "loss": 0.748,
+      "step": 6995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8352424863612593,
+      "learning_rate": 6.08024441858702e-06,
+      "loss": 0.7759,
+      "step": 6996
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8513458239644427,
+      "learning_rate": 6.079232387386793e-06,
+      "loss": 0.6962,
+      "step": 6997
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7749318055044432,
+      "learning_rate": 6.078220309809546e-06,
+      "loss": 0.7357,
+      "step": 6998
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6278371048409195,
+      "learning_rate": 6.077208185898772e-06,
+      "loss": 0.7956,
+      "step": 6999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6795013266706151,
+      "learning_rate": 6.076196015697963e-06,
+      "loss": 0.75,
+      "step": 7000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.603189552967509,
+      "learning_rate": 6.0751837992506126e-06,
+      "loss": 0.6932,
+      "step": 7001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7499992406174345,
+      "learning_rate": 6.074171536600222e-06,
+      "loss": 0.792,
+      "step": 7002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5497853251822435,
+      "learning_rate": 6.073159227790287e-06,
+      "loss": 0.8377,
+      "step": 7003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6611683188811253,
+      "learning_rate": 6.07214687286431e-06,
+      "loss": 0.6863,
+      "step": 7004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.565398532494567,
+      "learning_rate": 6.0711344718657934e-06,
+      "loss": 0.8827,
+      "step": 7005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7761234452751529,
+      "learning_rate": 6.070122024838245e-06,
+      "loss": 0.8373,
+      "step": 7006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8694278948370693,
+      "learning_rate": 6.069109531825169e-06,
+      "loss": 0.7514,
+      "step": 7007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4627082277295709,
+      "learning_rate": 6.068096992870075e-06,
+      "loss": 0.7686,
+      "step": 7008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6840389038849244,
+      "learning_rate": 6.067084408016475e-06,
+      "loss": 0.6845,
+      "step": 7009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7830489373027405,
+      "learning_rate": 6.06607177730788e-06,
+      "loss": 0.8029,
+      "step": 7010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5353153922050173,
+      "learning_rate": 6.065059100787808e-06,
+      "loss": 0.6947,
+      "step": 7011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1745156337788973,
+      "learning_rate": 6.064046378499775e-06,
+      "loss": 0.9262,
+      "step": 7012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8631110713134007,
+      "learning_rate": 6.063033610487298e-06,
+      "loss": 0.7341,
+      "step": 7013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8573311582309953,
+      "learning_rate": 6.0620207967939e-06,
+      "loss": 0.7251,
+      "step": 7014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6811207824229002,
+      "learning_rate": 6.061007937463101e-06,
+      "loss": 0.6912,
+      "step": 7015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7009785547611485,
+      "learning_rate": 6.05999503253843e-06,
+      "loss": 0.6915,
+      "step": 7016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7379816940487531,
+      "learning_rate": 6.058982082063413e-06,
+      "loss": 0.8303,
+      "step": 7017
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.676775922970699,
+      "learning_rate": 6.057969086081575e-06,
+      "loss": 0.7267,
+      "step": 7018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5349667792437334,
+      "learning_rate": 6.0569560446364495e-06,
+      "loss": 0.6886,
+      "step": 7019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6830622378031777,
+      "learning_rate": 6.05594295777157e-06,
+      "loss": 0.8338,
+      "step": 7020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6877725011764193,
+      "learning_rate": 6.054929825530469e-06,
+      "loss": 0.686,
+      "step": 7021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.57606040025295,
+      "learning_rate": 6.0539166479566844e-06,
+      "loss": 0.7525,
+      "step": 7022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1212785913661032,
+      "learning_rate": 6.052903425093752e-06,
+      "loss": 0.5728,
+      "step": 7023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6692812758221727,
+      "learning_rate": 6.051890156985217e-06,
+      "loss": 0.7643,
+      "step": 7024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.58420189378665,
+      "learning_rate": 6.050876843674619e-06,
+      "loss": 0.7183,
+      "step": 7025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1918524120556957,
+      "learning_rate": 6.049863485205503e-06,
+      "loss": 0.8423,
+      "step": 7026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.889889504386999,
+      "learning_rate": 6.048850081621415e-06,
+      "loss": 0.7108,
+      "step": 7027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1567158415754264,
+      "learning_rate": 6.047836632965901e-06,
+      "loss": 0.7577,
+      "step": 7028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.589299091195435,
+      "learning_rate": 6.046823139282515e-06,
+      "loss": 0.7056,
+      "step": 7029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.850435855698079,
+      "learning_rate": 6.045809600614807e-06,
+      "loss": 0.7539,
+      "step": 7030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1947230940411664,
+      "learning_rate": 6.044796017006331e-06,
+      "loss": 0.692,
+      "step": 7031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0911864643504856,
+      "learning_rate": 6.043782388500642e-06,
+      "loss": 0.7322,
+      "step": 7032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4952420759901068,
+      "learning_rate": 6.0427687151413004e-06,
+      "loss": 0.5692,
+      "step": 7033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.535635510778326,
+      "learning_rate": 6.041754996971866e-06,
+      "loss": 0.7324,
+      "step": 7034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.543495972537071,
+      "learning_rate": 6.040741234035898e-06,
+      "loss": 0.6917,
+      "step": 7035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7943736379671487,
+      "learning_rate": 6.039727426376961e-06,
+      "loss": 0.6558,
+      "step": 7036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7722313170742392,
+      "learning_rate": 6.038713574038622e-06,
+      "loss": 0.87,
+      "step": 7037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1802765202749357,
+      "learning_rate": 6.037699677064446e-06,
+      "loss": 0.7412,
+      "step": 7038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7962570659796095,
+      "learning_rate": 6.036685735498004e-06,
+      "loss": 0.7219,
+      "step": 7039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.087406742870312,
+      "learning_rate": 6.035671749382869e-06,
+      "loss": 0.6472,
+      "step": 7040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6473198611924385,
+      "learning_rate": 6.03465771876261e-06,
+      "loss": 0.7058,
+      "step": 7041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6469182279077905,
+      "learning_rate": 6.0336436436808054e-06,
+      "loss": 0.6951,
+      "step": 7042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6887796092300793,
+      "learning_rate": 6.032629524181031e-06,
+      "loss": 0.8461,
+      "step": 7043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4646504930333117,
+      "learning_rate": 6.031615360306867e-06,
+      "loss": 0.714,
+      "step": 7044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3002989579456623,
+      "learning_rate": 6.030601152101892e-06,
+      "loss": 0.6787,
+      "step": 7045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7161575647408027,
+      "learning_rate": 6.029586899609692e-06,
+      "loss": 0.7105,
+      "step": 7046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9571973969017018,
+      "learning_rate": 6.02857260287385e-06,
+      "loss": 0.8849,
+      "step": 7047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8679539132423928,
+      "learning_rate": 6.02755826193795e-06,
+      "loss": 0.6705,
+      "step": 7048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.75524263088909,
+      "learning_rate": 6.026543876845586e-06,
+      "loss": 0.7134,
+      "step": 7049
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6176864500044061,
+      "learning_rate": 6.025529447640343e-06,
+      "loss": 0.653,
+      "step": 7050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5604670445611437,
+      "learning_rate": 6.024514974365818e-06,
+      "loss": 0.6822,
+      "step": 7051
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6494796947064987,
+      "learning_rate": 6.0235004570656005e-06,
+      "loss": 0.7818,
+      "step": 7052
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3794735946129153,
+      "learning_rate": 6.02248589578329e-06,
+      "loss": 0.6638,
+      "step": 7053
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7610055593387821,
+      "learning_rate": 6.021471290562484e-06,
+      "loss": 0.8645,
+      "step": 7054
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8532891155633162,
+      "learning_rate": 6.020456641446781e-06,
+      "loss": 0.837,
+      "step": 7055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6438747465272188,
+      "learning_rate": 6.019441948479784e-06,
+      "loss": 0.6649,
+      "step": 7056
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.185100295456144,
+      "learning_rate": 6.018427211705094e-06,
+      "loss": 0.7342,
+      "step": 7057
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5670208503175924,
+      "learning_rate": 6.017412431166322e-06,
+      "loss": 0.7968,
+      "step": 7058
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7343452647282518,
+      "learning_rate": 6.016397606907069e-06,
+      "loss": 0.6868,
+      "step": 7059
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0965739608214207,
+      "learning_rate": 6.015382738970949e-06,
+      "loss": 0.6765,
+      "step": 7060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5412732838257814,
+      "learning_rate": 6.01436782740157e-06,
+      "loss": 0.6752,
+      "step": 7061
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7782664364411662,
+      "learning_rate": 6.013352872242546e-06,
+      "loss": 0.7118,
+      "step": 7062
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.131771342847608,
+      "learning_rate": 6.012337873537494e-06,
+      "loss": 0.6731,
+      "step": 7063
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.643144293750578,
+      "learning_rate": 6.011322831330028e-06,
+      "loss": 0.5993,
+      "step": 7064
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6199728881741677,
+      "learning_rate": 6.010307745663768e-06,
+      "loss": 0.7634,
+      "step": 7065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5913213630415333,
+      "learning_rate": 6.009292616582333e-06,
+      "loss": 0.7492,
+      "step": 7066
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8220998540756284,
+      "learning_rate": 6.008277444129347e-06,
+      "loss": 0.8558,
+      "step": 7067
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4917060039994383,
+      "learning_rate": 6.007262228348434e-06,
+      "loss": 0.6986,
+      "step": 7068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3145935813090714,
+      "learning_rate": 6.0062469692832205e-06,
+      "loss": 0.7754,
+      "step": 7069
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.23087911178832,
+      "learning_rate": 6.005231666977331e-06,
+      "loss": 0.6202,
+      "step": 7070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5887203392829472,
+      "learning_rate": 6.0042163214744006e-06,
+      "loss": 0.7092,
+      "step": 7071
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7609283176522437,
+      "learning_rate": 6.003200932818058e-06,
+      "loss": 0.7724,
+      "step": 7072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6204352323449809,
+      "learning_rate": 6.002185501051937e-06,
+      "loss": 0.8611,
+      "step": 7073
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7359246481316992,
+      "learning_rate": 6.001170026219673e-06,
+      "loss": 0.6863,
+      "step": 7074
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.643771342418592,
+      "learning_rate": 6.000154508364901e-06,
+      "loss": 0.7702,
+      "step": 7075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9027811537348995,
+      "learning_rate": 5.999138947531265e-06,
+      "loss": 0.7744,
+      "step": 7076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5669321506881257,
+      "learning_rate": 5.998123343762403e-06,
+      "loss": 0.7095,
+      "step": 7077
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8081527376265625,
+      "learning_rate": 5.997107697101957e-06,
+      "loss": 0.7861,
+      "step": 7078
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7428950392830866,
+      "learning_rate": 5.996092007593572e-06,
+      "loss": 0.6558,
+      "step": 7079
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9574588222126246,
+      "learning_rate": 5.9950762752808965e-06,
+      "loss": 0.8277,
+      "step": 7080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1548079041256558,
+      "learning_rate": 5.994060500207577e-06,
+      "loss": 0.5523,
+      "step": 7081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7870695547646351,
+      "learning_rate": 5.993044682417264e-06,
+      "loss": 0.7544,
+      "step": 7082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4597839123223824,
+      "learning_rate": 5.992028821953609e-06,
+      "loss": 0.973,
+      "step": 7083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6748399176623248,
+      "learning_rate": 5.9910129188602665e-06,
+      "loss": 0.7093,
+      "step": 7084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8145501714983665,
+      "learning_rate": 5.989996973180892e-06,
+      "loss": 0.6955,
+      "step": 7085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4793598678624855,
+      "learning_rate": 5.988980984959142e-06,
+      "loss": 0.8275,
+      "step": 7086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.614316606137556,
+      "learning_rate": 5.987964954238677e-06,
+      "loss": 0.6444,
+      "step": 7087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8591008545626035,
+      "learning_rate": 5.986948881063157e-06,
+      "loss": 0.7507,
+      "step": 7088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8043529070920699,
+      "learning_rate": 5.985932765476246e-06,
+      "loss": 0.8316,
+      "step": 7089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6345845437342954,
+      "learning_rate": 5.984916607521608e-06,
+      "loss": 0.715,
+      "step": 7090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4217388726047662,
+      "learning_rate": 5.983900407242911e-06,
+      "loss": 0.6188,
+      "step": 7091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7658986500126745,
+      "learning_rate": 5.982884164683821e-06,
+      "loss": 0.6898,
+      "step": 7092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0733849943305334,
+      "learning_rate": 5.9818678798880085e-06,
+      "loss": 0.6635,
+      "step": 7093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8192184215055218,
+      "learning_rate": 5.9808515528991486e-06,
+      "loss": 0.7991,
+      "step": 7094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5903026762314365,
+      "learning_rate": 5.979835183760911e-06,
+      "loss": 0.7848,
+      "step": 7095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9178796505349933,
+      "learning_rate": 5.978818772516976e-06,
+      "loss": 0.7667,
+      "step": 7096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7091366430232575,
+      "learning_rate": 5.977802319211016e-06,
+      "loss": 0.7339,
+      "step": 7097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.731299925096254,
+      "learning_rate": 5.976785823886713e-06,
+      "loss": 0.7882,
+      "step": 7098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7330062448740093,
+      "learning_rate": 5.975769286587747e-06,
+      "loss": 0.7887,
+      "step": 7099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6281084443260547,
+      "learning_rate": 5.974752707357805e-06,
+      "loss": 0.6776,
+      "step": 7100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5753315908822958,
+      "learning_rate": 5.973736086240566e-06,
+      "loss": 0.7672,
+      "step": 7101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7799206541876376,
+      "learning_rate": 5.9727194232797185e-06,
+      "loss": 0.8876,
+      "step": 7102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6974859385270373,
+      "learning_rate": 5.97170271851895e-06,
+      "loss": 0.6974,
+      "step": 7103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.774784908749254,
+      "learning_rate": 5.970685972001953e-06,
+      "loss": 0.8954,
+      "step": 7104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4410001540329704,
+      "learning_rate": 5.96966918377242e-06,
+      "loss": 0.7368,
+      "step": 7105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.58605213510814,
+      "learning_rate": 5.9686523538740385e-06,
+      "loss": 0.7629,
+      "step": 7106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8068695623418276,
+      "learning_rate": 5.96763548235051e-06,
+      "loss": 0.6714,
+      "step": 7107
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8309408770615383,
+      "learning_rate": 5.966618569245529e-06,
+      "loss": 0.6356,
+      "step": 7108
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0655753990895547,
+      "learning_rate": 5.965601614602798e-06,
+      "loss": 0.6915,
+      "step": 7109
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7668770987391411,
+      "learning_rate": 5.964584618466014e-06,
+      "loss": 0.6982,
+      "step": 7110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7339739732226773,
+      "learning_rate": 5.9635675808788785e-06,
+      "loss": 0.7326,
+      "step": 7111
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8407890574835128,
+      "learning_rate": 5.9625505018851e-06,
+      "loss": 0.809,
+      "step": 7112
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6921685342007542,
+      "learning_rate": 5.961533381528382e-06,
+      "loss": 0.837,
+      "step": 7113
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7732245432146896,
+      "learning_rate": 5.960516219852433e-06,
+      "loss": 0.6939,
+      "step": 7114
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.001005840345451,
+      "learning_rate": 5.959499016900963e-06,
+      "loss": 0.851,
+      "step": 7115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6669825099530375,
+      "learning_rate": 5.9584817727176836e-06,
+      "loss": 0.6274,
+      "step": 7116
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7022931740396943,
+      "learning_rate": 5.9574644873463075e-06,
+      "loss": 0.7832,
+      "step": 7117
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6169188401489243,
+      "learning_rate": 5.95644716083055e-06,
+      "loss": 0.7163,
+      "step": 7118
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6067463900431966,
+      "learning_rate": 5.955429793214129e-06,
+      "loss": 0.7717,
+      "step": 7119
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7715140694351112,
+      "learning_rate": 5.954412384540761e-06,
+      "loss": 0.6992,
+      "step": 7120
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8353482151127656,
+      "learning_rate": 5.953394934854167e-06,
+      "loss": 0.792,
+      "step": 7121
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9740153643175127,
+      "learning_rate": 5.952377444198068e-06,
+      "loss": 0.8219,
+      "step": 7122
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9474541672372772,
+      "learning_rate": 5.951359912616192e-06,
+      "loss": 0.7406,
+      "step": 7123
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9483215257068869,
+      "learning_rate": 5.950342340152261e-06,
+      "loss": 0.7537,
+      "step": 7124
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6693413560205328,
+      "learning_rate": 5.949324726850002e-06,
+      "loss": 0.7559,
+      "step": 7125
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3993758338837081,
+      "learning_rate": 5.948307072753146e-06,
+      "loss": 0.5911,
+      "step": 7126
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0941096067567364,
+      "learning_rate": 5.947289377905422e-06,
+      "loss": 0.7519,
+      "step": 7127
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.611142175355835,
+      "learning_rate": 5.946271642350565e-06,
+      "loss": 0.7411,
+      "step": 7128
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7399591958723062,
+      "learning_rate": 5.945253866132308e-06,
+      "loss": 0.6821,
+      "step": 7129
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5881509385174315,
+      "learning_rate": 5.9442360492943876e-06,
+      "loss": 0.761,
+      "step": 7130
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.157809187853498,
+      "learning_rate": 5.94321819188054e-06,
+      "loss": 0.7746,
+      "step": 7131
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7143744931806684,
+      "learning_rate": 5.942200293934506e-06,
+      "loss": 0.8814,
+      "step": 7132
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7468969317680434,
+      "learning_rate": 5.941182355500028e-06,
+      "loss": 0.7536,
+      "step": 7133
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7778700569493753,
+      "learning_rate": 5.940164376620847e-06,
+      "loss": 0.8217,
+      "step": 7134
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9384840358518247,
+      "learning_rate": 5.93914635734071e-06,
+      "loss": 0.7319,
+      "step": 7135
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8258175662564258,
+      "learning_rate": 5.938128297703361e-06,
+      "loss": 0.7112,
+      "step": 7136
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9557957150983085,
+      "learning_rate": 5.937110197752551e-06,
+      "loss": 0.8371,
+      "step": 7137
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6861170868567887,
+      "learning_rate": 5.936092057532029e-06,
+      "loss": 0.6872,
+      "step": 7138
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7125021676435108,
+      "learning_rate": 5.935073877085546e-06,
+      "loss": 0.7141,
+      "step": 7139
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7791693642246003,
+      "learning_rate": 5.934055656456855e-06,
+      "loss": 0.765,
+      "step": 7140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.56592185000554,
+      "learning_rate": 5.9330373956897135e-06,
+      "loss": 0.7644,
+      "step": 7141
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.138033945060271,
+      "learning_rate": 5.932019094827877e-06,
+      "loss": 0.8913,
+      "step": 7142
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7139129511904785,
+      "learning_rate": 5.931000753915104e-06,
+      "loss": 0.6303,
+      "step": 7143
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5986190984026407,
+      "learning_rate": 5.9299823729951544e-06,
+      "loss": 0.7074,
+      "step": 7144
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7749478158062932,
+      "learning_rate": 5.928963952111791e-06,
+      "loss": 0.8122,
+      "step": 7145
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.669462301205644,
+      "learning_rate": 5.9279454913087795e-06,
+      "loss": 0.7571,
+      "step": 7146
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.564583989128945,
+      "learning_rate": 5.926926990629883e-06,
+      "loss": 0.7673,
+      "step": 7147
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.751102776468043,
+      "learning_rate": 5.9259084501188695e-06,
+      "loss": 0.8126,
+      "step": 7148
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6615774550619031,
+      "learning_rate": 5.9248898698195054e-06,
+      "loss": 0.6084,
+      "step": 7149
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9223234003735044,
+      "learning_rate": 5.923871249775566e-06,
+      "loss": 0.7714,
+      "step": 7150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8596946224590138,
+      "learning_rate": 5.922852590030821e-06,
+      "loss": 0.7617,
+      "step": 7151
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.751564910241884,
+      "learning_rate": 5.921833890629045e-06,
+      "loss": 0.7649,
+      "step": 7152
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8075788452732249,
+      "learning_rate": 5.920815151614012e-06,
+      "loss": 0.658,
+      "step": 7153
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.634021233628217,
+      "learning_rate": 5.919796373029504e-06,
+      "loss": 0.6737,
+      "step": 7154
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0775644644900484,
+      "learning_rate": 5.918777554919296e-06,
+      "loss": 0.6744,
+      "step": 7155
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6543618125811825,
+      "learning_rate": 5.917758697327171e-06,
+      "loss": 0.758,
+      "step": 7156
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6480397792142736,
+      "learning_rate": 5.9167398002969104e-06,
+      "loss": 0.7773,
+      "step": 7157
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8196273782919405,
+      "learning_rate": 5.915720863872297e-06,
+      "loss": 0.869,
+      "step": 7158
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2196876391910885,
+      "learning_rate": 5.914701888097121e-06,
+      "loss": 0.7662,
+      "step": 7159
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5703454335174787,
+      "learning_rate": 5.913682873015166e-06,
+      "loss": 0.7679,
+      "step": 7160
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.076330543146294,
+      "learning_rate": 5.912663818670224e-06,
+      "loss": 0.9292,
+      "step": 7161
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1845105631234543,
+      "learning_rate": 5.911644725106085e-06,
+      "loss": 0.6347,
+      "step": 7162
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.611798029454967,
+      "learning_rate": 5.910625592366542e-06,
+      "loss": 0.7612,
+      "step": 7163
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0502538158425239,
+      "learning_rate": 5.90960642049539e-06,
+      "loss": 0.587,
+      "step": 7164
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5091010972137298,
+      "learning_rate": 5.9085872095364236e-06,
+      "loss": 0.7827,
+      "step": 7165
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7394176802434282,
+      "learning_rate": 5.907567959533441e-06,
+      "loss": 0.6716,
+      "step": 7166
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9263249435509002,
+      "learning_rate": 5.906548670530241e-06,
+      "loss": 0.7093,
+      "step": 7167
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.590617233585951,
+      "learning_rate": 5.905529342570627e-06,
+      "loss": 0.7863,
+      "step": 7168
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.314292542066483,
+      "learning_rate": 5.904509975698399e-06,
+      "loss": 0.6286,
+      "step": 7169
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7752512789915142,
+      "learning_rate": 5.903490569957366e-06,
+      "loss": 0.8059,
+      "step": 7170
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6776600899871201,
+      "learning_rate": 5.902471125391328e-06,
+      "loss": 0.6823,
+      "step": 7171
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.778397047042455,
+      "learning_rate": 5.901451642044097e-06,
+      "loss": 0.7731,
+      "step": 7172
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0751699906560122,
+      "learning_rate": 5.90043211995948e-06,
+      "loss": 0.745,
+      "step": 7173
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3800875955037768,
+      "learning_rate": 5.8994125591812914e-06,
+      "loss": 0.7814,
+      "step": 7174
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.180347491611148,
+      "learning_rate": 5.898392959753343e-06,
+      "loss": 0.7337,
+      "step": 7175
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2278636374285827,
+      "learning_rate": 5.897373321719445e-06,
+      "loss": 0.5708,
+      "step": 7176
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1611900253118834,
+      "learning_rate": 5.896353645123418e-06,
+      "loss": 0.7437,
+      "step": 7177
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1422832582710063,
+      "learning_rate": 5.895333930009079e-06,
+      "loss": 0.6989,
+      "step": 7178
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0393213468070863,
+      "learning_rate": 5.894314176420247e-06,
+      "loss": 0.6826,
+      "step": 7179
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8892212925114944,
+      "learning_rate": 5.893294384400743e-06,
+      "loss": 0.6383,
+      "step": 7180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6123524796591875,
+      "learning_rate": 5.89227455399439e-06,
+      "loss": 0.7527,
+      "step": 7181
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.831268326673753,
+      "learning_rate": 5.8912546852450116e-06,
+      "loss": 0.7028,
+      "step": 7182
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.688909143226616,
+      "learning_rate": 5.890234778196435e-06,
+      "loss": 0.7475,
+      "step": 7183
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6939973226289025,
+      "learning_rate": 5.889214832892489e-06,
+      "loss": 0.8381,
+      "step": 7184
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.165926706332095,
+      "learning_rate": 5.888194849376998e-06,
+      "loss": 0.7452,
+      "step": 7185
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5747694309586269,
+      "learning_rate": 5.8871748276937975e-06,
+      "loss": 0.6965,
+      "step": 7186
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6451895440797166,
+      "learning_rate": 5.886154767886719e-06,
+      "loss": 0.6628,
+      "step": 7187
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6874667340962082,
+      "learning_rate": 5.885134669999597e-06,
+      "loss": 0.8055,
+      "step": 7188
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7358941616865002,
+      "learning_rate": 5.8841145340762665e-06,
+      "loss": 0.796,
+      "step": 7189
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0840474384198533,
+      "learning_rate": 5.8830943601605665e-06,
+      "loss": 0.6913,
+      "step": 7190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1796310768911855,
+      "learning_rate": 5.882074148296332e-06,
+      "loss": 0.8211,
+      "step": 7191
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9090072101611593,
+      "learning_rate": 5.881053898527409e-06,
+      "loss": 0.791,
+      "step": 7192
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5458639453306267,
+      "learning_rate": 5.880033610897638e-06,
+      "loss": 0.6332,
+      "step": 7193
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7122716323667622,
+      "learning_rate": 5.879013285450863e-06,
+      "loss": 0.6896,
+      "step": 7194
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.983684966898967,
+      "learning_rate": 5.877992922230928e-06,
+      "loss": 0.8124,
+      "step": 7195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9315342682247323,
+      "learning_rate": 5.876972521281683e-06,
+      "loss": 0.7916,
+      "step": 7196
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4848885199980852,
+      "learning_rate": 5.875952082646975e-06,
+      "loss": 0.7266,
+      "step": 7197
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8248678396611198,
+      "learning_rate": 5.8749316063706565e-06,
+      "loss": 0.717,
+      "step": 7198
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6555381311790731,
+      "learning_rate": 5.873911092496577e-06,
+      "loss": 0.6268,
+      "step": 7199
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7080108793757693,
+      "learning_rate": 5.872890541068591e-06,
+      "loss": 0.7456,
+      "step": 7200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7653714308190358,
+      "learning_rate": 5.871869952130556e-06,
+      "loss": 0.7018,
+      "step": 7201
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.889154122717826,
+      "learning_rate": 5.870849325726326e-06,
+      "loss": 0.6626,
+      "step": 7202
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7081186264493322,
+      "learning_rate": 5.869828661899761e-06,
+      "loss": 0.7291,
+      "step": 7203
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.289611595091256,
+      "learning_rate": 5.8688079606947226e-06,
+      "loss": 0.7059,
+      "step": 7204
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8399423766324654,
+      "learning_rate": 5.86778722215507e-06,
+      "loss": 0.8133,
+      "step": 7205
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5967441729025025,
+      "learning_rate": 5.866766446324668e-06,
+      "loss": 0.7526,
+      "step": 7206
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5684463271942153,
+      "learning_rate": 5.865745633247381e-06,
+      "loss": 0.7309,
+      "step": 7207
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.221940464895738,
+      "learning_rate": 5.864724782967077e-06,
+      "loss": 0.6351,
+      "step": 7208
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.77220157819897,
+      "learning_rate": 5.8637038955276225e-06,
+      "loss": 0.7925,
+      "step": 7209
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.194980326606325,
+      "learning_rate": 5.862682970972888e-06,
+      "loss": 0.7504,
+      "step": 7210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6460153925119323,
+      "learning_rate": 5.861662009346747e-06,
+      "loss": 0.7637,
+      "step": 7211
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7145469428291813,
+      "learning_rate": 5.8606410106930675e-06,
+      "loss": 0.6815,
+      "step": 7212
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7028844988363225,
+      "learning_rate": 5.85961997505573e-06,
+      "loss": 0.6623,
+      "step": 7213
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5665461602877984,
+      "learning_rate": 5.858598902478604e-06,
+      "loss": 0.6932,
+      "step": 7214
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.511438169640806,
+      "learning_rate": 5.857577793005573e-06,
+      "loss": 0.8214,
+      "step": 7215
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.54989613996416,
+      "learning_rate": 5.856556646680514e-06,
+      "loss": 0.6988,
+      "step": 7216
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8451822880863407,
+      "learning_rate": 5.855535463547309e-06,
+      "loss": 0.8121,
+      "step": 7217
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6882306079318297,
+      "learning_rate": 5.854514243649839e-06,
+      "loss": 0.8263,
+      "step": 7218
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1387854380178069,
+      "learning_rate": 5.853492987031989e-06,
+      "loss": 0.5867,
+      "step": 7219
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8595920646067363,
+      "learning_rate": 5.8524716937376446e-06,
+      "loss": 0.7122,
+      "step": 7220
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6661555213167807,
+      "learning_rate": 5.851450363810694e-06,
+      "loss": 0.7745,
+      "step": 7221
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7046834194262523,
+      "learning_rate": 5.850428997295023e-06,
+      "loss": 0.7524,
+      "step": 7222
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.035138941234602,
+      "learning_rate": 5.8494075942345244e-06,
+      "loss": 0.7921,
+      "step": 7223
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8504667110494075,
+      "learning_rate": 5.8483861546730915e-06,
+      "loss": 0.9102,
+      "step": 7224
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1669891499713012,
+      "learning_rate": 5.8473646786546134e-06,
+      "loss": 0.6396,
+      "step": 7225
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.506007190001064,
+      "learning_rate": 5.846343166222991e-06,
+      "loss": 0.7626,
+      "step": 7226
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7384814195046772,
+      "learning_rate": 5.845321617422115e-06,
+      "loss": 0.6316,
+      "step": 7227
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6640651366533772,
+      "learning_rate": 5.8443000322958875e-06,
+      "loss": 0.754,
+      "step": 7228
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7798857056036885,
+      "learning_rate": 5.843278410888208e-06,
+      "loss": 0.8163,
+      "step": 7229
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7025485994234801,
+      "learning_rate": 5.842256753242975e-06,
+      "loss": 0.7308,
+      "step": 7230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6043417668921143,
+      "learning_rate": 5.841235059404097e-06,
+      "loss": 0.7601,
+      "step": 7231
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4348842195713383,
+      "learning_rate": 5.840213329415471e-06,
+      "loss": 0.6825,
+      "step": 7232
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8436100310397856,
+      "learning_rate": 5.839191563321009e-06,
+      "loss": 0.7456,
+      "step": 7233
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3138610942904814,
+      "learning_rate": 5.838169761164616e-06,
+      "loss": 0.6765,
+      "step": 7234
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6584105503315738,
+      "learning_rate": 5.837147922990204e-06,
+      "loss": 0.8621,
+      "step": 7235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1651083740159813,
+      "learning_rate": 5.836126048841678e-06,
+      "loss": 0.8166,
+      "step": 7236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.650114103761055,
+      "learning_rate": 5.8351041387629546e-06,
+      "loss": 0.7233,
+      "step": 7237
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.650098149949854,
+      "learning_rate": 5.834082192797948e-06,
+      "loss": 0.7291,
+      "step": 7238
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9565320423211746,
+      "learning_rate": 5.83306021099057e-06,
+      "loss": 0.5288,
+      "step": 7239
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6895540435751213,
+      "learning_rate": 5.832038193384741e-06,
+      "loss": 0.7956,
+      "step": 7240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6574188487627834,
+      "learning_rate": 5.831016140024377e-06,
+      "loss": 0.7394,
+      "step": 7241
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.085485515698926,
+      "learning_rate": 5.829994050953398e-06,
+      "loss": 0.7553,
+      "step": 7242
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6816067880885894,
+      "learning_rate": 5.828971926215726e-06,
+      "loss": 0.7352,
+      "step": 7243
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6786154579544554,
+      "learning_rate": 5.827949765855285e-06,
+      "loss": 0.7748,
+      "step": 7244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.267591138294332,
+      "learning_rate": 5.826927569915999e-06,
+      "loss": 0.7827,
+      "step": 7245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.808936093470966,
+      "learning_rate": 5.825905338441794e-06,
+      "loss": 0.777,
+      "step": 7246
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.042333621686125,
+      "learning_rate": 5.824883071476597e-06,
+      "loss": 0.7164,
+      "step": 7247
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2195084602203234,
+      "learning_rate": 5.823860769064335e-06,
+      "loss": 0.6201,
+      "step": 7248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.736923861058753,
+      "learning_rate": 5.822838431248943e-06,
+      "loss": 0.6686,
+      "step": 7249
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7936803796002656,
+      "learning_rate": 5.821816058074351e-06,
+      "loss": 0.7951,
+      "step": 7250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6677839981959948,
+      "learning_rate": 5.820793649584493e-06,
+      "loss": 0.6513,
+      "step": 7251
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6108793335237435,
+      "learning_rate": 5.819771205823303e-06,
+      "loss": 0.6776,
+      "step": 7252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.745689782845322,
+      "learning_rate": 5.818748726834718e-06,
+      "loss": 0.7542,
+      "step": 7253
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.04905746349162,
+      "learning_rate": 5.817726212662678e-06,
+      "loss": 0.604,
+      "step": 7254
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4032498202227435,
+      "learning_rate": 5.8167036633511206e-06,
+      "loss": 0.7741,
+      "step": 7255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6358779529410465,
+      "learning_rate": 5.815681078943989e-06,
+      "loss": 0.6784,
+      "step": 7256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4379396229657961,
+      "learning_rate": 5.814658459485223e-06,
+      "loss": 0.7087,
+      "step": 7257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5165712081327818,
+      "learning_rate": 5.81363580501877e-06,
+      "loss": 0.7343,
+      "step": 7258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7551836977490574,
+      "learning_rate": 5.812613115588575e-06,
+      "loss": 0.7864,
+      "step": 7259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5567143002233994,
+      "learning_rate": 5.811590391238584e-06,
+      "loss": 0.6831,
+      "step": 7260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5292823915950013,
+      "learning_rate": 5.8105676320127455e-06,
+      "loss": 0.7002,
+      "step": 7261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.156035613332329,
+      "learning_rate": 5.809544837955011e-06,
+      "loss": 0.7401,
+      "step": 7262
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.541248258813399,
+      "learning_rate": 5.808522009109333e-06,
+      "loss": 0.6775,
+      "step": 7263
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7453501042906698,
+      "learning_rate": 5.807499145519663e-06,
+      "loss": 0.7665,
+      "step": 7264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9947613484826785,
+      "learning_rate": 5.806476247229956e-06,
+      "loss": 0.7695,
+      "step": 7265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8968510760606911,
+      "learning_rate": 5.805453314284168e-06,
+      "loss": 0.7262,
+      "step": 7266
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6515291353488417,
+      "learning_rate": 5.804430346726258e-06,
+      "loss": 0.7286,
+      "step": 7267
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7595792429581576,
+      "learning_rate": 5.803407344600185e-06,
+      "loss": 0.6799,
+      "step": 7268
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5571158668930678,
+      "learning_rate": 5.802384307949909e-06,
+      "loss": 0.7144,
+      "step": 7269
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7218232286000021,
+      "learning_rate": 5.8013612368193905e-06,
+      "loss": 0.7664,
+      "step": 7270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.057266842216034,
+      "learning_rate": 5.800338131252597e-06,
+      "loss": 0.8122,
+      "step": 7271
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8476456548758147,
+      "learning_rate": 5.79931499129349e-06,
+      "loss": 0.7639,
+      "step": 7272
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.725037590067836,
+      "learning_rate": 5.7982918169860395e-06,
+      "loss": 0.6982,
+      "step": 7273
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8858059030834144,
+      "learning_rate": 5.79726860837421e-06,
+      "loss": 0.8068,
+      "step": 7274
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.687186440508739,
+      "learning_rate": 5.796245365501973e-06,
+      "loss": 0.8343,
+      "step": 7275
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.388765463001988,
+      "learning_rate": 5.795222088413299e-06,
+      "loss": 0.7898,
+      "step": 7276
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.54324322410973,
+      "learning_rate": 5.794198777152162e-06,
+      "loss": 0.6832,
+      "step": 7277
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.460747931685601,
+      "learning_rate": 5.793175431762534e-06,
+      "loss": 0.7471,
+      "step": 7278
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.736749921378703,
+      "learning_rate": 5.792152052288391e-06,
+      "loss": 0.7631,
+      "step": 7279
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4454506738007837,
+      "learning_rate": 5.791128638773711e-06,
+      "loss": 0.7237,
+      "step": 7280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6295371586439442,
+      "learning_rate": 5.790105191262471e-06,
+      "loss": 0.7844,
+      "step": 7281
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.032136316043192,
+      "learning_rate": 5.789081709798652e-06,
+      "loss": 0.7322,
+      "step": 7282
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.31294261814576,
+      "learning_rate": 5.788058194426234e-06,
+      "loss": 0.7011,
+      "step": 7283
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.031093469748077,
+      "learning_rate": 5.787034645189199e-06,
+      "loss": 0.6873,
+      "step": 7284
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6405693688146197,
+      "learning_rate": 5.786011062131535e-06,
+      "loss": 0.6268,
+      "step": 7285
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1790492528198186,
+      "learning_rate": 5.784987445297222e-06,
+      "loss": 0.7072,
+      "step": 7286
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8874622327251336,
+      "learning_rate": 5.783963794730254e-06,
+      "loss": 0.7373,
+      "step": 7287
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.041645307422203,
+      "learning_rate": 5.782940110474612e-06,
+      "loss": 0.4937,
+      "step": 7288
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.058389921062226,
+      "learning_rate": 5.7819163925742915e-06,
+      "loss": 0.62,
+      "step": 7289
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7124063625498342,
+      "learning_rate": 5.780892641073281e-06,
+      "loss": 0.7458,
+      "step": 7290
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7209015772473286,
+      "learning_rate": 5.779868856015578e-06,
+      "loss": 0.7387,
+      "step": 7291
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5947442796556501,
+      "learning_rate": 5.778845037445171e-06,
+      "loss": 0.5544,
+      "step": 7292
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4894773462992363,
+      "learning_rate": 5.7778211854060575e-06,
+      "loss": 0.6709,
+      "step": 7293
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7118265556285328,
+      "learning_rate": 5.776797299942236e-06,
+      "loss": 0.7914,
+      "step": 7294
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4176229399629623,
+      "learning_rate": 5.775773381097705e-06,
+      "loss": 0.7691,
+      "step": 7295
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6639664059417703,
+      "learning_rate": 5.774749428916464e-06,
+      "loss": 0.7213,
+      "step": 7296
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.929185074555467,
+      "learning_rate": 5.773725443442514e-06,
+      "loss": 0.7274,
+      "step": 7297
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0993171707183131,
+      "learning_rate": 5.7727014247198585e-06,
+      "loss": 0.7407,
+      "step": 7298
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7037349802013602,
+      "learning_rate": 5.771677372792502e-06,
+      "loss": 0.6688,
+      "step": 7299
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7508857366124202,
+      "learning_rate": 5.770653287704451e-06,
+      "loss": 0.7877,
+      "step": 7300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8743175017676652,
+      "learning_rate": 5.7696291694997105e-06,
+      "loss": 0.8017,
+      "step": 7301
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7040620197307357,
+      "learning_rate": 5.768605018222292e-06,
+      "loss": 0.7525,
+      "step": 7302
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.888919983484522,
+      "learning_rate": 5.767580833916203e-06,
+      "loss": 0.7342,
+      "step": 7303
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2925150511457058,
+      "learning_rate": 5.766556616625456e-06,
+      "loss": 0.778,
+      "step": 7304
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1105266884056288,
+      "learning_rate": 5.765532366394064e-06,
+      "loss": 0.6124,
+      "step": 7305
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8295661785178468,
+      "learning_rate": 5.764508083266043e-06,
+      "loss": 0.8599,
+      "step": 7306
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7839855404531328,
+      "learning_rate": 5.763483767285405e-06,
+      "loss": 0.6756,
+      "step": 7307
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.709020735112219,
+      "learning_rate": 5.762459418496169e-06,
+      "loss": 0.7548,
+      "step": 7308
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.376848603311801,
+      "learning_rate": 5.7614350369423555e-06,
+      "loss": 0.7061,
+      "step": 7309
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5753959038551975,
+      "learning_rate": 5.760410622667981e-06,
+      "loss": 0.7306,
+      "step": 7310
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.672206352615218,
+      "learning_rate": 5.75938617571707e-06,
+      "loss": 0.7261,
+      "step": 7311
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3317049406403663,
+      "learning_rate": 5.758361696133643e-06,
+      "loss": 0.808,
+      "step": 7312
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8181312048422218,
+      "learning_rate": 5.757337183961724e-06,
+      "loss": 0.769,
+      "step": 7313
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5129014653941077,
+      "learning_rate": 5.7563126392453415e-06,
+      "loss": 0.7351,
+      "step": 7314
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7366123512769263,
+      "learning_rate": 5.755288062028519e-06,
+      "loss": 0.7482,
+      "step": 7315
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7380639567037046,
+      "learning_rate": 5.754263452355287e-06,
+      "loss": 0.6763,
+      "step": 7316
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1880786907674963,
+      "learning_rate": 5.753238810269674e-06,
+      "loss": 0.6821,
+      "step": 7317
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.625116927909006,
+      "learning_rate": 5.752214135815712e-06,
+      "loss": 0.7787,
+      "step": 7318
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.583394200573293,
+      "learning_rate": 5.751189429037435e-06,
+      "loss": 0.7325,
+      "step": 7319
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0020999633140657,
+      "learning_rate": 5.750164689978873e-06,
+      "loss": 0.697,
+      "step": 7320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6914623037953567,
+      "learning_rate": 5.7491399186840646e-06,
+      "loss": 0.821,
+      "step": 7321
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7252898445253317,
+      "learning_rate": 5.748115115197045e-06,
+      "loss": 0.7045,
+      "step": 7322
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6388168331997486,
+      "learning_rate": 5.7470902795618535e-06,
+      "loss": 0.7474,
+      "step": 7323
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.600875558404842,
+      "learning_rate": 5.746065411822528e-06,
+      "loss": 0.775,
+      "step": 7324
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8778497265446137,
+      "learning_rate": 5.745040512023111e-06,
+      "loss": 0.6187,
+      "step": 7325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4560604201180172,
+      "learning_rate": 5.744015580207642e-06,
+      "loss": 0.6327,
+      "step": 7326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.064730859580756,
+      "learning_rate": 5.742990616420169e-06,
+      "loss": 0.7798,
+      "step": 7327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.742670856210051,
+      "learning_rate": 5.741965620704734e-06,
+      "loss": 0.7465,
+      "step": 7328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5534945372420068,
+      "learning_rate": 5.740940593105383e-06,
+      "loss": 0.639,
+      "step": 7329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8727830157924217,
+      "learning_rate": 5.7399155336661645e-06,
+      "loss": 0.7645,
+      "step": 7330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8271740668618743,
+      "learning_rate": 5.738890442431128e-06,
+      "loss": 0.7113,
+      "step": 7331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6510370149616784,
+      "learning_rate": 5.737865319444324e-06,
+      "loss": 0.7963,
+      "step": 7332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4789349020931777,
+      "learning_rate": 5.736840164749803e-06,
+      "loss": 0.657,
+      "step": 7333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5935984883227081,
+      "learning_rate": 5.73581497839162e-06,
+      "loss": 0.6307,
+      "step": 7334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6300942455928704,
+      "learning_rate": 5.734789760413827e-06,
+      "loss": 0.7855,
+      "step": 7335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7163893445487723,
+      "learning_rate": 5.733764510860482e-06,
+      "loss": 0.5738,
+      "step": 7336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6826763967565919,
+      "learning_rate": 5.7327392297756426e-06,
+      "loss": 0.7942,
+      "step": 7337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5658431255036533,
+      "learning_rate": 5.731713917203368e-06,
+      "loss": 0.6402,
+      "step": 7338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1691428467779628,
+      "learning_rate": 5.730688573187715e-06,
+      "loss": 0.6364,
+      "step": 7339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.834807651181348,
+      "learning_rate": 5.729663197772745e-06,
+      "loss": 0.7863,
+      "step": 7340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8037545659802838,
+      "learning_rate": 5.728637791002525e-06,
+      "loss": 0.7924,
+      "step": 7341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7129018397191729,
+      "learning_rate": 5.727612352921116e-06,
+      "loss": 0.7076,
+      "step": 7342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.950283474928144,
+      "learning_rate": 5.726586883572584e-06,
+      "loss": 0.6575,
+      "step": 7343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.743150400360149,
+      "learning_rate": 5.725561383000994e-06,
+      "loss": 0.7383,
+      "step": 7344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6006292661084787,
+      "learning_rate": 5.724535851250416e-06,
+      "loss": 0.784,
+      "step": 7345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6363489753291056,
+      "learning_rate": 5.723510288364918e-06,
+      "loss": 0.6872,
+      "step": 7346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7350274559273042,
+      "learning_rate": 5.722484694388573e-06,
+      "loss": 0.6437,
+      "step": 7347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4489472584727683,
+      "learning_rate": 5.721459069365452e-06,
+      "loss": 0.6161,
+      "step": 7348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8330351429929015,
+      "learning_rate": 5.720433413339627e-06,
+      "loss": 0.7325,
+      "step": 7349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6942479314168555,
+      "learning_rate": 5.719407726355174e-06,
+      "loss": 0.7485,
+      "step": 7350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1949393863585072,
+      "learning_rate": 5.718382008456168e-06,
+      "loss": 0.6534,
+      "step": 7351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7286357225586457,
+      "learning_rate": 5.717356259686689e-06,
+      "loss": 0.7278,
+      "step": 7352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5626147154497922,
+      "learning_rate": 5.716330480090812e-06,
+      "loss": 0.7637,
+      "step": 7353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1635926056472743,
+      "learning_rate": 5.71530466971262e-06,
+      "loss": 0.6708,
+      "step": 7354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7353402666549598,
+      "learning_rate": 5.714278828596192e-06,
+      "loss": 0.7765,
+      "step": 7355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7431502530212704,
+      "learning_rate": 5.713252956785613e-06,
+      "loss": 0.8086,
+      "step": 7356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6533178216869633,
+      "learning_rate": 5.712227054324968e-06,
+      "loss": 0.6683,
+      "step": 7357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1215658643601818,
+      "learning_rate": 5.711201121258337e-06,
+      "loss": 0.6281,
+      "step": 7358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4781379453889296,
+      "learning_rate": 5.710175157629812e-06,
+      "loss": 0.6923,
+      "step": 7359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.946481071935962,
+      "learning_rate": 5.709149163483476e-06,
+      "loss": 0.7418,
+      "step": 7360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5015093949398663,
+      "learning_rate": 5.708123138863424e-06,
+      "loss": 0.6169,
+      "step": 7361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8669639163495566,
+      "learning_rate": 5.707097083813743e-06,
+      "loss": 0.7243,
+      "step": 7362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7741102854585806,
+      "learning_rate": 5.706070998378524e-06,
+      "loss": 0.8545,
+      "step": 7363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1219807034398714,
+      "learning_rate": 5.705044882601862e-06,
+      "loss": 0.6269,
+      "step": 7364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7957461861770248,
+      "learning_rate": 5.704018736527853e-06,
+      "loss": 0.8522,
+      "step": 7365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7026944906877153,
+      "learning_rate": 5.7029925602005905e-06,
+      "loss": 0.6739,
+      "step": 7366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8035687083766008,
+      "learning_rate": 5.701966353664171e-06,
+      "loss": 0.8309,
+      "step": 7367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.567202677947694,
+      "learning_rate": 5.7009401169626955e-06,
+      "loss": 0.7814,
+      "step": 7368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.234624444678863,
+      "learning_rate": 5.69991385014026e-06,
+      "loss": 0.758,
+      "step": 7369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.645474469124295,
+      "learning_rate": 5.698887553240968e-06,
+      "loss": 0.7178,
+      "step": 7370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.519350319917618,
+      "learning_rate": 5.697861226308923e-06,
+      "loss": 0.7898,
+      "step": 7371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7633869630441674,
+      "learning_rate": 5.696834869388226e-06,
+      "loss": 0.7391,
+      "step": 7372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.685443423692735,
+      "learning_rate": 5.695808482522982e-06,
+      "loss": 0.7447,
+      "step": 7373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6114660169260218,
+      "learning_rate": 5.694782065757298e-06,
+      "loss": 0.6981,
+      "step": 7374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7320734754777884,
+      "learning_rate": 5.693755619135282e-06,
+      "loss": 0.812,
+      "step": 7375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7490790031595784,
+      "learning_rate": 5.692729142701042e-06,
+      "loss": 0.6522,
+      "step": 7376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7973106873563067,
+      "learning_rate": 5.691702636498688e-06,
+      "loss": 0.7888,
+      "step": 7377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.777532091916344,
+      "learning_rate": 5.69067610057233e-06,
+      "loss": 0.7925,
+      "step": 7378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5107653589276357,
+      "learning_rate": 5.689649534966083e-06,
+      "loss": 0.6908,
+      "step": 7379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7158939007570018,
+      "learning_rate": 5.6886229397240585e-06,
+      "loss": 0.8137,
+      "step": 7380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0219841715321776,
+      "learning_rate": 5.6875963148903735e-06,
+      "loss": 0.6799,
+      "step": 7381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1616792011487225,
+      "learning_rate": 5.6865696605091426e-06,
+      "loss": 0.6856,
+      "step": 7382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.792316689266636,
+      "learning_rate": 5.685542976624485e-06,
+      "loss": 0.7588,
+      "step": 7383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0915110068907157,
+      "learning_rate": 5.684516263280519e-06,
+      "loss": 0.7629,
+      "step": 7384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.571207497766288,
+      "learning_rate": 5.683489520521365e-06,
+      "loss": 0.7077,
+      "step": 7385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3360891545980518,
+      "learning_rate": 5.682462748391144e-06,
+      "loss": 0.6705,
+      "step": 7386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1584938252594545,
+      "learning_rate": 5.681435946933978e-06,
+      "loss": 0.7424,
+      "step": 7387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7066939895549247,
+      "learning_rate": 5.680409116193991e-06,
+      "loss": 0.7021,
+      "step": 7388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0296493410766527,
+      "learning_rate": 5.679382256215311e-06,
+      "loss": 0.6131,
+      "step": 7389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8288417086453383,
+      "learning_rate": 5.678355367042061e-06,
+      "loss": 0.8666,
+      "step": 7390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5551114361769585,
+      "learning_rate": 5.67732844871837e-06,
+      "loss": 0.6767,
+      "step": 7391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5916181318442604,
+      "learning_rate": 5.6763015012883686e-06,
+      "loss": 0.6609,
+      "step": 7392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6659057632861634,
+      "learning_rate": 5.675274524796184e-06,
+      "loss": 0.8411,
+      "step": 7393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5999896945743506,
+      "learning_rate": 5.674247519285951e-06,
+      "loss": 0.7312,
+      "step": 7394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0959666481033945,
+      "learning_rate": 5.673220484801798e-06,
+      "loss": 0.8024,
+      "step": 7395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7085824063516089,
+      "learning_rate": 5.672193421387862e-06,
+      "loss": 0.6968,
+      "step": 7396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6295858544337227,
+      "learning_rate": 5.671166329088278e-06,
+      "loss": 0.7349,
+      "step": 7397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.561872295300854,
+      "learning_rate": 5.6701392079471825e-06,
+      "loss": 0.7065,
+      "step": 7398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5608042965507924,
+      "learning_rate": 5.6691120580087126e-06,
+      "loss": 0.7062,
+      "step": 7399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4955994134907449,
+      "learning_rate": 5.668084879317006e-06,
+      "loss": 0.6742,
+      "step": 7400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4437900876989618,
+      "learning_rate": 5.6670576719162065e-06,
+      "loss": 0.6355,
+      "step": 7401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.960752185484751,
+      "learning_rate": 5.6660304358504524e-06,
+      "loss": 0.8037,
+      "step": 7402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7079889541089477,
+      "learning_rate": 5.665003171163886e-06,
+      "loss": 0.6411,
+      "step": 7403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7559530990986383,
+      "learning_rate": 5.6639758779006535e-06,
+      "loss": 0.7499,
+      "step": 7404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5335660685932972,
+      "learning_rate": 5.662948556104898e-06,
+      "loss": 0.7856,
+      "step": 7405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7445821338382945,
+      "learning_rate": 5.661921205820767e-06,
+      "loss": 0.6908,
+      "step": 7406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8909212639138298,
+      "learning_rate": 5.660893827092407e-06,
+      "loss": 0.7282,
+      "step": 7407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2503265063154212,
+      "learning_rate": 5.659866419963969e-06,
+      "loss": 0.6363,
+      "step": 7408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5273824427465834,
+      "learning_rate": 5.6588389844796e-06,
+      "loss": 0.6592,
+      "step": 7409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8484149630403135,
+      "learning_rate": 5.657811520683454e-06,
+      "loss": 0.7036,
+      "step": 7410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6521745428473682,
+      "learning_rate": 5.656784028619681e-06,
+      "loss": 0.6474,
+      "step": 7411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6305450299626847,
+      "learning_rate": 5.655756508332434e-06,
+      "loss": 0.8397,
+      "step": 7412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.774507907380422,
+      "learning_rate": 5.654728959865872e-06,
+      "loss": 0.6209,
+      "step": 7413
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4582006251787858,
+      "learning_rate": 5.653701383264147e-06,
+      "loss": 0.7772,
+      "step": 7414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.781468908927668,
+      "learning_rate": 5.652673778571418e-06,
+      "loss": 0.7249,
+      "step": 7415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6632800878251153,
+      "learning_rate": 5.651646145831842e-06,
+      "loss": 0.8106,
+      "step": 7416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2153637119954905,
+      "learning_rate": 5.650618485089582e-06,
+      "loss": 0.8469,
+      "step": 7417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2547283303977204,
+      "learning_rate": 5.649590796388794e-06,
+      "loss": 0.6218,
+      "step": 7418
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0498721539646705,
+      "learning_rate": 5.648563079773646e-06,
+      "loss": 0.8384,
+      "step": 7419
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.686247256426362,
+      "learning_rate": 5.647535335288296e-06,
+      "loss": 0.6734,
+      "step": 7420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.760735383437705,
+      "learning_rate": 5.6465075629769095e-06,
+      "loss": 0.794,
+      "step": 7421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8468793936177328,
+      "learning_rate": 5.645479762883657e-06,
+      "loss": 0.7795,
+      "step": 7422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.765893151528746,
+      "learning_rate": 5.644451935052697e-06,
+      "loss": 0.8033,
+      "step": 7423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2568935953872185,
+      "learning_rate": 5.6434240795282045e-06,
+      "loss": 0.7605,
+      "step": 7424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8237322942212313,
+      "learning_rate": 5.642396196354345e-06,
+      "loss": 0.7969,
+      "step": 7425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2461337232505167,
+      "learning_rate": 5.641368285575292e-06,
+      "loss": 0.6956,
+      "step": 7426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7895685000996264,
+      "learning_rate": 5.640340347235215e-06,
+      "loss": 0.7,
+      "step": 7427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7328157236699033,
+      "learning_rate": 5.639312381378286e-06,
+      "loss": 0.6845,
+      "step": 7428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5215285166563566,
+      "learning_rate": 5.63828438804868e-06,
+      "loss": 0.6985,
+      "step": 7429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6851993946061206,
+      "learning_rate": 5.637256367290573e-06,
+      "loss": 0.7181,
+      "step": 7430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4831718572497468,
+      "learning_rate": 5.636228319148141e-06,
+      "loss": 0.786,
+      "step": 7431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7916680906287328,
+      "learning_rate": 5.635200243665562e-06,
+      "loss": 0.8751,
+      "step": 7432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.73324422858378,
+      "learning_rate": 5.634172140887013e-06,
+      "loss": 0.7093,
+      "step": 7433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7676965876218724,
+      "learning_rate": 5.6331440108566735e-06,
+      "loss": 0.7699,
+      "step": 7434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6574536795424788,
+      "learning_rate": 5.632115853618727e-06,
+      "loss": 0.8,
+      "step": 7435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9738539995661288,
+      "learning_rate": 5.631087669217356e-06,
+      "loss": 0.7638,
+      "step": 7436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7669070524942934,
+      "learning_rate": 5.630059457696741e-06,
+      "loss": 0.6713,
+      "step": 7437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8465919472699919,
+      "learning_rate": 5.629031219101065e-06,
+      "loss": 0.7711,
+      "step": 7438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6964165566992941,
+      "learning_rate": 5.628002953474521e-06,
+      "loss": 0.7798,
+      "step": 7439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4109060786777652,
+      "learning_rate": 5.626974660861288e-06,
+      "loss": 0.7722,
+      "step": 7440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2244934576897524,
+      "learning_rate": 5.6259463413055604e-06,
+      "loss": 0.6866,
+      "step": 7441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.620855668148381,
+      "learning_rate": 5.624917994851523e-06,
+      "loss": 0.6847,
+      "step": 7442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5396796103435653,
+      "learning_rate": 5.6238896215433666e-06,
+      "loss": 0.789,
+      "step": 7443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0443819223253956,
+      "learning_rate": 5.622861221425286e-06,
+      "loss": 0.7668,
+      "step": 7444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.639145124864632,
+      "learning_rate": 5.6218327945414695e-06,
+      "loss": 0.6581,
+      "step": 7445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6936573187396629,
+      "learning_rate": 5.620804340936114e-06,
+      "loss": 0.6947,
+      "step": 7446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1879644753256662,
+      "learning_rate": 5.619775860653412e-06,
+      "loss": 0.7308,
+      "step": 7447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4922824272045991,
+      "learning_rate": 5.6187473537375635e-06,
+      "loss": 0.7117,
+      "step": 7448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6039343350947137,
+      "learning_rate": 5.617718820232762e-06,
+      "loss": 0.8824,
+      "step": 7449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5857095253866234,
+      "learning_rate": 5.616690260183206e-06,
+      "loss": 0.7346,
+      "step": 7450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7778716277243665,
+      "learning_rate": 5.615661673633098e-06,
+      "loss": 0.7091,
+      "step": 7451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4783799660705628,
+      "learning_rate": 5.614633060626634e-06,
+      "loss": 0.6717,
+      "step": 7452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7802538933755294,
+      "learning_rate": 5.613604421208021e-06,
+      "loss": 0.7549,
+      "step": 7453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6218471798104122,
+      "learning_rate": 5.612575755421459e-06,
+      "loss": 0.6962,
+      "step": 7454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.616574680336746,
+      "learning_rate": 5.611547063311152e-06,
+      "loss": 0.7126,
+      "step": 7455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7300501529394443,
+      "learning_rate": 5.610518344921305e-06,
+      "loss": 0.726,
+      "step": 7456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5948489946753583,
+      "learning_rate": 5.609489600296127e-06,
+      "loss": 0.5955,
+      "step": 7457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1263958155179714,
+      "learning_rate": 5.608460829479821e-06,
+      "loss": 0.7721,
+      "step": 7458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.816921971990411,
+      "learning_rate": 5.607432032516601e-06,
+      "loss": 0.6707,
+      "step": 7459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.82740458419387,
+      "learning_rate": 5.606403209450672e-06,
+      "loss": 0.8114,
+      "step": 7460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6073482876651164,
+      "learning_rate": 5.605374360326246e-06,
+      "loss": 0.6095,
+      "step": 7461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5280059132172616,
+      "learning_rate": 5.604345485187535e-06,
+      "loss": 0.7844,
+      "step": 7462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8905918049688364,
+      "learning_rate": 5.603316584078754e-06,
+      "loss": 0.7007,
+      "step": 7463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5523269179908386,
+      "learning_rate": 5.602287657044116e-06,
+      "loss": 0.7463,
+      "step": 7464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6143824891672685,
+      "learning_rate": 5.601258704127835e-06,
+      "loss": 0.6702,
+      "step": 7465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6587526310768463,
+      "learning_rate": 5.600229725374129e-06,
+      "loss": 0.7467,
+      "step": 7466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7422736254834466,
+      "learning_rate": 5.599200720827215e-06,
+      "loss": 0.752,
+      "step": 7467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7302837858473943,
+      "learning_rate": 5.598171690531312e-06,
+      "loss": 0.7121,
+      "step": 7468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1273584692403713,
+      "learning_rate": 5.597142634530639e-06,
+      "loss": 0.7282,
+      "step": 7469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7332549537185804,
+      "learning_rate": 5.5961135528694155e-06,
+      "loss": 0.8002,
+      "step": 7470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6646514483782957,
+      "learning_rate": 5.595084445591866e-06,
+      "loss": 0.6606,
+      "step": 7471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6985125814942594,
+      "learning_rate": 5.594055312742213e-06,
+      "loss": 0.645,
+      "step": 7472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4498106660634758,
+      "learning_rate": 5.59302615436468e-06,
+      "loss": 0.6669,
+      "step": 7473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8447676078122952,
+      "learning_rate": 5.5919969705034914e-06,
+      "loss": 0.7423,
+      "step": 7474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6388327240247234,
+      "learning_rate": 5.590967761202876e-06,
+      "loss": 0.6896,
+      "step": 7475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.695426352312091,
+      "learning_rate": 5.589938526507059e-06,
+      "loss": 0.7636,
+      "step": 7476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7808735406040495,
+      "learning_rate": 5.5889092664602696e-06,
+      "loss": 0.6783,
+      "step": 7477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.699114351876886,
+      "learning_rate": 5.5878799811067384e-06,
+      "loss": 0.6924,
+      "step": 7478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0281859187243088,
+      "learning_rate": 5.586850670490694e-06,
+      "loss": 0.6213,
+      "step": 7479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5501998669351713,
+      "learning_rate": 5.58582133465637e-06,
+      "loss": 0.6773,
+      "step": 7480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.594902643710725,
+      "learning_rate": 5.584791973647997e-06,
+      "loss": 0.7647,
+      "step": 7481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.063676088107756,
+      "learning_rate": 5.583762587509812e-06,
+      "loss": 0.6601,
+      "step": 7482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7374826406327133,
+      "learning_rate": 5.582733176286048e-06,
+      "loss": 0.7881,
+      "step": 7483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9100096211109772,
+      "learning_rate": 5.581703740020943e-06,
+      "loss": 0.678,
+      "step": 7484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8333142388827124,
+      "learning_rate": 5.580674278758732e-06,
+      "loss": 0.656,
+      "step": 7485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5770962465977743,
+      "learning_rate": 5.579644792543653e-06,
+      "loss": 0.7277,
+      "step": 7486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7396597311559212,
+      "learning_rate": 5.578615281419947e-06,
+      "loss": 0.7925,
+      "step": 7487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.666286394318608,
+      "learning_rate": 5.577585745431854e-06,
+      "loss": 0.8134,
+      "step": 7488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.652109635310369,
+      "learning_rate": 5.576556184623615e-06,
+      "loss": 0.7945,
+      "step": 7489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.810522115317186,
+      "learning_rate": 5.575526599039472e-06,
+      "loss": 0.8186,
+      "step": 7490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.045801659915348,
+      "learning_rate": 5.574496988723669e-06,
+      "loss": 0.7064,
+      "step": 7491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5863799887608951,
+      "learning_rate": 5.573467353720452e-06,
+      "loss": 0.8225,
+      "step": 7492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6072430873788641,
+      "learning_rate": 5.572437694074065e-06,
+      "loss": 0.6983,
+      "step": 7493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5606466623543347,
+      "learning_rate": 5.571408009828757e-06,
+      "loss": 0.6853,
+      "step": 7494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7045866045497329,
+      "learning_rate": 5.570378301028771e-06,
+      "loss": 0.773,
+      "step": 7495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.016227432809367,
+      "learning_rate": 5.569348567718363e-06,
+      "loss": 0.5608,
+      "step": 7496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4936380527566195,
+      "learning_rate": 5.568318809941777e-06,
+      "loss": 0.745,
+      "step": 7497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5652700205104695,
+      "learning_rate": 5.567289027743266e-06,
+      "loss": 0.6804,
+      "step": 7498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.697284851817034,
+      "learning_rate": 5.56625922116708e-06,
+      "loss": 0.6804,
+      "step": 7499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3629713504524275,
+      "learning_rate": 5.565229390257478e-06,
+      "loss": 0.6266,
+      "step": 7500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6288690024136074,
+      "learning_rate": 5.564199535058708e-06,
+      "loss": 0.7515,
+      "step": 7501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9064324651186386,
+      "learning_rate": 5.563169655615029e-06,
+      "loss": 0.7299,
+      "step": 7502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7136050087707275,
+      "learning_rate": 5.562139751970694e-06,
+      "loss": 0.7582,
+      "step": 7503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5765283446214389,
+      "learning_rate": 5.561109824169962e-06,
+      "loss": 0.7143,
+      "step": 7504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6703819627265137,
+      "learning_rate": 5.5600798722570925e-06,
+      "loss": 0.7054,
+      "step": 7505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4059087739182048,
+      "learning_rate": 5.559049896276343e-06,
+      "loss": 0.7031,
+      "step": 7506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.90308032247517,
+      "learning_rate": 5.558019896271975e-06,
+      "loss": 0.7339,
+      "step": 7507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.082479766101908,
+      "learning_rate": 5.556989872288248e-06,
+      "loss": 0.7864,
+      "step": 7508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5894002712843096,
+      "learning_rate": 5.555959824369426e-06,
+      "loss": 0.7006,
+      "step": 7509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6004343367219263,
+      "learning_rate": 5.554929752559772e-06,
+      "loss": 0.7282,
+      "step": 7510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8598053510797594,
+      "learning_rate": 5.553899656903552e-06,
+      "loss": 0.6528,
+      "step": 7511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.015295269819532,
+      "learning_rate": 5.552869537445029e-06,
+      "loss": 0.5765,
+      "step": 7512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7515336876164032,
+      "learning_rate": 5.55183939422847e-06,
+      "loss": 0.7621,
+      "step": 7513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6818687429603685,
+      "learning_rate": 5.550809227298144e-06,
+      "loss": 0.8136,
+      "step": 7514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.97369839189723,
+      "learning_rate": 5.54977903669832e-06,
+      "loss": 0.6005,
+      "step": 7515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8666964110664686,
+      "learning_rate": 5.548748822473265e-06,
+      "loss": 0.7901,
+      "step": 7516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5339037500406398,
+      "learning_rate": 5.54771858466725e-06,
+      "loss": 0.688,
+      "step": 7517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1201757606012452,
+      "learning_rate": 5.546688323324548e-06,
+      "loss": 0.603,
+      "step": 7518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.661861263474399,
+      "learning_rate": 5.545658038489433e-06,
+      "loss": 0.7199,
+      "step": 7519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4721487241820148,
+      "learning_rate": 5.544627730206176e-06,
+      "loss": 0.7497,
+      "step": 7520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5541856360727375,
+      "learning_rate": 5.543597398519053e-06,
+      "loss": 0.7956,
+      "step": 7521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6288944624075559,
+      "learning_rate": 5.542567043472337e-06,
+      "loss": 0.6503,
+      "step": 7522
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.815629643030386,
+      "learning_rate": 5.541536665110309e-06,
+      "loss": 0.7134,
+      "step": 7523
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6872146643666845,
+      "learning_rate": 5.540506263477243e-06,
+      "loss": 0.7874,
+      "step": 7524
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5120599181083378,
+      "learning_rate": 5.53947583861742e-06,
+      "loss": 0.7472,
+      "step": 7525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1994061466221144,
+      "learning_rate": 5.5384453905751175e-06,
+      "loss": 0.6499,
+      "step": 7526
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7841144481953028,
+      "learning_rate": 5.537414919394618e-06,
+      "loss": 0.8335,
+      "step": 7527
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1891863218035308,
+      "learning_rate": 5.536384425120202e-06,
+      "loss": 0.6692,
+      "step": 7528
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1180528427839485,
+      "learning_rate": 5.535353907796155e-06,
+      "loss": 0.6737,
+      "step": 7529
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8937116724379468,
+      "learning_rate": 5.534323367466758e-06,
+      "loss": 0.7341,
+      "step": 7530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3266852798578976,
+      "learning_rate": 5.5332928041762936e-06,
+      "loss": 0.7217,
+      "step": 7531
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.872123261113515,
+      "learning_rate": 5.5322622179690514e-06,
+      "loss": 0.8225,
+      "step": 7532
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.750405545964471,
+      "learning_rate": 5.531231608889313e-06,
+      "loss": 0.8594,
+      "step": 7533
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7520564822501166,
+      "learning_rate": 5.530200976981375e-06,
+      "loss": 0.6907,
+      "step": 7534
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6168195183902199,
+      "learning_rate": 5.529170322289515e-06,
+      "loss": 0.8008,
+      "step": 7535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5977297875540526,
+      "learning_rate": 5.528139644858031e-06,
+      "loss": 0.7311,
+      "step": 7536
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.663602388958666,
+      "learning_rate": 5.527108944731207e-06,
+      "loss": 0.7114,
+      "step": 7537
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7044251513459172,
+      "learning_rate": 5.526078221953341e-06,
+      "loss": 0.7496,
+      "step": 7538
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7280883402692144,
+      "learning_rate": 5.525047476568722e-06,
+      "loss": 0.8918,
+      "step": 7539
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2518056251855052,
+      "learning_rate": 5.524016708621641e-06,
+      "loss": 0.6097,
+      "step": 7540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0192452803567424,
+      "learning_rate": 5.522985918156397e-06,
+      "loss": 0.758,
+      "step": 7541
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4392255664321212,
+      "learning_rate": 5.5219551052172825e-06,
+      "loss": 0.7785,
+      "step": 7542
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5427321430827579,
+      "learning_rate": 5.5209242698485955e-06,
+      "loss": 0.7126,
+      "step": 7543
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.878421842246138,
+      "learning_rate": 5.519893412094631e-06,
+      "loss": 0.7315,
+      "step": 7544
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.663806909290669,
+      "learning_rate": 5.518862531999691e-06,
+      "loss": 0.6897,
+      "step": 7545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.83541310738147,
+      "learning_rate": 5.51783162960807e-06,
+      "loss": 0.7961,
+      "step": 7546
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2844188036007729,
+      "learning_rate": 5.516800704964071e-06,
+      "loss": 0.6716,
+      "step": 7547
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8114961403727075,
+      "learning_rate": 5.515769758111996e-06,
+      "loss": 0.7351,
+      "step": 7548
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4890440954812183,
+      "learning_rate": 5.514738789096146e-06,
+      "loss": 0.6907,
+      "step": 7549
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4864745316247578,
+      "learning_rate": 5.513707797960823e-06,
+      "loss": 0.7197,
+      "step": 7550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6083560308984581,
+      "learning_rate": 5.512676784750332e-06,
+      "loss": 0.6758,
+      "step": 7551
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2213013205371974,
+      "learning_rate": 5.511645749508977e-06,
+      "loss": 0.7525,
+      "step": 7552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7488582440738323,
+      "learning_rate": 5.5106146922810664e-06,
+      "loss": 0.7234,
+      "step": 7553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6935241710149191,
+      "learning_rate": 5.509583613110904e-06,
+      "loss": 0.6679,
+      "step": 7554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7652392069729081,
+      "learning_rate": 5.508552512042799e-06,
+      "loss": 0.7113,
+      "step": 7555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.684041975078135,
+      "learning_rate": 5.50752138912106e-06,
+      "loss": 0.7093,
+      "step": 7556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7942624034135337,
+      "learning_rate": 5.506490244389997e-06,
+      "loss": 0.8568,
+      "step": 7557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5656924260216911,
+      "learning_rate": 5.50545907789392e-06,
+      "loss": 0.7353,
+      "step": 7558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7709499960415083,
+      "learning_rate": 5.504427889677141e-06,
+      "loss": 0.6892,
+      "step": 7559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.637544231405416,
+      "learning_rate": 5.50339667978397e-06,
+      "loss": 0.7675,
+      "step": 7560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9826940599918519,
+      "learning_rate": 5.502365448258724e-06,
+      "loss": 0.7585,
+      "step": 7561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.108930880548422,
+      "learning_rate": 5.5013341951457166e-06,
+      "loss": 0.6608,
+      "step": 7562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1927920339583538,
+      "learning_rate": 5.500302920489262e-06,
+      "loss": 0.6253,
+      "step": 7563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4834327941409806,
+      "learning_rate": 5.499271624333676e-06,
+      "loss": 0.603,
+      "step": 7564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8546733081851523,
+      "learning_rate": 5.498240306723277e-06,
+      "loss": 0.9159,
+      "step": 7565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6510645664177517,
+      "learning_rate": 5.497208967702382e-06,
+      "loss": 0.7381,
+      "step": 7566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8098415345906451,
+      "learning_rate": 5.496177607315312e-06,
+      "loss": 0.7093,
+      "step": 7567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.80906310273073,
+      "learning_rate": 5.495146225606383e-06,
+      "loss": 0.7298,
+      "step": 7568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.044519110115154,
+      "learning_rate": 5.494114822619918e-06,
+      "loss": 0.6329,
+      "step": 7569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6408615494025736,
+      "learning_rate": 5.493083398400239e-06,
+      "loss": 0.7036,
+      "step": 7570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.769509160138884,
+      "learning_rate": 5.492051952991669e-06,
+      "loss": 0.7053,
+      "step": 7571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4060501995145538,
+      "learning_rate": 5.4910204864385306e-06,
+      "loss": 0.5907,
+      "step": 7572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6860887418717092,
+      "learning_rate": 5.489988998785147e-06,
+      "loss": 0.7891,
+      "step": 7573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5808063107138115,
+      "learning_rate": 5.488957490075846e-06,
+      "loss": 0.6969,
+      "step": 7574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5450406605006057,
+      "learning_rate": 5.487925960354953e-06,
+      "loss": 0.5783,
+      "step": 7575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5165086636535519,
+      "learning_rate": 5.486894409666794e-06,
+      "loss": 0.7649,
+      "step": 7576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.795213142901571,
+      "learning_rate": 5.485862838055699e-06,
+      "loss": 0.7803,
+      "step": 7577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.525294776159687,
+      "learning_rate": 5.4848312455659935e-06,
+      "loss": 0.6251,
+      "step": 7578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5766170145894858,
+      "learning_rate": 5.483799632242012e-06,
+      "loss": 0.7728,
+      "step": 7579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5819638078104892,
+      "learning_rate": 5.482767998128081e-06,
+      "loss": 0.7733,
+      "step": 7580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4876081532229877,
+      "learning_rate": 5.4817363432685355e-06,
+      "loss": 0.6944,
+      "step": 7581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7278936077903255,
+      "learning_rate": 5.4807046677077045e-06,
+      "loss": 0.7103,
+      "step": 7582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6957004723471578,
+      "learning_rate": 5.479672971489924e-06,
+      "loss": 0.7222,
+      "step": 7583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7157297950648132,
+      "learning_rate": 5.478641254659528e-06,
+      "loss": 0.7902,
+      "step": 7584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6738037451192436,
+      "learning_rate": 5.4776095172608535e-06,
+      "loss": 0.6984,
+      "step": 7585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6368957716625485,
+      "learning_rate": 5.476577759338232e-06,
+      "loss": 0.7348,
+      "step": 7586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6289421917006706,
+      "learning_rate": 5.475545980936003e-06,
+      "loss": 0.7178,
+      "step": 7587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8476876478009414,
+      "learning_rate": 5.474514182098504e-06,
+      "loss": 0.7868,
+      "step": 7588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7200014903189194,
+      "learning_rate": 5.473482362870073e-06,
+      "loss": 0.7067,
+      "step": 7589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5659954205077085,
+      "learning_rate": 5.472450523295052e-06,
+      "loss": 0.7148,
+      "step": 7590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.514049540826455,
+      "learning_rate": 5.471418663417779e-06,
+      "loss": 0.82,
+      "step": 7591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0392164901943635,
+      "learning_rate": 5.470386783282597e-06,
+      "loss": 0.8021,
+      "step": 7592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7937185975114855,
+      "learning_rate": 5.469354882933845e-06,
+      "loss": 0.7548,
+      "step": 7593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6692904700546543,
+      "learning_rate": 5.468322962415871e-06,
+      "loss": 0.6223,
+      "step": 7594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9741218098057072,
+      "learning_rate": 5.4672910217730155e-06,
+      "loss": 0.7215,
+      "step": 7595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4202434306600316,
+      "learning_rate": 5.4662590610496235e-06,
+      "loss": 0.74,
+      "step": 7596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7338091279623558,
+      "learning_rate": 5.465227080290042e-06,
+      "loss": 0.677,
+      "step": 7597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4837746776799823,
+      "learning_rate": 5.464195079538615e-06,
+      "loss": 0.7463,
+      "step": 7598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2195201779467661,
+      "learning_rate": 5.463163058839694e-06,
+      "loss": 0.6466,
+      "step": 7599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.707869480951546,
+      "learning_rate": 5.462131018237623e-06,
+      "loss": 0.8049,
+      "step": 7600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1408961747029838,
+      "learning_rate": 5.461098957776755e-06,
+      "loss": 0.6495,
+      "step": 7601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5319678594502897,
+      "learning_rate": 5.4600668775014355e-06,
+      "loss": 0.7598,
+      "step": 7602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6679267992169788,
+      "learning_rate": 5.459034777456018e-06,
+      "loss": 0.8395,
+      "step": 7603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6832422096336224,
+      "learning_rate": 5.4580026576848565e-06,
+      "loss": 0.6923,
+      "step": 7604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5777156475642915,
+      "learning_rate": 5.4569705182322975e-06,
+      "loss": 0.7106,
+      "step": 7605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0708026849989811,
+      "learning_rate": 5.455938359142699e-06,
+      "loss": 0.5794,
+      "step": 7606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7919973769006052,
+      "learning_rate": 5.454906180460413e-06,
+      "loss": 0.8208,
+      "step": 7607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9738787562217202,
+      "learning_rate": 5.453873982229797e-06,
+      "loss": 0.7634,
+      "step": 7608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7223355860892848,
+      "learning_rate": 5.452841764495203e-06,
+      "loss": 0.6701,
+      "step": 7609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6298314411910013,
+      "learning_rate": 5.451809527300992e-06,
+      "loss": 0.6653,
+      "step": 7610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5827379970340292,
+      "learning_rate": 5.450777270691517e-06,
+      "loss": 0.7268,
+      "step": 7611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6030594312154438,
+      "learning_rate": 5.449744994711141e-06,
+      "loss": 0.7338,
+      "step": 7612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7440903576973552,
+      "learning_rate": 5.448712699404221e-06,
+      "loss": 0.6835,
+      "step": 7613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8552996381345022,
+      "learning_rate": 5.4476803848151146e-06,
+      "loss": 0.7948,
+      "step": 7614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7097723117468895,
+      "learning_rate": 5.446648050988187e-06,
+      "loss": 0.7737,
+      "step": 7615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5489569928774765,
+      "learning_rate": 5.445615697967797e-06,
+      "loss": 0.7779,
+      "step": 7616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6047827443499443,
+      "learning_rate": 5.444583325798308e-06,
+      "loss": 0.6172,
+      "step": 7617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3638116102764668,
+      "learning_rate": 5.443550934524085e-06,
+      "loss": 0.6477,
+      "step": 7618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.583751045212211,
+      "learning_rate": 5.442518524189489e-06,
+      "loss": 0.7071,
+      "step": 7619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7369388642834174,
+      "learning_rate": 5.441486094838886e-06,
+      "loss": 0.6879,
+      "step": 7620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.021688719635058,
+      "learning_rate": 5.440453646516642e-06,
+      "loss": 0.6304,
+      "step": 7621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7990493621830925,
+      "learning_rate": 5.439421179267126e-06,
+      "loss": 0.7345,
+      "step": 7622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7064087546195097,
+      "learning_rate": 5.438388693134702e-06,
+      "loss": 0.7291,
+      "step": 7623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3119512678231438,
+      "learning_rate": 5.4373561881637405e-06,
+      "loss": 0.8295,
+      "step": 7624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4990414616567413,
+      "learning_rate": 5.436323664398607e-06,
+      "loss": 0.6743,
+      "step": 7625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.84428423572651,
+      "learning_rate": 5.435291121883675e-06,
+      "loss": 0.7324,
+      "step": 7626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5383387513799152,
+      "learning_rate": 5.434258560663316e-06,
+      "loss": 0.882,
+      "step": 7627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7702231880350252,
+      "learning_rate": 5.433225980781898e-06,
+      "loss": 0.8207,
+      "step": 7628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.617685489895309,
+      "learning_rate": 5.432193382283794e-06,
+      "loss": 0.6352,
+      "step": 7629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.811271617080683,
+      "learning_rate": 5.431160765213379e-06,
+      "loss": 0.7581,
+      "step": 7630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9809116141180104,
+      "learning_rate": 5.430128129615025e-06,
+      "loss": 0.5798,
+      "step": 7631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6963018836029187,
+      "learning_rate": 5.429095475533108e-06,
+      "loss": 0.7806,
+      "step": 7632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6039808783707106,
+      "learning_rate": 5.428062803012003e-06,
+      "loss": 0.7619,
+      "step": 7633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6371311003846825,
+      "learning_rate": 5.4270301120960856e-06,
+      "loss": 0.7061,
+      "step": 7634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.662502034036486,
+      "learning_rate": 5.425997402829732e-06,
+      "loss": 0.6369,
+      "step": 7635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5869420496028235,
+      "learning_rate": 5.424964675257324e-06,
+      "loss": 0.6208,
+      "step": 7636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.216942024539661,
+      "learning_rate": 5.423931929423235e-06,
+      "loss": 0.7134,
+      "step": 7637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9358421424903542,
+      "learning_rate": 5.422899165371846e-06,
+      "loss": 0.7263,
+      "step": 7638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5966854675775064,
+      "learning_rate": 5.421866383147541e-06,
+      "loss": 0.7486,
+      "step": 7639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.371999407328986,
+      "learning_rate": 5.420833582794696e-06,
+      "loss": 0.796,
+      "step": 7640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5934054197596617,
+      "learning_rate": 5.419800764357697e-06,
+      "loss": 0.7497,
+      "step": 7641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0436225804819048,
+      "learning_rate": 5.418767927880923e-06,
+      "loss": 0.5965,
+      "step": 7642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.521854507488094,
+      "learning_rate": 5.417735073408756e-06,
+      "loss": 0.7468,
+      "step": 7643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4522479038668688,
+      "learning_rate": 5.416702200985585e-06,
+      "loss": 0.6945,
+      "step": 7644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7823981248082792,
+      "learning_rate": 5.415669310655791e-06,
+      "loss": 0.682,
+      "step": 7645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9237365790976617,
+      "learning_rate": 5.414636402463764e-06,
+      "loss": 0.7472,
+      "step": 7646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.601742824942308,
+      "learning_rate": 5.413603476453884e-06,
+      "loss": 0.7087,
+      "step": 7647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7489835332268713,
+      "learning_rate": 5.4125705326705425e-06,
+      "loss": 0.6963,
+      "step": 7648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9196478536242798,
+      "learning_rate": 5.411537571158127e-06,
+      "loss": 0.7404,
+      "step": 7649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7035834093807702,
+      "learning_rate": 5.410504591961025e-06,
+      "loss": 0.7998,
+      "step": 7650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2032852462323527,
+      "learning_rate": 5.409471595123628e-06,
+      "loss": 0.6697,
+      "step": 7651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6246774448935608,
+      "learning_rate": 5.408438580690322e-06,
+      "loss": 0.6763,
+      "step": 7652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1855669895523033,
+      "learning_rate": 5.4074055487055025e-06,
+      "loss": 0.6233,
+      "step": 7653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8504075208604291,
+      "learning_rate": 5.406372499213557e-06,
+      "loss": 0.8249,
+      "step": 7654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.497857653229911,
+      "learning_rate": 5.4053394322588825e-06,
+      "loss": 0.7842,
+      "step": 7655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6576887453860667,
+      "learning_rate": 5.404306347885868e-06,
+      "loss": 0.7516,
+      "step": 7656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0437095205173803,
+      "learning_rate": 5.403273246138912e-06,
+      "loss": 0.5941,
+      "step": 7657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6641864283754166,
+      "learning_rate": 5.4022401270624036e-06,
+      "loss": 0.6379,
+      "step": 7658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.036798792810019,
+      "learning_rate": 5.401206990700741e-06,
+      "loss": 0.6523,
+      "step": 7659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.726383040612301,
+      "learning_rate": 5.400173837098324e-06,
+      "loss": 0.6258,
+      "step": 7660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.038769565995671,
+      "learning_rate": 5.399140666299543e-06,
+      "loss": 0.6431,
+      "step": 7661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5620983910001878,
+      "learning_rate": 5.3981074783488e-06,
+      "loss": 0.6667,
+      "step": 7662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6855874176713703,
+      "learning_rate": 5.3970742732904905e-06,
+      "loss": 0.759,
+      "step": 7663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.716128950415977,
+      "learning_rate": 5.396041051169016e-06,
+      "loss": 0.8249,
+      "step": 7664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0943546678920033,
+      "learning_rate": 5.395007812028775e-06,
+      "loss": 0.6173,
+      "step": 7665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8380519051405886,
+      "learning_rate": 5.39397455591417e-06,
+      "loss": 0.6273,
+      "step": 7666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.589636137556641,
+      "learning_rate": 5.3929412828696006e-06,
+      "loss": 0.6847,
+      "step": 7667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.99058616884271,
+      "learning_rate": 5.3919079929394665e-06,
+      "loss": 0.6869,
+      "step": 7668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1063260334239755,
+      "learning_rate": 5.390874686168176e-06,
+      "loss": 0.6047,
+      "step": 7669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6800968016533895,
+      "learning_rate": 5.389841362600129e-06,
+      "loss": 0.8203,
+      "step": 7670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8170287790687811,
+      "learning_rate": 5.388808022279731e-06,
+      "loss": 0.7575,
+      "step": 7671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2341942727475463,
+      "learning_rate": 5.387774665251385e-06,
+      "loss": 0.7844,
+      "step": 7672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5919732033917404,
+      "learning_rate": 5.386741291559499e-06,
+      "loss": 0.6306,
+      "step": 7673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.717146904511777,
+      "learning_rate": 5.385707901248478e-06,
+      "loss": 0.7379,
+      "step": 7674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5707348873907878,
+      "learning_rate": 5.3846744943627295e-06,
+      "loss": 0.7489,
+      "step": 7675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.673225011119813,
+      "learning_rate": 5.38364107094666e-06,
+      "loss": 0.7707,
+      "step": 7676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.475712364131529,
+      "learning_rate": 5.382607631044682e-06,
+      "loss": 0.7822,
+      "step": 7677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4054689214181426,
+      "learning_rate": 5.381574174701201e-06,
+      "loss": 0.6856,
+      "step": 7678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.711883011526424,
+      "learning_rate": 5.380540701960627e-06,
+      "loss": 0.6686,
+      "step": 7679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5184357458539313,
+      "learning_rate": 5.379507212867372e-06,
+      "loss": 0.7194,
+      "step": 7680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8859264540937353,
+      "learning_rate": 5.378473707465847e-06,
+      "loss": 0.7111,
+      "step": 7681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.434061736056138,
+      "learning_rate": 5.377440185800463e-06,
+      "loss": 0.7627,
+      "step": 7682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6185924152789741,
+      "learning_rate": 5.376406647915635e-06,
+      "loss": 0.7137,
+      "step": 7683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.764237023930959,
+      "learning_rate": 5.375373093855774e-06,
+      "loss": 0.7853,
+      "step": 7684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1410672751160167,
+      "learning_rate": 5.374339523665295e-06,
+      "loss": 0.6855,
+      "step": 7685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9006212610747146,
+      "learning_rate": 5.373305937388613e-06,
+      "loss": 0.8289,
+      "step": 7686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0969784976798307,
+      "learning_rate": 5.372272335070145e-06,
+      "loss": 0.7349,
+      "step": 7687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4844654581730765,
+      "learning_rate": 5.371238716754305e-06,
+      "loss": 0.9668,
+      "step": 7688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1750525338801878,
+      "learning_rate": 5.37020508248551e-06,
+      "loss": 0.6714,
+      "step": 7689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6967560739897274,
+      "learning_rate": 5.369171432308178e-06,
+      "loss": 0.776,
+      "step": 7690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8397131984801947,
+      "learning_rate": 5.368137766266728e-06,
+      "loss": 0.6957,
+      "step": 7691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.028194084101294,
+      "learning_rate": 5.367104084405578e-06,
+      "loss": 0.7157,
+      "step": 7692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2490394459485854,
+      "learning_rate": 5.3660703867691475e-06,
+      "loss": 0.6896,
+      "step": 7693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8775661825578027,
+      "learning_rate": 5.365036673401857e-06,
+      "loss": 0.7643,
+      "step": 7694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0616065107481814,
+      "learning_rate": 5.364002944348129e-06,
+      "loss": 0.5735,
+      "step": 7695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.67605893612148,
+      "learning_rate": 5.362969199652382e-06,
+      "loss": 0.7238,
+      "step": 7696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5458967429039814,
+      "learning_rate": 5.361935439359042e-06,
+      "loss": 0.7408,
+      "step": 7697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6760598304726635,
+      "learning_rate": 5.36090166351253e-06,
+      "loss": 0.7864,
+      "step": 7698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2646584594277008,
+      "learning_rate": 5.359867872157267e-06,
+      "loss": 0.7029,
+      "step": 7699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8649994213202636,
+      "learning_rate": 5.358834065337684e-06,
+      "loss": 0.7478,
+      "step": 7700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.743263535861343,
+      "learning_rate": 5.357800243098199e-06,
+      "loss": 0.7746,
+      "step": 7701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7572846137895595,
+      "learning_rate": 5.3567664054832426e-06,
+      "loss": 0.6826,
+      "step": 7702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8321329144811638,
+      "learning_rate": 5.355732552537237e-06,
+      "loss": 0.735,
+      "step": 7703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.81880184781535,
+      "learning_rate": 5.354698684304613e-06,
+      "loss": 0.7132,
+      "step": 7704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7597624897536825,
+      "learning_rate": 5.353664800829796e-06,
+      "loss": 0.8154,
+      "step": 7705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.660613883498111,
+      "learning_rate": 5.352630902157217e-06,
+      "loss": 0.6358,
+      "step": 7706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1998031839330505,
+      "learning_rate": 5.3515969883313e-06,
+      "loss": 0.5726,
+      "step": 7707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3716454299029623,
+      "learning_rate": 5.350563059396478e-06,
+      "loss": 0.7461,
+      "step": 7708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6199923498284166,
+      "learning_rate": 5.3495291153971806e-06,
+      "loss": 0.7779,
+      "step": 7709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.761771904130704,
+      "learning_rate": 5.348495156377839e-06,
+      "loss": 0.7066,
+      "step": 7710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5624144549999865,
+      "learning_rate": 5.347461182382886e-06,
+      "loss": 0.7897,
+      "step": 7711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5355766790307963,
+      "learning_rate": 5.346427193456749e-06,
+      "loss": 0.7533,
+      "step": 7712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7772770529387953,
+      "learning_rate": 5.345393189643866e-06,
+      "loss": 0.8326,
+      "step": 7713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5664521099403084,
+      "learning_rate": 5.344359170988668e-06,
+      "loss": 0.7846,
+      "step": 7714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5933467574108235,
+      "learning_rate": 5.3433251375355884e-06,
+      "loss": 0.7486,
+      "step": 7715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.525275064250936,
+      "learning_rate": 5.342291089329066e-06,
+      "loss": 0.7626,
+      "step": 7716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.81053636484869,
+      "learning_rate": 5.34125702641353e-06,
+      "loss": 0.7857,
+      "step": 7717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4752536204781368,
+      "learning_rate": 5.340222948833421e-06,
+      "loss": 0.7528,
+      "step": 7718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6854022554490236,
+      "learning_rate": 5.339188856633173e-06,
+      "loss": 0.6959,
+      "step": 7719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5258116411411144,
+      "learning_rate": 5.338154749857225e-06,
+      "loss": 0.8146,
+      "step": 7720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9794105178517033,
+      "learning_rate": 5.337120628550016e-06,
+      "loss": 0.5907,
+      "step": 7721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7596696559019476,
+      "learning_rate": 5.3360864927559805e-06,
+      "loss": 0.7181,
+      "step": 7722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1633300266713962,
+      "learning_rate": 5.335052342519561e-06,
+      "loss": 0.6485,
+      "step": 7723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0865436468640417,
+      "learning_rate": 5.3340181778851954e-06,
+      "loss": 0.6929,
+      "step": 7724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1567506081581098,
+      "learning_rate": 5.332983998897327e-06,
+      "loss": 0.7075,
+      "step": 7725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5893288103659946,
+      "learning_rate": 5.331949805600391e-06,
+      "loss": 0.73,
+      "step": 7726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8191808627729515,
+      "learning_rate": 5.330915598038835e-06,
+      "loss": 0.7533,
+      "step": 7727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.680170907421823,
+      "learning_rate": 5.329881376257098e-06,
+      "loss": 0.805,
+      "step": 7728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1629202007645174,
+      "learning_rate": 5.328847140299624e-06,
+      "loss": 0.5798,
+      "step": 7729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7461495765891315,
+      "learning_rate": 5.327812890210856e-06,
+      "loss": 0.7058,
+      "step": 7730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6340221607256482,
+      "learning_rate": 5.326778626035238e-06,
+      "loss": 0.7263,
+      "step": 7731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6895713026200194,
+      "learning_rate": 5.3257443478172146e-06,
+      "loss": 0.7625,
+      "step": 7732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0900785578457932,
+      "learning_rate": 5.32471005560123e-06,
+      "loss": 0.6282,
+      "step": 7733
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.549028939328378,
+      "learning_rate": 5.323675749431732e-06,
+      "loss": 0.7614,
+      "step": 7734
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1470388750407137,
+      "learning_rate": 5.322641429353167e-06,
+      "loss": 0.6842,
+      "step": 7735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8251636741754078,
+      "learning_rate": 5.321607095409981e-06,
+      "loss": 0.8241,
+      "step": 7736
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6942604392353153,
+      "learning_rate": 5.3205727476466206e-06,
+      "loss": 0.699,
+      "step": 7737
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6353132144548397,
+      "learning_rate": 5.3195383861075365e-06,
+      "loss": 0.7518,
+      "step": 7738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3790288550480545,
+      "learning_rate": 5.318504010837175e-06,
+      "loss": 0.7213,
+      "step": 7739
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9417526308595425,
+      "learning_rate": 5.3174696218799894e-06,
+      "loss": 0.8308,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1259233252362755,
+      "learning_rate": 5.316435219280425e-06,
+      "loss": 0.7922,
+      "step": 7741
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.549565814560876,
+      "learning_rate": 5.315400803082934e-06,
+      "loss": 0.6847,
+      "step": 7742
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7438215521209288,
+      "learning_rate": 5.314366373331969e-06,
+      "loss": 0.778,
+      "step": 7743
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2842699876565324,
+      "learning_rate": 5.313331930071981e-06,
+      "loss": 0.7172,
+      "step": 7744
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.70667659786039,
+      "learning_rate": 5.312297473347422e-06,
+      "loss": 0.6394,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7102872855950864,
+      "learning_rate": 5.311263003202745e-06,
+      "loss": 0.8461,
+      "step": 7746
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6611720927604083,
+      "learning_rate": 5.310228519682403e-06,
+      "loss": 0.7572,
+      "step": 7747
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.759311774131948,
+      "learning_rate": 5.309194022830852e-06,
+      "loss": 0.7,
+      "step": 7748
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3083797826556136,
+      "learning_rate": 5.308159512692544e-06,
+      "loss": 0.7607,
+      "step": 7749
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1489554235619153,
+      "learning_rate": 5.307124989311936e-06,
+      "loss": 0.6346,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6968141126584386,
+      "learning_rate": 5.306090452733482e-06,
+      "loss": 0.7533,
+      "step": 7751
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5768270256667773,
+      "learning_rate": 5.305055903001642e-06,
+      "loss": 0.6895,
+      "step": 7752
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5834350887360462,
+      "learning_rate": 5.304021340160868e-06,
+      "loss": 0.73,
+      "step": 7753
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5268773764399544,
+      "learning_rate": 5.302986764255621e-06,
+      "loss": 0.832,
+      "step": 7754
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6402520289415705,
+      "learning_rate": 5.301952175330357e-06,
+      "loss": 0.8595,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.18214233455152,
+      "learning_rate": 5.300917573429536e-06,
+      "loss": 0.6388,
+      "step": 7756
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5263323134529039,
+      "learning_rate": 5.299882958597617e-06,
+      "loss": 0.733,
+      "step": 7757
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.631715429604868,
+      "learning_rate": 5.298848330879061e-06,
+      "loss": 0.6178,
+      "step": 7758
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5215277275291597,
+      "learning_rate": 5.297813690318325e-06,
+      "loss": 0.7396,
+      "step": 7759
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7004262161794483,
+      "learning_rate": 5.29677903695987e-06,
+      "loss": 0.7384,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4995849635902332,
+      "learning_rate": 5.295744370848162e-06,
+      "loss": 0.7013,
+      "step": 7761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5989703712561953,
+      "learning_rate": 5.294709692027658e-06,
+      "loss": 0.7749,
+      "step": 7762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.839126619157761,
+      "learning_rate": 5.293675000542822e-06,
+      "loss": 0.8651,
+      "step": 7763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5468997126551374,
+      "learning_rate": 5.292640296438116e-06,
+      "loss": 0.6516,
+      "step": 7764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6350092642337752,
+      "learning_rate": 5.291605579758007e-06,
+      "loss": 0.6245,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5311102589378285,
+      "learning_rate": 5.290570850546954e-06,
+      "loss": 0.6751,
+      "step": 7766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8368994024257332,
+      "learning_rate": 5.289536108849429e-06,
+      "loss": 0.7074,
+      "step": 7767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6204208544511378,
+      "learning_rate": 5.28850135470989e-06,
+      "loss": 0.7023,
+      "step": 7768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5674991232677002,
+      "learning_rate": 5.287466588172804e-06,
+      "loss": 0.7943,
+      "step": 7769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3964556508800179,
+      "learning_rate": 5.286431809282639e-06,
+      "loss": 0.568,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.603700840489099,
+      "learning_rate": 5.2853970180838624e-06,
+      "loss": 0.8388,
+      "step": 7771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6792205717978397,
+      "learning_rate": 5.2843622146209405e-06,
+      "loss": 0.7196,
+      "step": 7772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.507508137296848,
+      "learning_rate": 5.283327398938339e-06,
+      "loss": 0.7537,
+      "step": 7773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7118657370073942,
+      "learning_rate": 5.2822925710805305e-06,
+      "loss": 0.8406,
+      "step": 7774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8759473895132923,
+      "learning_rate": 5.28125773109198e-06,
+      "loss": 0.8343,
+      "step": 7775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2204550216265666,
+      "learning_rate": 5.2802228790171596e-06,
+      "loss": 0.7055,
+      "step": 7776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.482157042365586,
+      "learning_rate": 5.279188014900537e-06,
+      "loss": 0.5986,
+      "step": 7777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8478890319709707,
+      "learning_rate": 5.278153138786583e-06,
+      "loss": 0.7211,
+      "step": 7778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8763444540898984,
+      "learning_rate": 5.27711825071977e-06,
+      "loss": 0.6825,
+      "step": 7779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3881277367798552,
+      "learning_rate": 5.276083350744568e-06,
+      "loss": 0.6703,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9792839346798294,
+      "learning_rate": 5.275048438905452e-06,
+      "loss": 0.7303,
+      "step": 7781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2803761124954802,
+      "learning_rate": 5.27401351524689e-06,
+      "loss": 0.7234,
+      "step": 7782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5995935298567485,
+      "learning_rate": 5.272978579813357e-06,
+      "loss": 0.749,
+      "step": 7783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6753900069402878,
+      "learning_rate": 5.2719436326493255e-06,
+      "loss": 0.7624,
+      "step": 7784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8017038852221139,
+      "learning_rate": 5.270908673799272e-06,
+      "loss": 0.7243,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9116341036682212,
+      "learning_rate": 5.26987370330767e-06,
+      "loss": 0.6682,
+      "step": 7786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6741840213900525,
+      "learning_rate": 5.2688387212189915e-06,
+      "loss": 0.7999,
+      "step": 7787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7072761471669156,
+      "learning_rate": 5.267803727577717e-06,
+      "loss": 0.7973,
+      "step": 7788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0195749008024177,
+      "learning_rate": 5.266768722428318e-06,
+      "loss": 0.7601,
+      "step": 7789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5571463464199162,
+      "learning_rate": 5.2657337058152735e-06,
+      "loss": 0.7399,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7402307661551002,
+      "learning_rate": 5.26469867778306e-06,
+      "loss": 0.7636,
+      "step": 7791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.535154944548455,
+      "learning_rate": 5.263663638376155e-06,
+      "loss": 0.6641,
+      "step": 7792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6983062414675643,
+      "learning_rate": 5.262628587639036e-06,
+      "loss": 0.8472,
+      "step": 7793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6754101958676129,
+      "learning_rate": 5.261593525616181e-06,
+      "loss": 0.7031,
+      "step": 7794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4466695316924203,
+      "learning_rate": 5.2605584523520715e-06,
+      "loss": 0.8534,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5410661891909194,
+      "learning_rate": 5.259523367891184e-06,
+      "loss": 0.7399,
+      "step": 7796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7907561471028177,
+      "learning_rate": 5.258488272277998e-06,
+      "loss": 0.7489,
+      "step": 7797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5227766719253888,
+      "learning_rate": 5.257453165556996e-06,
+      "loss": 0.7089,
+      "step": 7798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.165669276196909,
+      "learning_rate": 5.256418047772659e-06,
+      "loss": 0.6572,
+      "step": 7799
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4573842825510335,
+      "learning_rate": 5.255382918969467e-06,
+      "loss": 0.8413,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4702576170690425,
+      "learning_rate": 5.2543477791919025e-06,
+      "loss": 0.7088,
+      "step": 7801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5883942070602817,
+      "learning_rate": 5.253312628484447e-06,
+      "loss": 0.5916,
+      "step": 7802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.039534525450695,
+      "learning_rate": 5.252277466891583e-06,
+      "loss": 0.6641,
+      "step": 7803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.44982140256563,
+      "learning_rate": 5.251242294457796e-06,
+      "loss": 0.6961,
+      "step": 7804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.363308280757994,
+      "learning_rate": 5.2502071112275675e-06,
+      "loss": 0.6409,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5330182070393257,
+      "learning_rate": 5.249171917245382e-06,
+      "loss": 0.7463,
+      "step": 7806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.505033899847728,
+      "learning_rate": 5.248136712555724e-06,
+      "loss": 0.7176,
+      "step": 7807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6695420759311257,
+      "learning_rate": 5.247101497203079e-06,
+      "loss": 0.7721,
+      "step": 7808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.039073695714937,
+      "learning_rate": 5.2460662712319335e-06,
+      "loss": 0.6116,
+      "step": 7809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.913380175237871,
+      "learning_rate": 5.24503103468677e-06,
+      "loss": 0.6042,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6557816747144731,
+      "learning_rate": 5.243995787612078e-06,
+      "loss": 0.7508,
+      "step": 7811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6493026475708092,
+      "learning_rate": 5.242960530052344e-06,
+      "loss": 0.7297,
+      "step": 7812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7752610408166283,
+      "learning_rate": 5.241925262052055e-06,
+      "loss": 0.6167,
+      "step": 7813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5740891128302468,
+      "learning_rate": 5.240889983655701e-06,
+      "loss": 0.7049,
+      "step": 7814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6925012644496076,
+      "learning_rate": 5.239854694907764e-06,
+      "loss": 0.8097,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5867493680948865,
+      "learning_rate": 5.238819395852738e-06,
+      "loss": 0.7148,
+      "step": 7816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.478282941404453,
+      "learning_rate": 5.2377840865351115e-06,
+      "loss": 0.7177,
+      "step": 7817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1335327363495278,
+      "learning_rate": 5.236748766999372e-06,
+      "loss": 0.7074,
+      "step": 7818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8474475100057524,
+      "learning_rate": 5.235713437290012e-06,
+      "loss": 0.6813,
+      "step": 7819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.476993592615844,
+      "learning_rate": 5.234678097451519e-06,
+      "loss": 0.6888,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6488594709950954,
+      "learning_rate": 5.233642747528387e-06,
+      "loss": 0.6761,
+      "step": 7821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4748072062762667,
+      "learning_rate": 5.232607387565104e-06,
+      "loss": 0.6495,
+      "step": 7822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.813831531717927,
+      "learning_rate": 5.231572017606166e-06,
+      "loss": 0.6545,
+      "step": 7823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5010599228865702,
+      "learning_rate": 5.230536637696062e-06,
+      "loss": 0.7952,
+      "step": 7824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4561308231530514,
+      "learning_rate": 5.229501247879284e-06,
+      "loss": 0.6382,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7364569052064653,
+      "learning_rate": 5.228465848200327e-06,
+      "loss": 0.7328,
+      "step": 7826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6004390129417243,
+      "learning_rate": 5.227430438703683e-06,
+      "loss": 0.7967,
+      "step": 7827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.206502577557424,
+      "learning_rate": 5.226395019433847e-06,
+      "loss": 0.7359,
+      "step": 7828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5293229817420195,
+      "learning_rate": 5.225359590435312e-06,
+      "loss": 0.5805,
+      "step": 7829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7099289465210596,
+      "learning_rate": 5.224324151752575e-06,
+      "loss": 0.6968,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7515336416143157,
+      "learning_rate": 5.223288703430128e-06,
+      "loss": 0.7221,
+      "step": 7831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6925663093962144,
+      "learning_rate": 5.222253245512471e-06,
+      "loss": 0.7605,
+      "step": 7832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2454862715498893,
+      "learning_rate": 5.221217778044096e-06,
+      "loss": 0.7565,
+      "step": 7833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1585827674904516,
+      "learning_rate": 5.220182301069499e-06,
+      "loss": 0.8183,
+      "step": 7834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5822061118508632,
+      "learning_rate": 5.219146814633178e-06,
+      "loss": 0.6965,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4410305814832147,
+      "learning_rate": 5.218111318779631e-06,
+      "loss": 0.7876,
+      "step": 7836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9098444724017567,
+      "learning_rate": 5.2170758135533554e-06,
+      "loss": 0.7412,
+      "step": 7837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.153140136796043,
+      "learning_rate": 5.2160402989988475e-06,
+      "loss": 0.8312,
+      "step": 7838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7488620661957799,
+      "learning_rate": 5.215004775160608e-06,
+      "loss": 0.7728,
+      "step": 7839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.091594979221933,
+      "learning_rate": 5.2139692420831325e-06,
+      "loss": 0.5603,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0494215004868372,
+      "learning_rate": 5.212933699810924e-06,
+      "loss": 0.6779,
+      "step": 7841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5953846246840144,
+      "learning_rate": 5.211898148388481e-06,
+      "loss": 0.6841,
+      "step": 7842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.609993950069927,
+      "learning_rate": 5.2108625878603e-06,
+      "loss": 0.7214,
+      "step": 7843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.578705887033242,
+      "learning_rate": 5.209827018270886e-06,
+      "loss": 0.5864,
+      "step": 7844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8883013168297103,
+      "learning_rate": 5.208791439664736e-06,
+      "loss": 0.7747,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0766791719067488,
+      "learning_rate": 5.207755852086355e-06,
+      "loss": 0.699,
+      "step": 7846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5848254812424876,
+      "learning_rate": 5.206720255580241e-06,
+      "loss": 0.6807,
+      "step": 7847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.768957604203999,
+      "learning_rate": 5.2056846501908975e-06,
+      "loss": 0.7016,
+      "step": 7848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5635035638293555,
+      "learning_rate": 5.204649035962825e-06,
+      "loss": 0.6751,
+      "step": 7849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1106162849929102,
+      "learning_rate": 5.2036134129405305e-06,
+      "loss": 0.763,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8367764647390132,
+      "learning_rate": 5.202577781168513e-06,
+      "loss": 0.8101,
+      "step": 7851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8560024435940197,
+      "learning_rate": 5.201542140691277e-06,
+      "loss": 0.8151,
+      "step": 7852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.943628206976688,
+      "learning_rate": 5.200506491553326e-06,
+      "loss": 0.7454,
+      "step": 7853
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.515895431747729,
+      "learning_rate": 5.199470833799164e-06,
+      "loss": 0.732,
+      "step": 7854
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2058227857750203,
+      "learning_rate": 5.198435167473297e-06,
+      "loss": 0.6569,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4435148940311482,
+      "learning_rate": 5.1973994926202296e-06,
+      "loss": 0.6234,
+      "step": 7856
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9800176170940822,
+      "learning_rate": 5.196363809284466e-06,
+      "loss": 0.6649,
+      "step": 7857
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6282829763005027,
+      "learning_rate": 5.19532811751051e-06,
+      "loss": 0.631,
+      "step": 7858
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.660368644729554,
+      "learning_rate": 5.1942924173428725e-06,
+      "loss": 0.7286,
+      "step": 7859
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7001806601713427,
+      "learning_rate": 5.193256708826055e-06,
+      "loss": 0.7327,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5108557606209785,
+      "learning_rate": 5.192220992004569e-06,
+      "loss": 0.6252,
+      "step": 7861
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.764124569557034,
+      "learning_rate": 5.191185266922917e-06,
+      "loss": 0.8232,
+      "step": 7862
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.975881581891303,
+      "learning_rate": 5.190149533625607e-06,
+      "loss": 0.7387,
+      "step": 7863
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7633368828264122,
+      "learning_rate": 5.18911379215715e-06,
+      "loss": 0.7369,
+      "step": 7864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5067551186768027,
+      "learning_rate": 5.188078042562052e-06,
+      "loss": 0.5974,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6920967471076467,
+      "learning_rate": 5.187042284884821e-06,
+      "loss": 0.698,
+      "step": 7866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.053549143916168,
+      "learning_rate": 5.186006519169967e-06,
+      "loss": 0.6385,
+      "step": 7867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.721935833936345,
+      "learning_rate": 5.184970745461998e-06,
+      "loss": 0.7599,
+      "step": 7868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.624040974256335,
+      "learning_rate": 5.1839349638054245e-06,
+      "loss": 0.7804,
+      "step": 7869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5389275520958225,
+      "learning_rate": 5.182899174244757e-06,
+      "loss": 0.6133,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9366121520535495,
+      "learning_rate": 5.181863376824504e-06,
+      "loss": 0.7485,
+      "step": 7871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7139858236141967,
+      "learning_rate": 5.1808275715891756e-06,
+      "loss": 0.6622,
+      "step": 7872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6480863432861146,
+      "learning_rate": 5.1797917585832846e-06,
+      "loss": 0.8508,
+      "step": 7873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6949416729989377,
+      "learning_rate": 5.178755937851341e-06,
+      "loss": 0.7136,
+      "step": 7874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7313493154702515,
+      "learning_rate": 5.177720109437857e-06,
+      "loss": 0.6085,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5997451074425468,
+      "learning_rate": 5.176684273387344e-06,
+      "loss": 0.7609,
+      "step": 7876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3536643108598017,
+      "learning_rate": 5.1756484297443135e-06,
+      "loss": 0.7488,
+      "step": 7877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5711083198981324,
+      "learning_rate": 5.1746125785532795e-06,
+      "loss": 0.7693,
+      "step": 7878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1766002658046086,
+      "learning_rate": 5.173576719858755e-06,
+      "loss": 0.7384,
+      "step": 7879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8893485116899897,
+      "learning_rate": 5.172540853705251e-06,
+      "loss": 0.7228,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8523668589549942,
+      "learning_rate": 5.171504980137282e-06,
+      "loss": 0.7825,
+      "step": 7881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3752795231776322,
+      "learning_rate": 5.170469099199363e-06,
+      "loss": 0.7356,
+      "step": 7882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.721891859609702,
+      "learning_rate": 5.169433210936007e-06,
+      "loss": 0.682,
+      "step": 7883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8486580669144055,
+      "learning_rate": 5.168397315391729e-06,
+      "loss": 0.6861,
+      "step": 7884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6655394687733782,
+      "learning_rate": 5.167361412611042e-06,
+      "loss": 0.6717,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1510653541581077,
+      "learning_rate": 5.1663255026384625e-06,
+      "loss": 0.6428,
+      "step": 7886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.914979500289328,
+      "learning_rate": 5.165289585518507e-06,
+      "loss": 0.7463,
+      "step": 7887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9357593994154911,
+      "learning_rate": 5.164253661295689e-06,
+      "loss": 0.8919,
+      "step": 7888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.54000328339009,
+      "learning_rate": 5.1632177300145255e-06,
+      "loss": 0.7077,
+      "step": 7889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5577548187651369,
+      "learning_rate": 5.162181791719532e-06,
+      "loss": 0.7598,
+      "step": 7890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3714909903614427,
+      "learning_rate": 5.161145846455227e-06,
+      "loss": 0.7583,
+      "step": 7891
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5555990747304596,
+      "learning_rate": 5.160109894266125e-06,
+      "loss": 0.6791,
+      "step": 7892
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.530365772511538,
+      "learning_rate": 5.159073935196746e-06,
+      "loss": 0.7411,
+      "step": 7893
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6058469585981858,
+      "learning_rate": 5.1580379692916025e-06,
+      "loss": 0.8161,
+      "step": 7894
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4521342693346249,
+      "learning_rate": 5.157001996595218e-06,
+      "loss": 0.6582,
+      "step": 7895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7007141521938898,
+      "learning_rate": 5.155966017152108e-06,
+      "loss": 0.7084,
+      "step": 7896
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6654456924626537,
+      "learning_rate": 5.1549300310067896e-06,
+      "loss": 0.8593,
+      "step": 7897
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.008935352458989,
+      "learning_rate": 5.153894038203785e-06,
+      "loss": 0.6825,
+      "step": 7898
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.933740752243675,
+      "learning_rate": 5.152858038787608e-06,
+      "loss": 0.793,
+      "step": 7899
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9617042566102498,
+      "learning_rate": 5.151822032802782e-06,
+      "loss": 0.7367,
+      "step": 7900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6162276627479473,
+      "learning_rate": 5.150786020293824e-06,
+      "loss": 0.8479,
+      "step": 7901
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9873766251775333,
+      "learning_rate": 5.149750001305257e-06,
+      "loss": 0.7891,
+      "step": 7902
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6434647984497663,
+      "learning_rate": 5.148713975881598e-06,
+      "loss": 0.669,
+      "step": 7903
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9655195252966952,
+      "learning_rate": 5.147677944067368e-06,
+      "loss": 0.5785,
+      "step": 7904
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1584286179444327,
+      "learning_rate": 5.146641905907088e-06,
+      "loss": 0.6486,
+      "step": 7905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.534907531343163,
+      "learning_rate": 5.145605861445279e-06,
+      "loss": 0.6193,
+      "step": 7906
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6216732570732078,
+      "learning_rate": 5.144569810726462e-06,
+      "loss": 0.7188,
+      "step": 7907
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.747653846460813,
+      "learning_rate": 5.143533753795157e-06,
+      "loss": 0.8019,
+      "step": 7908
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7777755824723864,
+      "learning_rate": 5.142497690695888e-06,
+      "loss": 0.7552,
+      "step": 7909
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.799686514642581,
+      "learning_rate": 5.141461621473175e-06,
+      "loss": 0.8042,
+      "step": 7910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7004440828855136,
+      "learning_rate": 5.140425546171542e-06,
+      "loss": 0.7581,
+      "step": 7911
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7969916063904339,
+      "learning_rate": 5.13938946483551e-06,
+      "loss": 0.8486,
+      "step": 7912
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.41045705787896,
+      "learning_rate": 5.1383533775096025e-06,
+      "loss": 0.792,
+      "step": 7913
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0269242101103206,
+      "learning_rate": 5.137317284238344e-06,
+      "loss": 0.671,
+      "step": 7914
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3888491723479817,
+      "learning_rate": 5.136281185066253e-06,
+      "loss": 0.7276,
+      "step": 7915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.710412045393934,
+      "learning_rate": 5.135245080037857e-06,
+      "loss": 0.6752,
+      "step": 7916
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8196483281314908,
+      "learning_rate": 5.1342089691976794e-06,
+      "loss": 0.7535,
+      "step": 7917
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.587984957302592,
+      "learning_rate": 5.1331728525902435e-06,
+      "loss": 0.6702,
+      "step": 7918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5348288597529616,
+      "learning_rate": 5.1321367302600726e-06,
+      "loss": 0.657,
+      "step": 7919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.166316169524862,
+      "learning_rate": 5.131100602251692e-06,
+      "loss": 0.6828,
+      "step": 7920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6095793324360415,
+      "learning_rate": 5.130064468609627e-06,
+      "loss": 0.7081,
+      "step": 7921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6179908934126752,
+      "learning_rate": 5.129028329378403e-06,
+      "loss": 0.7373,
+      "step": 7922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.756455100451225,
+      "learning_rate": 5.127992184602542e-06,
+      "loss": 0.7246,
+      "step": 7923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6028771048256045,
+      "learning_rate": 5.126956034326573e-06,
+      "loss": 0.6681,
+      "step": 7924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.585572663381009,
+      "learning_rate": 5.125919878595022e-06,
+      "loss": 0.7051,
+      "step": 7925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8202311735787853,
+      "learning_rate": 5.124883717452412e-06,
+      "loss": 0.7059,
+      "step": 7926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.776107776841329,
+      "learning_rate": 5.123847550943272e-06,
+      "loss": 0.6688,
+      "step": 7927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.689066938530323,
+      "learning_rate": 5.122811379112124e-06,
+      "loss": 0.7582,
+      "step": 7928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9091303672857076,
+      "learning_rate": 5.121775202003499e-06,
+      "loss": 0.5686,
+      "step": 7929
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6817320684241008,
+      "learning_rate": 5.120739019661923e-06,
+      "loss": 0.7164,
+      "step": 7930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5743303189199418,
+      "learning_rate": 5.119702832131922e-06,
+      "loss": 0.7216,
+      "step": 7931
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.676213855295697,
+      "learning_rate": 5.118666639458023e-06,
+      "loss": 0.7054,
+      "step": 7932
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5462270285349278,
+      "learning_rate": 5.117630441684755e-06,
+      "loss": 0.7317,
+      "step": 7933
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4928853264912054,
+      "learning_rate": 5.116594238856645e-06,
+      "loss": 0.7788,
+      "step": 7934
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5792774061409072,
+      "learning_rate": 5.115558031018221e-06,
+      "loss": 0.7101,
+      "step": 7935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7096010710096035,
+      "learning_rate": 5.114521818214011e-06,
+      "loss": 0.6381,
+      "step": 7936
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9524447019241262,
+      "learning_rate": 5.113485600488541e-06,
+      "loss": 0.732,
+      "step": 7937
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7525236865143183,
+      "learning_rate": 5.112449377886345e-06,
+      "loss": 0.7621,
+      "step": 7938
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5823544316309726,
+      "learning_rate": 5.111413150451948e-06,
+      "loss": 0.6686,
+      "step": 7939
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5347281079647863,
+      "learning_rate": 5.11037691822988e-06,
+      "loss": 0.8059,
+      "step": 7940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1850885841873062,
+      "learning_rate": 5.10934068126467e-06,
+      "loss": 0.6567,
+      "step": 7941
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8617771801096938,
+      "learning_rate": 5.108304439600847e-06,
+      "loss": 0.8245,
+      "step": 7942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.622522151535485,
+      "learning_rate": 5.107268193282941e-06,
+      "loss": 0.7729,
+      "step": 7943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.488370801557423,
+      "learning_rate": 5.1062319423554815e-06,
+      "loss": 0.6901,
+      "step": 7944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6609479734161428,
+      "learning_rate": 5.105195686863e-06,
+      "loss": 0.7957,
+      "step": 7945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9790921666385244,
+      "learning_rate": 5.104159426850024e-06,
+      "loss": 0.771,
+      "step": 7946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.835482325933701,
+      "learning_rate": 5.103123162361087e-06,
+      "loss": 0.686,
+      "step": 7947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.753889410656475,
+      "learning_rate": 5.102086893440717e-06,
+      "loss": 0.8002,
+      "step": 7948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6471469963999594,
+      "learning_rate": 5.101050620133447e-06,
+      "loss": 0.6569,
+      "step": 7949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6243844479644105,
+      "learning_rate": 5.100014342483804e-06,
+      "loss": 0.6587,
+      "step": 7950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5772678484560592,
+      "learning_rate": 5.098978060536325e-06,
+      "loss": 0.737,
+      "step": 7951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2558363050936419,
+      "learning_rate": 5.097941774335538e-06,
+      "loss": 0.6914,
+      "step": 7952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1683526316464747,
+      "learning_rate": 5.096905483925972e-06,
+      "loss": 0.651,
+      "step": 7953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8363645240448223,
+      "learning_rate": 5.095869189352166e-06,
+      "loss": 0.7427,
+      "step": 7954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7916056261357725,
+      "learning_rate": 5.094832890658644e-06,
+      "loss": 0.6724,
+      "step": 7955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.536556646569109,
+      "learning_rate": 5.0937965878899435e-06,
+      "loss": 0.5909,
+      "step": 7956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7454038600111073,
+      "learning_rate": 5.092760281090592e-06,
+      "loss": 0.7837,
+      "step": 7957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7376978340095757,
+      "learning_rate": 5.091723970305128e-06,
+      "loss": 0.6355,
+      "step": 7958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7010428440771466,
+      "learning_rate": 5.090687655578078e-06,
+      "loss": 0.6635,
+      "step": 7959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6229802900954062,
+      "learning_rate": 5.089651336953979e-06,
+      "loss": 0.6764,
+      "step": 7960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0362512538428998,
+      "learning_rate": 5.088615014477362e-06,
+      "loss": 0.7287,
+      "step": 7961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.141209964550675,
+      "learning_rate": 5.087578688192759e-06,
+      "loss": 0.7771,
+      "step": 7962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7937449886219345,
+      "learning_rate": 5.086542358144708e-06,
+      "loss": 0.8374,
+      "step": 7963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.524309149997358,
+      "learning_rate": 5.0855060243777366e-06,
+      "loss": 0.7162,
+      "step": 7964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8700449014983336,
+      "learning_rate": 5.084469686936382e-06,
+      "loss": 0.6814,
+      "step": 7965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1128732229983993,
+      "learning_rate": 5.083433345865175e-06,
+      "loss": 0.695,
+      "step": 7966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4398784450333098,
+      "learning_rate": 5.082397001208653e-06,
+      "loss": 0.7354,
+      "step": 7967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5245015579172898,
+      "learning_rate": 5.081360653011349e-06,
+      "loss": 0.7407,
+      "step": 7968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7160263947134806,
+      "learning_rate": 5.080324301317795e-06,
+      "loss": 0.7114,
+      "step": 7969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3654598073673287,
+      "learning_rate": 5.079287946172529e-06,
+      "loss": 0.686,
+      "step": 7970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8386562475845707,
+      "learning_rate": 5.07825158762008e-06,
+      "loss": 0.8382,
+      "step": 7971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0597523161166094,
+      "learning_rate": 5.077215225704988e-06,
+      "loss": 0.7717,
+      "step": 7972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.71070234705447,
+      "learning_rate": 5.076178860471787e-06,
+      "loss": 0.8189,
+      "step": 7973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.895555540987407,
+      "learning_rate": 5.0751424919650085e-06,
+      "loss": 0.7406,
+      "step": 7974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4498212941843183,
+      "learning_rate": 5.07410612022919e-06,
+      "loss": 0.686,
+      "step": 7975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6177785018079744,
+      "learning_rate": 5.073069745308868e-06,
+      "loss": 0.7173,
+      "step": 7976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.686661766311077,
+      "learning_rate": 5.072033367248574e-06,
+      "loss": 0.7571,
+      "step": 7977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7484996232494392,
+      "learning_rate": 5.0709969860928485e-06,
+      "loss": 0.7756,
+      "step": 7978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7008824476040794,
+      "learning_rate": 5.069960601886224e-06,
+      "loss": 0.9137,
+      "step": 7979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5171471048948546,
+      "learning_rate": 5.068924214673234e-06,
+      "loss": 0.7778,
+      "step": 7980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7692905004878972,
+      "learning_rate": 5.067887824498419e-06,
+      "loss": 0.8083,
+      "step": 7981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7824247301956673,
+      "learning_rate": 5.0668514314063135e-06,
+      "loss": 0.8334,
+      "step": 7982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7275798576236667,
+      "learning_rate": 5.065815035441452e-06,
+      "loss": 0.6746,
+      "step": 7983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9546674772777404,
+      "learning_rate": 5.064778636648371e-06,
+      "loss": 0.6878,
+      "step": 7984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.769312106254405,
+      "learning_rate": 5.063742235071609e-06,
+      "loss": 0.7343,
+      "step": 7985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.706885735248818,
+      "learning_rate": 5.062705830755702e-06,
+      "loss": 0.762,
+      "step": 7986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5545021026047745,
+      "learning_rate": 5.061669423745185e-06,
+      "loss": 0.6746,
+      "step": 7987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4975554710531387,
+      "learning_rate": 5.060633014084595e-06,
+      "loss": 0.6524,
+      "step": 7988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6602406702066062,
+      "learning_rate": 5.05959660181847e-06,
+      "loss": 0.7587,
+      "step": 7989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1089290119741937,
+      "learning_rate": 5.058560186991346e-06,
+      "loss": 0.6345,
+      "step": 7990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8088703388969891,
+      "learning_rate": 5.0575237696477596e-06,
+      "loss": 0.7712,
+      "step": 7991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6288428424551995,
+      "learning_rate": 5.05648734983225e-06,
+      "loss": 0.641,
+      "step": 7992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8498849659367202,
+      "learning_rate": 5.055450927589352e-06,
+      "loss": 0.8061,
+      "step": 7993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5957357275193103,
+      "learning_rate": 5.054414502963605e-06,
+      "loss": 0.9315,
+      "step": 7994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6214524296128903,
+      "learning_rate": 5.053378075999545e-06,
+      "loss": 0.8806,
+      "step": 7995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0233610215995175,
+      "learning_rate": 5.052341646741711e-06,
+      "loss": 0.7084,
+      "step": 7996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0663750134635483,
+      "learning_rate": 5.05130521523464e-06,
+      "loss": 0.8274,
+      "step": 7997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8963478970584287,
+      "learning_rate": 5.050268781522868e-06,
+      "loss": 0.7346,
+      "step": 7998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4985936620608373,
+      "learning_rate": 5.049232345650936e-06,
+      "loss": 0.7028,
+      "step": 7999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2207304634605096,
+      "learning_rate": 5.04819590766338e-06,
+      "loss": 0.6452,
+      "step": 8000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.577967080057421,
+      "learning_rate": 5.0471594676047385e-06,
+      "loss": 0.9138,
+      "step": 8001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.895434384875941,
+      "learning_rate": 5.0461230255195495e-06,
+      "loss": 0.8844,
+      "step": 8002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5426648388298059,
+      "learning_rate": 5.045086581452352e-06,
+      "loss": 0.8668,
+      "step": 8003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.352898816264319,
+      "learning_rate": 5.044050135447682e-06,
+      "loss": 0.6837,
+      "step": 8004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.696723417625502,
+      "learning_rate": 5.043013687550083e-06,
+      "loss": 0.7757,
+      "step": 8005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6576315202905028,
+      "learning_rate": 5.041977237804088e-06,
+      "loss": 0.6192,
+      "step": 8006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1266119316274534,
+      "learning_rate": 5.040940786254236e-06,
+      "loss": 0.7275,
+      "step": 8007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5341769414167752,
+      "learning_rate": 5.039904332945069e-06,
+      "loss": 0.6664,
+      "step": 8008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8811547740617982,
+      "learning_rate": 5.038867877921124e-06,
+      "loss": 0.7073,
+      "step": 8009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9600587430085752,
+      "learning_rate": 5.037831421226939e-06,
+      "loss": 0.6161,
+      "step": 8010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7114664363938143,
+      "learning_rate": 5.036794962907054e-06,
+      "loss": 0.7809,
+      "step": 8011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2525773529361697,
+      "learning_rate": 5.035758503006007e-06,
+      "loss": 0.6154,
+      "step": 8012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5716765870570075,
+      "learning_rate": 5.034722041568337e-06,
+      "loss": 0.7925,
+      "step": 8013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9404067243155296,
+      "learning_rate": 5.033685578638586e-06,
+      "loss": 0.7465,
+      "step": 8014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8718998728281762,
+      "learning_rate": 5.03264911426129e-06,
+      "loss": 0.8334,
+      "step": 8015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.254345739075206,
+      "learning_rate": 5.031612648480987e-06,
+      "loss": 0.5684,
+      "step": 8016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2773391265155176,
+      "learning_rate": 5.03057618134222e-06,
+      "loss": 0.7329,
+      "step": 8017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2234848284816158,
+      "learning_rate": 5.029539712889524e-06,
+      "loss": 0.7352,
+      "step": 8018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.139188101948316,
+      "learning_rate": 5.028503243167443e-06,
+      "loss": 0.6872,
+      "step": 8019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7399120587392867,
+      "learning_rate": 5.027466772220513e-06,
+      "loss": 0.7943,
+      "step": 8020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7837884757843139,
+      "learning_rate": 5.026430300093275e-06,
+      "loss": 0.7084,
+      "step": 8021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7111838611740164,
+      "learning_rate": 5.025393826830267e-06,
+      "loss": 0.7608,
+      "step": 8022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2946996002403715,
+      "learning_rate": 5.024357352476032e-06,
+      "loss": 0.6185,
+      "step": 8023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8434869580060613,
+      "learning_rate": 5.023320877075107e-06,
+      "loss": 0.7804,
+      "step": 8024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7768125256868685,
+      "learning_rate": 5.02228440067203e-06,
+      "loss": 0.7678,
+      "step": 8025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6440894687654617,
+      "learning_rate": 5.0212479233113445e-06,
+      "loss": 0.6993,
+      "step": 8026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1602692949290088,
+      "learning_rate": 5.020211445037587e-06,
+      "loss": 0.7254,
+      "step": 8027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3465394217123599,
+      "learning_rate": 5.019174965895299e-06,
+      "loss": 0.6178,
+      "step": 8028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6634397131706613,
+      "learning_rate": 5.0181384859290215e-06,
+      "loss": 0.7365,
+      "step": 8029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.452489732164864,
+      "learning_rate": 5.017102005183292e-06,
+      "loss": 0.6759,
+      "step": 8030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.207007053183127,
+      "learning_rate": 5.01606552370265e-06,
+      "loss": 0.7155,
+      "step": 8031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6842429418085552,
+      "learning_rate": 5.015029041531639e-06,
+      "loss": 0.647,
+      "step": 8032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8412520005580375,
+      "learning_rate": 5.013992558714795e-06,
+      "loss": 0.7612,
+      "step": 8033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5769895974072097,
+      "learning_rate": 5.01295607529666e-06,
+      "loss": 0.7366,
+      "step": 8034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.32896119409382,
+      "learning_rate": 5.011919591321773e-06,
+      "loss": 0.8624,
+      "step": 8035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5513081871066865,
+      "learning_rate": 5.010883106834676e-06,
+      "loss": 0.695,
+      "step": 8036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8084123818270137,
+      "learning_rate": 5.009846621879906e-06,
+      "loss": 0.6807,
+      "step": 8037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8322921441267055,
+      "learning_rate": 5.008810136502006e-06,
+      "loss": 0.7553,
+      "step": 8038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6915969173177805,
+      "learning_rate": 5.007773650745514e-06,
+      "loss": 0.7483,
+      "step": 8039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6659502594525017,
+      "learning_rate": 5.00673716465497e-06,
+      "loss": 0.6465,
+      "step": 8040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4714306004494655,
+      "learning_rate": 5.005700678274918e-06,
+      "loss": 0.6353,
+      "step": 8041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0985730529270232,
+      "learning_rate": 5.004664191649893e-06,
+      "loss": 0.6524,
+      "step": 8042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7499398518650202,
+      "learning_rate": 5.003627704824438e-06,
+      "loss": 0.7069,
+      "step": 8043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0839649595199443,
+      "learning_rate": 5.0025912178430925e-06,
+      "loss": 0.7291,
+      "step": 8044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4655616248107337,
+      "learning_rate": 5.001554730750395e-06,
+      "loss": 0.7729,
+      "step": 8045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7186320508195527,
+      "learning_rate": 5.0005182435908885e-06,
+      "loss": 0.7004,
+      "step": 8046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7126023512748816,
+      "learning_rate": 4.999481756409112e-06,
+      "loss": 0.6573,
+      "step": 8047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1601335554462464,
+      "learning_rate": 4.998445269249606e-06,
+      "loss": 0.8736,
+      "step": 8048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9653438556454454,
+      "learning_rate": 4.997408782156909e-06,
+      "loss": 0.679,
+      "step": 8049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7582431117719763,
+      "learning_rate": 4.996372295175563e-06,
+      "loss": 0.7016,
+      "step": 8050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7198534809531518,
+      "learning_rate": 4.995335808350109e-06,
+      "loss": 0.7596,
+      "step": 8051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8668676378869171,
+      "learning_rate": 4.994299321725085e-06,
+      "loss": 0.8042,
+      "step": 8052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5633748234237275,
+      "learning_rate": 4.99326283534503e-06,
+      "loss": 0.7055,
+      "step": 8053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5717913976383528,
+      "learning_rate": 4.9922263492544885e-06,
+      "loss": 0.7056,
+      "step": 8054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0508042258777102,
+      "learning_rate": 4.991189863497996e-06,
+      "loss": 0.5603,
+      "step": 8055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6966073655949092,
+      "learning_rate": 4.990153378120096e-06,
+      "loss": 0.781,
+      "step": 8056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.535879093878998,
+      "learning_rate": 4.989116893165325e-06,
+      "loss": 0.7571,
+      "step": 8057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.612112756471766,
+      "learning_rate": 4.988080408678229e-06,
+      "loss": 0.7331,
+      "step": 8058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7983895075781986,
+      "learning_rate": 4.987043924703342e-06,
+      "loss": 0.7589,
+      "step": 8059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2149694411582566,
+      "learning_rate": 4.986007441285206e-06,
+      "loss": 0.6935,
+      "step": 8060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5357643124143692,
+      "learning_rate": 4.984970958468364e-06,
+      "loss": 0.7653,
+      "step": 8061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8866057939734833,
+      "learning_rate": 4.983934476297351e-06,
+      "loss": 0.6705,
+      "step": 8062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1791708384521844,
+      "learning_rate": 4.9828979948167095e-06,
+      "loss": 0.6824,
+      "step": 8063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2690962389113822,
+      "learning_rate": 4.981861514070979e-06,
+      "loss": 0.6314,
+      "step": 8064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6723049320445522,
+      "learning_rate": 4.980825034104702e-06,
+      "loss": 0.7265,
+      "step": 8065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5761646862090284,
+      "learning_rate": 4.979788554962415e-06,
+      "loss": 0.7552,
+      "step": 8066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.860117695494126,
+      "learning_rate": 4.978752076688657e-06,
+      "loss": 0.5772,
+      "step": 8067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6141565404831104,
+      "learning_rate": 4.97771559932797e-06,
+      "loss": 0.6504,
+      "step": 8068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7076328012945239,
+      "learning_rate": 4.976679122924896e-06,
+      "loss": 0.7916,
+      "step": 8069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5999798630373885,
+      "learning_rate": 4.97564264752397e-06,
+      "loss": 0.7647,
+      "step": 8070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6686122700290278,
+      "learning_rate": 4.974606173169733e-06,
+      "loss": 0.7199,
+      "step": 8071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.598813130006741,
+      "learning_rate": 4.973569699906727e-06,
+      "loss": 0.7481,
+      "step": 8072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6635852367442598,
+      "learning_rate": 4.972533227779489e-06,
+      "loss": 0.8621,
+      "step": 8073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0197634674116436,
+      "learning_rate": 4.971496756832557e-06,
+      "loss": 0.6698,
+      "step": 8074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5545484087115042,
+      "learning_rate": 4.9704602871104765e-06,
+      "loss": 0.6551,
+      "step": 8075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9361170725875294,
+      "learning_rate": 4.969423818657783e-06,
+      "loss": 0.7072,
+      "step": 8076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6532146811118738,
+      "learning_rate": 4.9683873515190145e-06,
+      "loss": 0.7652,
+      "step": 8077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.021482333974357,
+      "learning_rate": 4.9673508857387115e-06,
+      "loss": 0.7688,
+      "step": 8078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.036730409391179,
+      "learning_rate": 4.966314421361416e-06,
+      "loss": 0.7552,
+      "step": 8079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1458989853099606,
+      "learning_rate": 4.965277958431664e-06,
+      "loss": 0.7585,
+      "step": 8080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6026604342613304,
+      "learning_rate": 4.964241496993994e-06,
+      "loss": 0.6149,
+      "step": 8081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.57867163368099,
+      "learning_rate": 4.963205037092947e-06,
+      "loss": 0.7819,
+      "step": 8082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1238396531802868,
+      "learning_rate": 4.962168578773063e-06,
+      "loss": 0.6137,
+      "step": 8083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6756221919326595,
+      "learning_rate": 4.9611321220788775e-06,
+      "loss": 0.724,
+      "step": 8084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6641690779750584,
+      "learning_rate": 4.9600956670549324e-06,
+      "loss": 0.7657,
+      "step": 8085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1286219544347125,
+      "learning_rate": 4.959059213745764e-06,
+      "loss": 0.8148,
+      "step": 8086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4179625396585969,
+      "learning_rate": 4.9580227621959145e-06,
+      "loss": 0.7057,
+      "step": 8087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7113538921430882,
+      "learning_rate": 4.956986312449919e-06,
+      "loss": 0.7514,
+      "step": 8088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.616430051558961,
+      "learning_rate": 4.955949864552318e-06,
+      "loss": 0.6615,
+      "step": 8089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0610531278004554,
+      "learning_rate": 4.954913418547651e-06,
+      "loss": 0.5554,
+      "step": 8090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.45391902222075,
+      "learning_rate": 4.953876974480452e-06,
+      "loss": 0.6681,
+      "step": 8091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9467094862346643,
+      "learning_rate": 4.952840532395262e-06,
+      "loss": 0.7868,
+      "step": 8092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7770121945543662,
+      "learning_rate": 4.951804092336621e-06,
+      "loss": 0.7243,
+      "step": 8093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.557167336330192,
+      "learning_rate": 4.950767654349067e-06,
+      "loss": 0.7395,
+      "step": 8094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9909046477190788,
+      "learning_rate": 4.949731218477133e-06,
+      "loss": 0.7391,
+      "step": 8095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.772607967087778,
+      "learning_rate": 4.9486947847653615e-06,
+      "loss": 0.7164,
+      "step": 8096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7619282088713293,
+      "learning_rate": 4.947658353258291e-06,
+      "loss": 0.8099,
+      "step": 8097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7656194412837678,
+      "learning_rate": 4.946621924000456e-06,
+      "loss": 0.6455,
+      "step": 8098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8387194859472897,
+      "learning_rate": 4.945585497036396e-06,
+      "loss": 0.6648,
+      "step": 8099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6478044865138048,
+      "learning_rate": 4.944549072410648e-06,
+      "loss": 0.7067,
+      "step": 8100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6529849891985904,
+      "learning_rate": 4.9435126501677525e-06,
+      "loss": 0.7788,
+      "step": 8101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9085462034638183,
+      "learning_rate": 4.942476230352241e-06,
+      "loss": 0.7538,
+      "step": 8102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9591547138171582,
+      "learning_rate": 4.941439813008656e-06,
+      "loss": 0.4561,
+      "step": 8103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.14865992223908,
+      "learning_rate": 4.940403398181531e-06,
+      "loss": 0.5973,
+      "step": 8104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6583764008293456,
+      "learning_rate": 4.939366985915408e-06,
+      "loss": 0.7246,
+      "step": 8105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2258663164105366,
+      "learning_rate": 4.938330576254817e-06,
+      "loss": 0.6615,
+      "step": 8106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4210022737014216,
+      "learning_rate": 4.9372941692443e-06,
+      "loss": 0.6715,
+      "step": 8107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.417388355518647,
+      "learning_rate": 4.936257764928392e-06,
+      "loss": 0.6644,
+      "step": 8108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7885093782041308,
+      "learning_rate": 4.935221363351631e-06,
+      "loss": 0.7311,
+      "step": 8109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0174692440280955,
+      "learning_rate": 4.934184964558549e-06,
+      "loss": 0.7426,
+      "step": 8110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6159060989753824,
+      "learning_rate": 4.933148568593687e-06,
+      "loss": 0.7437,
+      "step": 8111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7688533660869776,
+      "learning_rate": 4.932112175501583e-06,
+      "loss": 0.735,
+      "step": 8112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2510516544911425,
+      "learning_rate": 4.931075785326767e-06,
+      "loss": 0.7433,
+      "step": 8113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4339848810845963,
+      "learning_rate": 4.930039398113779e-06,
+      "loss": 0.7512,
+      "step": 8114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8971354380241319,
+      "learning_rate": 4.929003013907152e-06,
+      "loss": 0.7047,
+      "step": 8115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7121174544319884,
+      "learning_rate": 4.927966632751427e-06,
+      "loss": 0.7415,
+      "step": 8116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.831379097187573,
+      "learning_rate": 4.926930254691134e-06,
+      "loss": 0.8209,
+      "step": 8117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.430977061639987,
+      "learning_rate": 4.92589387977081e-06,
+      "loss": 0.6653,
+      "step": 8118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7142178471781668,
+      "learning_rate": 4.924857508034994e-06,
+      "loss": 0.7506,
+      "step": 8119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6533860893692915,
+      "learning_rate": 4.9238211395282156e-06,
+      "loss": 0.7185,
+      "step": 8120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.217485195051953,
+      "learning_rate": 4.922784774295013e-06,
+      "loss": 0.6142,
+      "step": 8121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1060433890201198,
+      "learning_rate": 4.92174841237992e-06,
+      "loss": 0.7612,
+      "step": 8122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7328191415279082,
+      "learning_rate": 4.920712053827475e-06,
+      "loss": 0.8207,
+      "step": 8123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5739900134766722,
+      "learning_rate": 4.919675698682206e-06,
+      "loss": 0.7294,
+      "step": 8124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.644809169169365,
+      "learning_rate": 4.918639346988652e-06,
+      "loss": 0.6934,
+      "step": 8125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.441775397332316,
+      "learning_rate": 4.917602998791348e-06,
+      "loss": 0.7601,
+      "step": 8126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5048326176367608,
+      "learning_rate": 4.9165666541348265e-06,
+      "loss": 0.8192,
+      "step": 8127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8866007950991737,
+      "learning_rate": 4.91553031306362e-06,
+      "loss": 0.7573,
+      "step": 8128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5031454862416145,
+      "learning_rate": 4.914493975622263e-06,
+      "loss": 0.8132,
+      "step": 8129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0594103457934685,
+      "learning_rate": 4.9134576418552945e-06,
+      "loss": 0.6932,
+      "step": 8130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9563483196508588,
+      "learning_rate": 4.9124213118072415e-06,
+      "loss": 0.6405,
+      "step": 8131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.675585626864903,
+      "learning_rate": 4.911384985522639e-06,
+      "loss": 0.6578,
+      "step": 8132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.817901309197307,
+      "learning_rate": 4.910348663046021e-06,
+      "loss": 0.7238,
+      "step": 8133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7715864432725628,
+      "learning_rate": 4.909312344421923e-06,
+      "loss": 0.7916,
+      "step": 8134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3862520973898762,
+      "learning_rate": 4.908276029694873e-06,
+      "loss": 0.6619,
+      "step": 8135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.790535819859783,
+      "learning_rate": 4.907239718909408e-06,
+      "loss": 0.6527,
+      "step": 8136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7712615615259317,
+      "learning_rate": 4.90620341211006e-06,
+      "loss": 0.6744,
+      "step": 8137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0705817560861837,
+      "learning_rate": 4.905167109341357e-06,
+      "loss": 0.7695,
+      "step": 8138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0270963896194618,
+      "learning_rate": 4.904130810647836e-06,
+      "loss": 0.6528,
+      "step": 8139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6194396439505794,
+      "learning_rate": 4.903094516074027e-06,
+      "loss": 0.6554,
+      "step": 8140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5280977343085052,
+      "learning_rate": 4.902058225664465e-06,
+      "loss": 0.7172,
+      "step": 8141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6526828200318173,
+      "learning_rate": 4.901021939463676e-06,
+      "loss": 0.8422,
+      "step": 8142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4010183028320928,
+      "learning_rate": 4.899985657516195e-06,
+      "loss": 0.6193,
+      "step": 8143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.759663365736721,
+      "learning_rate": 4.898949379866556e-06,
+      "loss": 0.8498,
+      "step": 8144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6146468578364914,
+      "learning_rate": 4.897913106559285e-06,
+      "loss": 0.677,
+      "step": 8145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.74547396361004,
+      "learning_rate": 4.896876837638915e-06,
+      "loss": 0.672,
+      "step": 8146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5027981117459213,
+      "learning_rate": 4.895840573149977e-06,
+      "loss": 0.6911,
+      "step": 8147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0211118807801651,
+      "learning_rate": 4.8948043131370025e-06,
+      "loss": 0.6327,
+      "step": 8148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.731484110860571,
+      "learning_rate": 4.893768057644519e-06,
+      "loss": 0.7599,
+      "step": 8149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.645714025504405,
+      "learning_rate": 4.89273180671706e-06,
+      "loss": 0.6646,
+      "step": 8150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7951016435409215,
+      "learning_rate": 4.891695560399154e-06,
+      "loss": 0.7426,
+      "step": 8151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6266819913451882,
+      "learning_rate": 4.890659318735333e-06,
+      "loss": 0.7383,
+      "step": 8152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6365971303398228,
+      "learning_rate": 4.8896230817701214e-06,
+      "loss": 0.6589,
+      "step": 8153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6227860039196642,
+      "learning_rate": 4.888586849548053e-06,
+      "loss": 0.797,
+      "step": 8154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7923237588236551,
+      "learning_rate": 4.887550622113657e-06,
+      "loss": 0.7724,
+      "step": 8155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.676279907403954,
+      "learning_rate": 4.88651439951146e-06,
+      "loss": 0.7178,
+      "step": 8156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7232247295665688,
+      "learning_rate": 4.885478181785991e-06,
+      "loss": 0.7315,
+      "step": 8157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4554288872912993,
+      "learning_rate": 4.884441968981781e-06,
+      "loss": 0.6007,
+      "step": 8158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.771667209749387,
+      "learning_rate": 4.883405761143357e-06,
+      "loss": 0.8296,
+      "step": 8159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6050674185390787,
+      "learning_rate": 4.882369558315246e-06,
+      "loss": 0.6717,
+      "step": 8160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.045946432659539,
+      "learning_rate": 4.8813333605419775e-06,
+      "loss": 0.6511,
+      "step": 8161
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8313421936793877,
+      "learning_rate": 4.88029716786808e-06,
+      "loss": 0.7056,
+      "step": 8162
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.532655208927776,
+      "learning_rate": 4.879260980338078e-06,
+      "loss": 0.7371,
+      "step": 8163
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6486786949649368,
+      "learning_rate": 4.878224797996502e-06,
+      "loss": 0.642,
+      "step": 8164
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1888175869879922,
+      "learning_rate": 4.877188620887876e-06,
+      "loss": 0.7616,
+      "step": 8165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9632074764817167,
+      "learning_rate": 4.8761524490567315e-06,
+      "loss": 0.708,
+      "step": 8166
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5398800492456541,
+      "learning_rate": 4.875116282547589e-06,
+      "loss": 0.6706,
+      "step": 8167
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.967468074035581,
+      "learning_rate": 4.874080121404979e-06,
+      "loss": 0.8111,
+      "step": 8168
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.749400468459773,
+      "learning_rate": 4.873043965673427e-06,
+      "loss": 0.6705,
+      "step": 8169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8174155688494424,
+      "learning_rate": 4.872007815397459e-06,
+      "loss": 0.6798,
+      "step": 8170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6039229296430741,
+      "learning_rate": 4.870971670621599e-06,
+      "loss": 0.7656,
+      "step": 8171
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4829024721861235,
+      "learning_rate": 4.869935531390374e-06,
+      "loss": 0.6748,
+      "step": 8172
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5755419819228045,
+      "learning_rate": 4.86889939774831e-06,
+      "loss": 0.7362,
+      "step": 8173
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4668105396326407,
+      "learning_rate": 4.86786326973993e-06,
+      "loss": 0.781,
+      "step": 8174
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0910677486806803,
+      "learning_rate": 4.866827147409758e-06,
+      "loss": 0.733,
+      "step": 8175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4513928930107862,
+      "learning_rate": 4.8657910308023205e-06,
+      "loss": 0.6458,
+      "step": 8176
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6599292536442691,
+      "learning_rate": 4.864754919962144e-06,
+      "loss": 0.7687,
+      "step": 8177
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.526917202122999,
+      "learning_rate": 4.863718814933749e-06,
+      "loss": 0.7462,
+      "step": 8178
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.766498738356997,
+      "learning_rate": 4.862682715761658e-06,
+      "loss": 0.7099,
+      "step": 8179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7794169421204367,
+      "learning_rate": 4.861646622490399e-06,
+      "loss": 0.6776,
+      "step": 8180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7069616824458826,
+      "learning_rate": 4.860610535164491e-06,
+      "loss": 0.7572,
+      "step": 8181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7492571548712934,
+      "learning_rate": 4.85957445382846e-06,
+      "loss": 0.7955,
+      "step": 8182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.125785522320591,
+      "learning_rate": 4.858538378526825e-06,
+      "loss": 0.8827,
+      "step": 8183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.959716173418002,
+      "learning_rate": 4.857502309304114e-06,
+      "loss": 0.7398,
+      "step": 8184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.098324920389947,
+      "learning_rate": 4.856466246204844e-06,
+      "loss": 0.6525,
+      "step": 8185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6591600079319488,
+      "learning_rate": 4.8554301892735385e-06,
+      "loss": 0.6637,
+      "step": 8186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4235113161485198,
+      "learning_rate": 4.854394138554722e-06,
+      "loss": 0.818,
+      "step": 8187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0107453220035538,
+      "learning_rate": 4.853358094092913e-06,
+      "loss": 0.6136,
+      "step": 8188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7138912739014291,
+      "learning_rate": 4.852322055932633e-06,
+      "loss": 0.7479,
+      "step": 8189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.007740425422642,
+      "learning_rate": 4.851286024118402e-06,
+      "loss": 0.777,
+      "step": 8190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5558285823978044,
+      "learning_rate": 4.850249998694745e-06,
+      "loss": 0.8,
+      "step": 8191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7791312168702178,
+      "learning_rate": 4.849213979706177e-06,
+      "loss": 0.8004,
+      "step": 8192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8256032900740673,
+      "learning_rate": 4.848177967197219e-06,
+      "loss": 0.6727,
+      "step": 8193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8878822367577378,
+      "learning_rate": 4.8471419612123925e-06,
+      "loss": 0.5149,
+      "step": 8194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5790971760002035,
+      "learning_rate": 4.846105961796218e-06,
+      "loss": 0.7545,
+      "step": 8195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8872787792149714,
+      "learning_rate": 4.845069968993212e-06,
+      "loss": 0.693,
+      "step": 8196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2215813792061532,
+      "learning_rate": 4.844033982847893e-06,
+      "loss": 0.5704,
+      "step": 8197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7036616306129295,
+      "learning_rate": 4.842998003404784e-06,
+      "loss": 0.7416,
+      "step": 8198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.772947243827297,
+      "learning_rate": 4.841962030708398e-06,
+      "loss": 0.7401,
+      "step": 8199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8433268011210262,
+      "learning_rate": 4.840926064803256e-06,
+      "loss": 0.8136,
+      "step": 8200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9620605008038037,
+      "learning_rate": 4.839890105733876e-06,
+      "loss": 0.5645,
+      "step": 8201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.355334306858517,
+      "learning_rate": 4.8388541535447756e-06,
+      "loss": 0.7524,
+      "step": 8202
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3707491698575691,
+      "learning_rate": 4.837818208280469e-06,
+      "loss": 0.6505,
+      "step": 8203
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5437883641155616,
+      "learning_rate": 4.836782269985475e-06,
+      "loss": 0.7013,
+      "step": 8204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1074978682705348,
+      "learning_rate": 4.8357463387043124e-06,
+      "loss": 0.7118,
+      "step": 8205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2012182632191686,
+      "learning_rate": 4.834710414481496e-06,
+      "loss": 0.6662,
+      "step": 8206
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4308287892828169,
+      "learning_rate": 4.833674497361539e-06,
+      "loss": 0.7044,
+      "step": 8207
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6430634663888264,
+      "learning_rate": 4.832638587388959e-06,
+      "loss": 0.5792,
+      "step": 8208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.057839712904107,
+      "learning_rate": 4.831602684608274e-06,
+      "loss": 0.7158,
+      "step": 8209
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5845663699320056,
+      "learning_rate": 4.830566789063995e-06,
+      "loss": 0.8339,
+      "step": 8210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0132770399032203,
+      "learning_rate": 4.829530900800638e-06,
+      "loss": 0.7296,
+      "step": 8211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.544898475870902,
+      "learning_rate": 4.8284950198627186e-06,
+      "loss": 0.6722,
+      "step": 8212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7137907664301686,
+      "learning_rate": 4.827459146294751e-06,
+      "loss": 0.8381,
+      "step": 8213
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8391953350981776,
+      "learning_rate": 4.826423280141247e-06,
+      "loss": 0.7843,
+      "step": 8214
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.936329411482093,
+      "learning_rate": 4.825387421446721e-06,
+      "loss": 0.8431,
+      "step": 8215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.974669056710717,
+      "learning_rate": 4.824351570255689e-06,
+      "loss": 0.7215,
+      "step": 8216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.78985532176945,
+      "learning_rate": 4.823315726612659e-06,
+      "loss": 0.7074,
+      "step": 8217
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0533191383189706,
+      "learning_rate": 4.8222798905621445e-06,
+      "loss": 0.6093,
+      "step": 8218
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3912726249043172,
+      "learning_rate": 4.82124406214866e-06,
+      "loss": 0.6868,
+      "step": 8219
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5781178602209105,
+      "learning_rate": 4.820208241416719e-06,
+      "loss": 0.6811,
+      "step": 8220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5179846235906496,
+      "learning_rate": 4.819172428410826e-06,
+      "loss": 0.6988,
+      "step": 8221
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7864417381059368,
+      "learning_rate": 4.818136623175498e-06,
+      "loss": 0.8762,
+      "step": 8222
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.486830245575246,
+      "learning_rate": 4.817100825755244e-06,
+      "loss": 0.7344,
+      "step": 8223
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1120348528258637,
+      "learning_rate": 4.816065036194576e-06,
+      "loss": 0.7303,
+      "step": 8224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9067360158205635,
+      "learning_rate": 4.815029254538003e-06,
+      "loss": 0.7655,
+      "step": 8225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.36264084307817,
+      "learning_rate": 4.813993480830034e-06,
+      "loss": 0.8222,
+      "step": 8226
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7641111423077238,
+      "learning_rate": 4.812957715115181e-06,
+      "loss": 0.7943,
+      "step": 8227
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5360528872971921,
+      "learning_rate": 4.8119219574379495e-06,
+      "loss": 0.6327,
+      "step": 8228
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6348969056658593,
+      "learning_rate": 4.810886207842852e-06,
+      "loss": 0.7486,
+      "step": 8229
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.648369763567878,
+      "learning_rate": 4.809850466374393e-06,
+      "loss": 0.7988,
+      "step": 8230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0593678259475126,
+      "learning_rate": 4.8088147330770855e-06,
+      "loss": 0.6787,
+      "step": 8231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.699481661533193,
+      "learning_rate": 4.807779007995434e-06,
+      "loss": 0.8165,
+      "step": 8232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4664334340558458,
+      "learning_rate": 4.8067432911739455e-06,
+      "loss": 0.5665,
+      "step": 8233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.685745871197162,
+      "learning_rate": 4.80570758265713e-06,
+      "loss": 0.6458,
+      "step": 8234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0229897352013075,
+      "learning_rate": 4.804671882489491e-06,
+      "loss": 0.5611,
+      "step": 8235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.159923202297731,
+      "learning_rate": 4.803636190715536e-06,
+      "loss": 0.6436,
+      "step": 8236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2736514248093918,
+      "learning_rate": 4.802600507379771e-06,
+      "loss": 0.6259,
+      "step": 8237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8154620710299756,
+      "learning_rate": 4.801564832526704e-06,
+      "loss": 0.8237,
+      "step": 8238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9080803121106444,
+      "learning_rate": 4.800529166200837e-06,
+      "loss": 0.8105,
+      "step": 8239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0257800920083773,
+      "learning_rate": 4.7994935084466745e-06,
+      "loss": 0.5137,
+      "step": 8240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.105097471461232,
+      "learning_rate": 4.798457859308724e-06,
+      "loss": 0.8081,
+      "step": 8241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5724860828770075,
+      "learning_rate": 4.797422218831488e-06,
+      "loss": 0.7154,
+      "step": 8242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.689879042102217,
+      "learning_rate": 4.796386587059471e-06,
+      "loss": 0.708,
+      "step": 8243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.244165372040904,
+      "learning_rate": 4.795350964037174e-06,
+      "loss": 0.7441,
+      "step": 8244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5559864349564103,
+      "learning_rate": 4.794315349809105e-06,
+      "loss": 0.8129,
+      "step": 8245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.437209201096035,
+      "learning_rate": 4.7932797444197604e-06,
+      "loss": 0.7643,
+      "step": 8246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7007303721087697,
+      "learning_rate": 4.792244147913647e-06,
+      "loss": 0.6866,
+      "step": 8247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9319799824721495,
+      "learning_rate": 4.791208560335264e-06,
+      "loss": 0.679,
+      "step": 8248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.533487251437475,
+      "learning_rate": 4.790172981729116e-06,
+      "loss": 0.6232,
+      "step": 8249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3963082748702345,
+      "learning_rate": 4.789137412139701e-06,
+      "loss": 0.6476,
+      "step": 8250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.057552127279159,
+      "learning_rate": 4.7881018516115205e-06,
+      "loss": 0.7387,
+      "step": 8251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6367487082686523,
+      "learning_rate": 4.787066300189077e-06,
+      "loss": 0.7061,
+      "step": 8252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6746169586054545,
+      "learning_rate": 4.786030757916868e-06,
+      "loss": 0.7432,
+      "step": 8253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6715342810445786,
+      "learning_rate": 4.784995224839394e-06,
+      "loss": 0.7013,
+      "step": 8254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5522872564329153,
+      "learning_rate": 4.783959701001153e-06,
+      "loss": 0.7363,
+      "step": 8255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6352614468848234,
+      "learning_rate": 4.782924186446647e-06,
+      "loss": 0.757,
+      "step": 8256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9595700891174312,
+      "learning_rate": 4.7818886812203715e-06,
+      "loss": 0.6461,
+      "step": 8257
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.786152501594755,
+      "learning_rate": 4.780853185366823e-06,
+      "loss": 0.6903,
+      "step": 8258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5535182108651193,
+      "learning_rate": 4.779817698930502e-06,
+      "loss": 0.6905,
+      "step": 8259
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.728112572891154,
+      "learning_rate": 4.778782221955907e-06,
+      "loss": 0.8477,
+      "step": 8260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.308889915049971,
+      "learning_rate": 4.77774675448753e-06,
+      "loss": 0.5971,
+      "step": 8261
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8463106547503678,
+      "learning_rate": 4.776711296569872e-06,
+      "loss": 0.7596,
+      "step": 8262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.662202320793308,
+      "learning_rate": 4.775675848247427e-06,
+      "loss": 0.7897,
+      "step": 8263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6577286399643816,
+      "learning_rate": 4.774640409564688e-06,
+      "loss": 0.6638,
+      "step": 8264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6919305602269958,
+      "learning_rate": 4.773604980566154e-06,
+      "loss": 0.7139,
+      "step": 8265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6742369114140498,
+      "learning_rate": 4.772569561296318e-06,
+      "loss": 0.821,
+      "step": 8266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6450055663673147,
+      "learning_rate": 4.771534151799676e-06,
+      "loss": 0.7679,
+      "step": 8267
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.430452647396058,
+      "learning_rate": 4.770498752120718e-06,
+      "loss": 0.7105,
+      "step": 8268
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4201782226235364,
+      "learning_rate": 4.76946336230394e-06,
+      "loss": 0.7074,
+      "step": 8269
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8565633233501537,
+      "learning_rate": 4.768427982393836e-06,
+      "loss": 0.7992,
+      "step": 8270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8632268746581981,
+      "learning_rate": 4.767392612434897e-06,
+      "loss": 0.7381,
+      "step": 8271
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5327031905235753,
+      "learning_rate": 4.766357252471615e-06,
+      "loss": 0.6216,
+      "step": 8272
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6570601820332806,
+      "learning_rate": 4.765321902548482e-06,
+      "loss": 0.6904,
+      "step": 8273
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4827823749365137,
+      "learning_rate": 4.76428656270999e-06,
+      "loss": 0.7735,
+      "step": 8274
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7394720535764414,
+      "learning_rate": 4.7632512330006296e-06,
+      "loss": 0.7213,
+      "step": 8275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.424530294291724,
+      "learning_rate": 4.76221591346489e-06,
+      "loss": 0.7941,
+      "step": 8276
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7334156252015227,
+      "learning_rate": 4.761180604147262e-06,
+      "loss": 0.7871,
+      "step": 8277
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4700565087043984,
+      "learning_rate": 4.760145305092238e-06,
+      "loss": 0.7197,
+      "step": 8278
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8412082436029555,
+      "learning_rate": 4.759110016344302e-06,
+      "loss": 0.7738,
+      "step": 8279
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4915437044824285,
+      "learning_rate": 4.7580747379479455e-06,
+      "loss": 0.7969,
+      "step": 8280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3498816091165637,
+      "learning_rate": 4.757039469947658e-06,
+      "loss": 0.7837,
+      "step": 8281
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5997969296531744,
+      "learning_rate": 4.756004212387923e-06,
+      "loss": 0.7093,
+      "step": 8282
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.33483674743471,
+      "learning_rate": 4.7549689653132304e-06,
+      "loss": 0.5878,
+      "step": 8283
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.282751952033538,
+      "learning_rate": 4.753933728768069e-06,
+      "loss": 0.6976,
+      "step": 8284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4952266833484806,
+      "learning_rate": 4.752898502796922e-06,
+      "loss": 0.8156,
+      "step": 8285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.621819262929022,
+      "learning_rate": 4.751863287444278e-06,
+      "loss": 0.693,
+      "step": 8286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.409050232651201,
+      "learning_rate": 4.750828082754619e-06,
+      "loss": 0.758,
+      "step": 8287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.611956336066396,
+      "learning_rate": 4.7497928887724325e-06,
+      "loss": 0.7745,
+      "step": 8288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.566838624654462,
+      "learning_rate": 4.748757705542205e-06,
+      "loss": 0.6455,
+      "step": 8289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2803447356713145,
+      "learning_rate": 4.7477225331084174e-06,
+      "loss": 0.7393,
+      "step": 8290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7646241345724638,
+      "learning_rate": 4.746687371515554e-06,
+      "loss": 0.7053,
+      "step": 8291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1927115652860225,
+      "learning_rate": 4.7456522208081e-06,
+      "loss": 0.8764,
+      "step": 8292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8222396200213793,
+      "learning_rate": 4.7446170810305346e-06,
+      "loss": 0.8305,
+      "step": 8293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5036048075985404,
+      "learning_rate": 4.743581952227342e-06,
+      "loss": 0.7645,
+      "step": 8294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4781877950853626,
+      "learning_rate": 4.7425468344430035e-06,
+      "loss": 0.8321,
+      "step": 8295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0054996797753393,
+      "learning_rate": 4.741511727722003e-06,
+      "loss": 0.7494,
+      "step": 8296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.511930772926412,
+      "learning_rate": 4.740476632108818e-06,
+      "loss": 0.6708,
+      "step": 8297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7619065908266747,
+      "learning_rate": 4.73944154764793e-06,
+      "loss": 0.6354,
+      "step": 8298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8712325089023938,
+      "learning_rate": 4.73840647438382e-06,
+      "loss": 0.732,
+      "step": 8299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6627430494982078,
+      "learning_rate": 4.737371412360966e-06,
+      "loss": 0.6596,
+      "step": 8300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.631164816955673,
+      "learning_rate": 4.7363363616238465e-06,
+      "loss": 0.8291,
+      "step": 8301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5332993509592783,
+      "learning_rate": 4.73530132221694e-06,
+      "loss": 0.6822,
+      "step": 8302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1924823882895743,
+      "learning_rate": 4.734266294184728e-06,
+      "loss": 0.7136,
+      "step": 8303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6010645277965603,
+      "learning_rate": 4.733231277571683e-06,
+      "loss": 0.7908,
+      "step": 8304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2387497698364673,
+      "learning_rate": 4.732196272422285e-06,
+      "loss": 0.7927,
+      "step": 8305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.679838214325707,
+      "learning_rate": 4.731161278781008e-06,
+      "loss": 0.7237,
+      "step": 8306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1260160392107057,
+      "learning_rate": 4.730126296692332e-06,
+      "loss": 0.6089,
+      "step": 8307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5983028097051541,
+      "learning_rate": 4.72909132620073e-06,
+      "loss": 0.8064,
+      "step": 8308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4461046787181364,
+      "learning_rate": 4.7280563673506745e-06,
+      "loss": 0.7573,
+      "step": 8309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6841840953869343,
+      "learning_rate": 4.727021420186646e-06,
+      "loss": 0.7559,
+      "step": 8310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5346008662216195,
+      "learning_rate": 4.725986484753112e-06,
+      "loss": 0.6201,
+      "step": 8311
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.419340720953096,
+      "learning_rate": 4.72495156109455e-06,
+      "loss": 0.7602,
+      "step": 8312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7288679283473434,
+      "learning_rate": 4.723916649255432e-06,
+      "loss": 0.6907,
+      "step": 8313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.999545000041693,
+      "learning_rate": 4.722881749280232e-06,
+      "loss": 0.6915,
+      "step": 8314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1202922718059996,
+      "learning_rate": 4.7218468612134175e-06,
+      "loss": 0.5966,
+      "step": 8315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5512432908694505,
+      "learning_rate": 4.720811985099464e-06,
+      "loss": 0.7515,
+      "step": 8316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4328618877624288,
+      "learning_rate": 4.719777120982843e-06,
+      "loss": 0.8116,
+      "step": 8317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7565661563526906,
+      "learning_rate": 4.718742268908022e-06,
+      "loss": 0.7766,
+      "step": 8318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7604169458199284,
+      "learning_rate": 4.717707428919471e-06,
+      "loss": 0.7215,
+      "step": 8319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1751884758607014,
+      "learning_rate": 4.716672601061661e-06,
+      "loss": 0.5988,
+      "step": 8320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9682215014316453,
+      "learning_rate": 4.715637785379062e-06,
+      "loss": 0.6259,
+      "step": 8321
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3387535815082927,
+      "learning_rate": 4.714602981916139e-06,
+      "loss": 0.8025,
+      "step": 8322
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.646080680181077,
+      "learning_rate": 4.713568190717362e-06,
+      "loss": 0.7536,
+      "step": 8323
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.422866308446287,
+      "learning_rate": 4.712533411827197e-06,
+      "loss": 0.7098,
+      "step": 8324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1738032733772485,
+      "learning_rate": 4.711498645290113e-06,
+      "loss": 0.6834,
+      "step": 8325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4944370771416002,
+      "learning_rate": 4.710463891150573e-06,
+      "loss": 0.6993,
+      "step": 8326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.582362756156884,
+      "learning_rate": 4.709429149453046e-06,
+      "loss": 0.662,
+      "step": 8327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5874563877001695,
+      "learning_rate": 4.708394420241996e-06,
+      "loss": 0.8303,
+      "step": 8328
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9844950136689279,
+      "learning_rate": 4.707359703561885e-06,
+      "loss": 0.6797,
+      "step": 8329
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1764629620510232,
+      "learning_rate": 4.70632499945718e-06,
+      "loss": 0.6236,
+      "step": 8330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.718621817343712,
+      "learning_rate": 4.705290307972344e-06,
+      "loss": 0.7239,
+      "step": 8331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6558950444048375,
+      "learning_rate": 4.7042556291518415e-06,
+      "loss": 0.7785,
+      "step": 8332
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.666124412803934,
+      "learning_rate": 4.703220963040131e-06,
+      "loss": 0.7828,
+      "step": 8333
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.845216794737158,
+      "learning_rate": 4.702186309681677e-06,
+      "loss": 0.8489,
+      "step": 8334
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8344668351791578,
+      "learning_rate": 4.701151669120942e-06,
+      "loss": 0.7033,
+      "step": 8335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6858162392957587,
+      "learning_rate": 4.700117041402384e-06,
+      "loss": 0.6275,
+      "step": 8336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5890169466074027,
+      "learning_rate": 4.699082426570465e-06,
+      "loss": 0.7965,
+      "step": 8337
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5845102226628776,
+      "learning_rate": 4.6980478246696435e-06,
+      "loss": 0.7457,
+      "step": 8338
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.05124697860731,
+      "learning_rate": 4.697013235744382e-06,
+      "loss": 0.5863,
+      "step": 8339
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3718786642525043,
+      "learning_rate": 4.695978659839133e-06,
+      "loss": 0.7067,
+      "step": 8340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.601257130923925,
+      "learning_rate": 4.694944096998361e-06,
+      "loss": 0.6745,
+      "step": 8341
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1253890833125741,
+      "learning_rate": 4.693909547266518e-06,
+      "loss": 0.6428,
+      "step": 8342
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5322297143236436,
+      "learning_rate": 4.692875010688066e-06,
+      "loss": 0.6454,
+      "step": 8343
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.60767830667497,
+      "learning_rate": 4.6918404873074574e-06,
+      "loss": 0.7898,
+      "step": 8344
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.563462830669679,
+      "learning_rate": 4.69080597716915e-06,
+      "loss": 0.7408,
+      "step": 8345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6617946964433377,
+      "learning_rate": 4.6897714803175995e-06,
+      "loss": 0.6259,
+      "step": 8346
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6078773590027813,
+      "learning_rate": 4.688736996797257e-06,
+      "loss": 0.7892,
+      "step": 8347
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5425130980754114,
+      "learning_rate": 4.687702526652579e-06,
+      "loss": 0.7518,
+      "step": 8348
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.599789628650717,
+      "learning_rate": 4.68666806992802e-06,
+      "loss": 0.6578,
+      "step": 8349
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6868904877427637,
+      "learning_rate": 4.685633626668032e-06,
+      "loss": 0.685,
+      "step": 8350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.099065576581352,
+      "learning_rate": 4.684599196917067e-06,
+      "loss": 0.7749,
+      "step": 8351
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1420816330861012,
+      "learning_rate": 4.683564780719576e-06,
+      "loss": 0.6589,
+      "step": 8352
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0774357912112416,
+      "learning_rate": 4.682530378120014e-06,
+      "loss": 0.571,
+      "step": 8353
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6393777467624728,
+      "learning_rate": 4.681495989162826e-06,
+      "loss": 0.7459,
+      "step": 8354
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0229642023258358,
+      "learning_rate": 4.680461613892465e-06,
+      "loss": 0.5906,
+      "step": 8355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7731735571559792,
+      "learning_rate": 4.679427252353379e-06,
+      "loss": 0.6562,
+      "step": 8356
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5115124131572268,
+      "learning_rate": 4.678392904590021e-06,
+      "loss": 0.6153,
+      "step": 8357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0046623752838175,
+      "learning_rate": 4.677358570646834e-06,
+      "loss": 0.691,
+      "step": 8358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.954535245245959,
+      "learning_rate": 4.676324250568269e-06,
+      "loss": 0.7816,
+      "step": 8359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9665148784801456,
+      "learning_rate": 4.6752899443987694e-06,
+      "loss": 0.7343,
+      "step": 8360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.042268542916504,
+      "learning_rate": 4.674255652182788e-06,
+      "loss": 0.6583,
+      "step": 8361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.153033908389317,
+      "learning_rate": 4.673221373964764e-06,
+      "loss": 0.7273,
+      "step": 8362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3953458436259645,
+      "learning_rate": 4.672187109789144e-06,
+      "loss": 0.7401,
+      "step": 8363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9724154811976378,
+      "learning_rate": 4.671152859700377e-06,
+      "loss": 0.8776,
+      "step": 8364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0740430981238014,
+      "learning_rate": 4.670118623742904e-06,
+      "loss": 0.5899,
+      "step": 8365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5112027036403237,
+      "learning_rate": 4.669084401961166e-06,
+      "loss": 0.7845,
+      "step": 8366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9495242968605264,
+      "learning_rate": 4.668050194399609e-06,
+      "loss": 0.814,
+      "step": 8367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6687631345470808,
+      "learning_rate": 4.667016001102675e-06,
+      "loss": 0.6868,
+      "step": 8368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7335738477001987,
+      "learning_rate": 4.665981822114805e-06,
+      "loss": 0.689,
+      "step": 8369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.62171409818517,
+      "learning_rate": 4.66494765748044e-06,
+      "loss": 0.6704,
+      "step": 8370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5592274098814438,
+      "learning_rate": 4.6639135072440195e-06,
+      "loss": 0.6998,
+      "step": 8371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.552850065856012,
+      "learning_rate": 4.662879371449987e-06,
+      "loss": 0.7136,
+      "step": 8372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3776718952094884,
+      "learning_rate": 4.6618452501427755e-06,
+      "loss": 0.7286,
+      "step": 8373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5907873341309517,
+      "learning_rate": 4.660811143366828e-06,
+      "loss": 0.7373,
+      "step": 8374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8316977904245786,
+      "learning_rate": 4.6597770511665814e-06,
+      "loss": 0.7642,
+      "step": 8375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5124717412348694,
+      "learning_rate": 4.658742973586471e-06,
+      "loss": 0.7024,
+      "step": 8376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.797611623377723,
+      "learning_rate": 4.657708910670936e-06,
+      "loss": 0.7213,
+      "step": 8377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7000448252920355,
+      "learning_rate": 4.656674862464412e-06,
+      "loss": 0.6614,
+      "step": 8378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6787077845286353,
+      "learning_rate": 4.655640829011335e-06,
+      "loss": 0.6392,
+      "step": 8379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1667022392829693,
+      "learning_rate": 4.654606810356135e-06,
+      "loss": 0.672,
+      "step": 8380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.589643508106399,
+      "learning_rate": 4.653572806543251e-06,
+      "loss": 0.7553,
+      "step": 8381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5547801700923423,
+      "learning_rate": 4.652538817617117e-06,
+      "loss": 0.6636,
+      "step": 8382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7223674116495842,
+      "learning_rate": 4.651504843622163e-06,
+      "loss": 0.7678,
+      "step": 8383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6840175677795217,
+      "learning_rate": 4.65047088460282e-06,
+      "loss": 0.6757,
+      "step": 8384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.327673936175427,
+      "learning_rate": 4.6494369406035225e-06,
+      "loss": 0.6238,
+      "step": 8385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6452243589719604,
+      "learning_rate": 4.6484030116687014e-06,
+      "loss": 0.7452,
+      "step": 8386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5607656398044853,
+      "learning_rate": 4.647369097842785e-06,
+      "loss": 0.703,
+      "step": 8387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1398510972764044,
+      "learning_rate": 4.646335199170205e-06,
+      "loss": 0.7064,
+      "step": 8388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.767825211631413,
+      "learning_rate": 4.645301315695387e-06,
+      "loss": 0.7371,
+      "step": 8389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1753203966657353,
+      "learning_rate": 4.6442674474627645e-06,
+      "loss": 0.7165,
+      "step": 8390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4854118059770594,
+      "learning_rate": 4.643233594516759e-06,
+      "loss": 0.7562,
+      "step": 8391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8638506361684513,
+      "learning_rate": 4.642199756901802e-06,
+      "loss": 0.7844,
+      "step": 8392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.41716408471408,
+      "learning_rate": 4.64116593466232e-06,
+      "loss": 0.6998,
+      "step": 8393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6859936552150552,
+      "learning_rate": 4.6401321278427334e-06,
+      "loss": 0.7772,
+      "step": 8394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5459542711326504,
+      "learning_rate": 4.639098336487472e-06,
+      "loss": 0.6971,
+      "step": 8395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.094134846205592,
+      "learning_rate": 4.638064560640959e-06,
+      "loss": 0.6516,
+      "step": 8396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4858332012490252,
+      "learning_rate": 4.63703080034762e-06,
+      "loss": 0.8459,
+      "step": 8397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0631991033813835,
+      "learning_rate": 4.635997055651873e-06,
+      "loss": 0.7818,
+      "step": 8398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9194612426731064,
+      "learning_rate": 4.634963326598143e-06,
+      "loss": 0.7369,
+      "step": 8399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4612645207908086,
+      "learning_rate": 4.633929613230855e-06,
+      "loss": 0.8341,
+      "step": 8400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.176954067349415,
+      "learning_rate": 4.632895915594424e-06,
+      "loss": 0.636,
+      "step": 8401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7917324051370753,
+      "learning_rate": 4.631862233733274e-06,
+      "loss": 0.6808,
+      "step": 8402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.659446434347785,
+      "learning_rate": 4.630828567691823e-06,
+      "loss": 0.8014,
+      "step": 8403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6221839895942303,
+      "learning_rate": 4.629794917514492e-06,
+      "loss": 0.802,
+      "step": 8404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8201333416531136,
+      "learning_rate": 4.628761283245697e-06,
+      "loss": 0.737,
+      "step": 8405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.472819278701895,
+      "learning_rate": 4.627727664929856e-06,
+      "loss": 0.7733,
+      "step": 8406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7303007593216206,
+      "learning_rate": 4.626694062611387e-06,
+      "loss": 0.7494,
+      "step": 8407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3419039102423276,
+      "learning_rate": 4.6256604763347066e-06,
+      "loss": 0.7719,
+      "step": 8408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8581066561260728,
+      "learning_rate": 4.624626906144227e-06,
+      "loss": 0.7784,
+      "step": 8409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.126278367723696,
+      "learning_rate": 4.623593352084367e-06,
+      "loss": 0.7462,
+      "step": 8410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6227132840535117,
+      "learning_rate": 4.622559814199538e-06,
+      "loss": 0.7393,
+      "step": 8411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4425833848145595,
+      "learning_rate": 4.621526292534156e-06,
+      "loss": 0.7467,
+      "step": 8412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.760229724800663,
+      "learning_rate": 4.6204927871326295e-06,
+      "loss": 0.7574,
+      "step": 8413
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6930933984222123,
+      "learning_rate": 4.619459298039373e-06,
+      "loss": 0.7493,
+      "step": 8414
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5561418114758785,
+      "learning_rate": 4.6184258252988016e-06,
+      "loss": 0.7285,
+      "step": 8415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6510398982169225,
+      "learning_rate": 4.6173923689553205e-06,
+      "loss": 0.7255,
+      "step": 8416
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8129757536883813,
+      "learning_rate": 4.616358929053339e-06,
+      "loss": 0.8048,
+      "step": 8417
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6929873857812594,
+      "learning_rate": 4.615325505637272e-06,
+      "loss": 0.7636,
+      "step": 8418
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5964080148616508,
+      "learning_rate": 4.614292098751524e-06,
+      "loss": 0.8126,
+      "step": 8419
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7338558844113618,
+      "learning_rate": 4.613258708440503e-06,
+      "loss": 0.6687,
+      "step": 8420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6434720396164835,
+      "learning_rate": 4.612225334748616e-06,
+      "loss": 0.758,
+      "step": 8421
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5855422197620779,
+      "learning_rate": 4.611191977720272e-06,
+      "loss": 0.7187,
+      "step": 8422
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8794545096914832,
+      "learning_rate": 4.610158637399872e-06,
+      "loss": 0.6823,
+      "step": 8423
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0038888025803745,
+      "learning_rate": 4.609125313831826e-06,
+      "loss": 0.6288,
+      "step": 8424
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7493134184066463,
+      "learning_rate": 4.608092007060533e-06,
+      "loss": 0.7395,
+      "step": 8425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7903053949473542,
+      "learning_rate": 4.607058717130403e-06,
+      "loss": 0.7093,
+      "step": 8426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.146013468976803,
+      "learning_rate": 4.6060254440858315e-06,
+      "loss": 0.5776,
+      "step": 8427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5762806885304463,
+      "learning_rate": 4.6049921879712254e-06,
+      "loss": 0.7158,
+      "step": 8428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1722595244218152,
+      "learning_rate": 4.603958948830985e-06,
+      "loss": 0.6065,
+      "step": 8429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2128906140956959,
+      "learning_rate": 4.602925726709512e-06,
+      "loss": 0.6963,
+      "step": 8430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7432948542995457,
+      "learning_rate": 4.6018925216512025e-06,
+      "loss": 0.7545,
+      "step": 8431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7411303571060361,
+      "learning_rate": 4.600859333700457e-06,
+      "loss": 0.7539,
+      "step": 8432
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7061220803588042,
+      "learning_rate": 4.599826162901679e-06,
+      "loss": 0.6586,
+      "step": 8433
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4969063748665599,
+      "learning_rate": 4.5987930092992596e-06,
+      "loss": 0.715,
+      "step": 8434
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2040325217564438,
+      "learning_rate": 4.597759872937597e-06,
+      "loss": 0.5858,
+      "step": 8435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8780049818869862,
+      "learning_rate": 4.5967267538610915e-06,
+      "loss": 0.7631,
+      "step": 8436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7430064273624335,
+      "learning_rate": 4.595693652114133e-06,
+      "loss": 0.7404,
+      "step": 8437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.592662654981369,
+      "learning_rate": 4.594660567741118e-06,
+      "loss": 0.7328,
+      "step": 8438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6389392889657968,
+      "learning_rate": 4.593627500786444e-06,
+      "loss": 0.8472,
+      "step": 8439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7065703929877953,
+      "learning_rate": 4.592594451294501e-06,
+      "loss": 0.7312,
+      "step": 8440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.564719462906776,
+      "learning_rate": 4.59156141930968e-06,
+      "loss": 0.6374,
+      "step": 8441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7605728962294809,
+      "learning_rate": 4.590528404876374e-06,
+      "loss": 0.7368,
+      "step": 8442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7216035926418212,
+      "learning_rate": 4.5894954080389755e-06,
+      "loss": 0.7057,
+      "step": 8443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1069106796837374,
+      "learning_rate": 4.588462428841875e-06,
+      "loss": 0.8077,
+      "step": 8444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5140367200047002,
+      "learning_rate": 4.587429467329458e-06,
+      "loss": 0.8239,
+      "step": 8445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6615281783389788,
+      "learning_rate": 4.586396523546116e-06,
+      "loss": 0.7563,
+      "step": 8446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8820370446307195,
+      "learning_rate": 4.585363597536239e-06,
+      "loss": 0.7743,
+      "step": 8447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6664826274981432,
+      "learning_rate": 4.584330689344211e-06,
+      "loss": 0.823,
+      "step": 8448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7885533373603824,
+      "learning_rate": 4.5832977990144165e-06,
+      "loss": 0.5683,
+      "step": 8449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8068091560227035,
+      "learning_rate": 4.582264926591244e-06,
+      "loss": 0.6753,
+      "step": 8450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0093067137749163,
+      "learning_rate": 4.581232072119081e-06,
+      "loss": 0.6287,
+      "step": 8451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0505112188903487,
+      "learning_rate": 4.580199235642306e-06,
+      "loss": 0.7828,
+      "step": 8452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6694510277972294,
+      "learning_rate": 4.5791664172053044e-06,
+      "loss": 0.6307,
+      "step": 8453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5188315479323713,
+      "learning_rate": 4.578133616852462e-06,
+      "loss": 0.6308,
+      "step": 8454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8759600990685157,
+      "learning_rate": 4.577100834628155e-06,
+      "loss": 0.638,
+      "step": 8455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.059369649694723,
+      "learning_rate": 4.5760680705767665e-06,
+      "loss": 0.7302,
+      "step": 8456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4618334627592344,
+      "learning_rate": 4.5750353247426785e-06,
+      "loss": 0.6634,
+      "step": 8457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6540581407127115,
+      "learning_rate": 4.5740025971702695e-06,
+      "loss": 0.7399,
+      "step": 8458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2950034043604486,
+      "learning_rate": 4.572969887903916e-06,
+      "loss": 0.6207,
+      "step": 8459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7053220080738338,
+      "learning_rate": 4.571937196987998e-06,
+      "loss": 0.8139,
+      "step": 8460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.493148520761759,
+      "learning_rate": 4.570904524466893e-06,
+      "loss": 0.6541,
+      "step": 8461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.609203590072895,
+      "learning_rate": 4.5698718703849755e-06,
+      "loss": 0.6964,
+      "step": 8462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8067858962575447,
+      "learning_rate": 4.5688392347866226e-06,
+      "loss": 0.6514,
+      "step": 8463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7396848275861443,
+      "learning_rate": 4.5678066177162065e-06,
+      "loss": 0.8063,
+      "step": 8464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0865614426473975,
+      "learning_rate": 4.566774019218104e-06,
+      "loss": 0.5749,
+      "step": 8465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5178282093934465,
+      "learning_rate": 4.565741439336686e-06,
+      "loss": 0.6712,
+      "step": 8466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3066647584299231,
+      "learning_rate": 4.5647088781163255e-06,
+      "loss": 0.7106,
+      "step": 8467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6195678521379935,
+      "learning_rate": 4.563676335601393e-06,
+      "loss": 0.8029,
+      "step": 8468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.498027317882548,
+      "learning_rate": 4.562643811836263e-06,
+      "loss": 0.7524,
+      "step": 8469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8476229792530956,
+      "learning_rate": 4.561611306865299e-06,
+      "loss": 0.7594,
+      "step": 8470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7432721056990073,
+      "learning_rate": 4.560578820732876e-06,
+      "loss": 0.6519,
+      "step": 8471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0960831661809811,
+      "learning_rate": 4.559546353483359e-06,
+      "loss": 0.6351,
+      "step": 8472
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9327498241227756,
+      "learning_rate": 4.558513905161116e-06,
+      "loss": 0.6685,
+      "step": 8473
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6863117825838703,
+      "learning_rate": 4.557481475810512e-06,
+      "loss": 0.7223,
+      "step": 8474
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7353276116749752,
+      "learning_rate": 4.556449065475917e-06,
+      "loss": 0.778,
+      "step": 8475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0485307384791416,
+      "learning_rate": 4.555416674201693e-06,
+      "loss": 0.584,
+      "step": 8476
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7604149260928053,
+      "learning_rate": 4.554384302032204e-06,
+      "loss": 0.7632,
+      "step": 8477
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.770483641170293,
+      "learning_rate": 4.553351949011814e-06,
+      "loss": 0.7269,
+      "step": 8478
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5691914684144235,
+      "learning_rate": 4.5523196151848846e-06,
+      "loss": 0.6954,
+      "step": 8479
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8732448057025055,
+      "learning_rate": 4.551287300595781e-06,
+      "loss": 0.6928,
+      "step": 8480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7662184209999345,
+      "learning_rate": 4.550255005288861e-06,
+      "loss": 0.6867,
+      "step": 8481
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5772478923387927,
+      "learning_rate": 4.549222729308483e-06,
+      "loss": 0.6933,
+      "step": 8482
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5648788997352927,
+      "learning_rate": 4.548190472699011e-06,
+      "loss": 0.8273,
+      "step": 8483
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6871229677077113,
+      "learning_rate": 4.547158235504797e-06,
+      "loss": 0.7248,
+      "step": 8484
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5789784691932807,
+      "learning_rate": 4.546126017770205e-06,
+      "loss": 0.7173,
+      "step": 8485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2369260901832158,
+      "learning_rate": 4.5450938195395875e-06,
+      "loss": 0.6448,
+      "step": 8486
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.112451725512762,
+      "learning_rate": 4.544061640857303e-06,
+      "loss": 0.8761,
+      "step": 8487
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7331298791960497,
+      "learning_rate": 4.543029481767703e-06,
+      "loss": 0.7551,
+      "step": 8488
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.850750418065733,
+      "learning_rate": 4.541997342315145e-06,
+      "loss": 0.7239,
+      "step": 8489
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.616639119335143,
+      "learning_rate": 4.540965222543981e-06,
+      "loss": 0.9067,
+      "step": 8490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5253271758960374,
+      "learning_rate": 4.539933122498566e-06,
+      "loss": 0.6969,
+      "step": 8491
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.78760167525411,
+      "learning_rate": 4.5389010422232474e-06,
+      "loss": 0.755,
+      "step": 8492
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6020541734488365,
+      "learning_rate": 4.5378689817623765e-06,
+      "loss": 0.6927,
+      "step": 8493
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7801685039161035,
+      "learning_rate": 4.536836941160308e-06,
+      "loss": 0.759,
+      "step": 8494
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.666722042402242,
+      "learning_rate": 4.535804920461386e-06,
+      "loss": 0.8296,
+      "step": 8495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3017506816072713,
+      "learning_rate": 4.53477291970996e-06,
+      "loss": 0.6991,
+      "step": 8496
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6285195286633196,
+      "learning_rate": 4.5337409389503764e-06,
+      "loss": 0.7786,
+      "step": 8497
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8037526213771478,
+      "learning_rate": 4.532708978226987e-06,
+      "loss": 0.716,
+      "step": 8498
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2109230831660898,
+      "learning_rate": 4.5316770375841315e-06,
+      "loss": 0.6487,
+      "step": 8499
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8403637131984416,
+      "learning_rate": 4.530645117066155e-06,
+      "loss": 0.7054,
+      "step": 8500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5945410278079888,
+      "learning_rate": 4.529613216717406e-06,
+      "loss": 0.6762,
+      "step": 8501
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.108473721419886,
+      "learning_rate": 4.528581336582223e-06,
+      "loss": 0.6879,
+      "step": 8502
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7275672506118838,
+      "learning_rate": 4.527549476704949e-06,
+      "loss": 0.6932,
+      "step": 8503
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5758028770044907,
+      "learning_rate": 4.526517637129927e-06,
+      "loss": 0.7084,
+      "step": 8504
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.771413331379663,
+      "learning_rate": 4.525485817901499e-06,
+      "loss": 0.7867,
+      "step": 8505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.066244292157976,
+      "learning_rate": 4.524454019063999e-06,
+      "loss": 0.6691,
+      "step": 8506
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0896036635866357,
+      "learning_rate": 4.52342224066177e-06,
+      "loss": 0.8056,
+      "step": 8507
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.168638031447497,
+      "learning_rate": 4.522390482739148e-06,
+      "loss": 0.7428,
+      "step": 8508
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8344490211739444,
+      "learning_rate": 4.5213587453404736e-06,
+      "loss": 0.8397,
+      "step": 8509
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.498377964127965,
+      "learning_rate": 4.520327028510076e-06,
+      "loss": 0.6841,
+      "step": 8510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6342883325580453,
+      "learning_rate": 4.5192953322922955e-06,
+      "loss": 0.6853,
+      "step": 8511
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8388771248633287,
+      "learning_rate": 4.518263656731468e-06,
+      "loss": 0.7222,
+      "step": 8512
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5850332335960207,
+      "learning_rate": 4.5172320018719205e-06,
+      "loss": 0.7494,
+      "step": 8513
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4608826434001467,
+      "learning_rate": 4.5162003677579905e-06,
+      "loss": 0.6711,
+      "step": 8514
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8188102339172978,
+      "learning_rate": 4.5151687544340065e-06,
+      "loss": 0.7032,
+      "step": 8515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8969423353890068,
+      "learning_rate": 4.514137161944304e-06,
+      "loss": 0.6859,
+      "step": 8516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.62364956944679,
+      "learning_rate": 4.513105590333207e-06,
+      "loss": 0.7721,
+      "step": 8517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5544187706235073,
+      "learning_rate": 4.512074039645049e-06,
+      "loss": 0.5979,
+      "step": 8518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6096994550181984,
+      "learning_rate": 4.511042509924157e-06,
+      "loss": 0.6332,
+      "step": 8519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5195989522224207,
+      "learning_rate": 4.5100110012148546e-06,
+      "loss": 0.7292,
+      "step": 8520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8885423295897157,
+      "learning_rate": 4.508979513561471e-06,
+      "loss": 0.6954,
+      "step": 8521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9165690149248766,
+      "learning_rate": 4.507948047008332e-06,
+      "loss": 0.6359,
+      "step": 8522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6237698783452756,
+      "learning_rate": 4.506916601599763e-06,
+      "loss": 0.6557,
+      "step": 8523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6795542969734152,
+      "learning_rate": 4.505885177380083e-06,
+      "loss": 0.7077,
+      "step": 8524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4646522317657882,
+      "learning_rate": 4.504853774393618e-06,
+      "loss": 0.7396,
+      "step": 8525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9930698791612416,
+      "learning_rate": 4.5038223926846905e-06,
+      "loss": 0.6379,
+      "step": 8526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5656798271946093,
+      "learning_rate": 4.5027910322976186e-06,
+      "loss": 0.6239,
+      "step": 8527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5839086295521931,
+      "learning_rate": 4.501759693276724e-06,
+      "loss": 0.7147,
+      "step": 8528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3810186553416857,
+      "learning_rate": 4.5007283756663245e-06,
+      "loss": 0.7216,
+      "step": 8529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8298553727381148,
+      "learning_rate": 4.49969707951074e-06,
+      "loss": 0.8435,
+      "step": 8530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.984951141218946,
+      "learning_rate": 4.498665804854285e-06,
+      "loss": 0.6138,
+      "step": 8531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5425329865276833,
+      "learning_rate": 4.497634551741277e-06,
+      "loss": 0.7385,
+      "step": 8532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7334975411686637,
+      "learning_rate": 4.49660332021603e-06,
+      "loss": 0.8129,
+      "step": 8533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.931944427961078,
+      "learning_rate": 4.495572110322862e-06,
+      "loss": 0.7657,
+      "step": 8534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1234418662658032,
+      "learning_rate": 4.494540922106082e-06,
+      "loss": 0.8249,
+      "step": 8535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6486803014718958,
+      "learning_rate": 4.4935097556100045e-06,
+      "loss": 0.6608,
+      "step": 8536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.742810575210834,
+      "learning_rate": 4.492478610878942e-06,
+      "loss": 0.6803,
+      "step": 8537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8845047484729316,
+      "learning_rate": 4.491447487957203e-06,
+      "loss": 0.7384,
+      "step": 8538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7959010704028875,
+      "learning_rate": 4.490416386889097e-06,
+      "loss": 0.7631,
+      "step": 8539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0334236973450084,
+      "learning_rate": 4.489385307718934e-06,
+      "loss": 0.6721,
+      "step": 8540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5991625174483184,
+      "learning_rate": 4.488354250491024e-06,
+      "loss": 0.6283,
+      "step": 8541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5001840908513098,
+      "learning_rate": 4.48732321524967e-06,
+      "loss": 0.7115,
+      "step": 8542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4364517391479141,
+      "learning_rate": 4.486292202039178e-06,
+      "loss": 0.6394,
+      "step": 8543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9321591877607773,
+      "learning_rate": 4.485261210903854e-06,
+      "loss": 0.6661,
+      "step": 8544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.807445672287689,
+      "learning_rate": 4.484230241888005e-06,
+      "loss": 0.7302,
+      "step": 8545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7246460504435848,
+      "learning_rate": 4.48319929503593e-06,
+      "loss": 0.8319,
+      "step": 8546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7832138572779128,
+      "learning_rate": 4.482168370391931e-06,
+      "loss": 0.7116,
+      "step": 8547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.67485029563173,
+      "learning_rate": 4.481137468000312e-06,
+      "loss": 0.5745,
+      "step": 8548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2128603735562504,
+      "learning_rate": 4.48010658790537e-06,
+      "loss": 0.7626,
+      "step": 8549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5507957634728524,
+      "learning_rate": 4.479075730151406e-06,
+      "loss": 0.7776,
+      "step": 8550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7531395660208913,
+      "learning_rate": 4.478044894782718e-06,
+      "loss": 0.6773,
+      "step": 8551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2657688215266383,
+      "learning_rate": 4.477014081843605e-06,
+      "loss": 0.618,
+      "step": 8552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9181521021753245,
+      "learning_rate": 4.47598329137836e-06,
+      "loss": 0.7599,
+      "step": 8553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4961118544397929,
+      "learning_rate": 4.47495252343128e-06,
+      "loss": 0.6629,
+      "step": 8554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.462627546000898,
+      "learning_rate": 4.473921778046661e-06,
+      "loss": 0.6656,
+      "step": 8555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8252455795996063,
+      "learning_rate": 4.4728910552687935e-06,
+      "loss": 0.7263,
+      "step": 8556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5954267227106338,
+      "learning_rate": 4.471860355141971e-06,
+      "loss": 0.7338,
+      "step": 8557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.818110697119175,
+      "learning_rate": 4.470829677710485e-06,
+      "loss": 0.7958,
+      "step": 8558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6554581515297027,
+      "learning_rate": 4.469799023018628e-06,
+      "loss": 0.6895,
+      "step": 8559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6366863899111843,
+      "learning_rate": 4.468768391110687e-06,
+      "loss": 0.6207,
+      "step": 8560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.108474451651856,
+      "learning_rate": 4.467737782030951e-06,
+      "loss": 0.7457,
+      "step": 8561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8543737753957967,
+      "learning_rate": 4.466707195823707e-06,
+      "loss": 0.7777,
+      "step": 8562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6013972881447092,
+      "learning_rate": 4.465676632533245e-06,
+      "loss": 0.6899,
+      "step": 8563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.741264126573674,
+      "learning_rate": 4.464646092203846e-06,
+      "loss": 0.7135,
+      "step": 8564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2314551845597306,
+      "learning_rate": 4.463615574879798e-06,
+      "loss": 0.6837,
+      "step": 8565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9822202770678827,
+      "learning_rate": 4.462585080605384e-06,
+      "loss": 0.7469,
+      "step": 8566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6441504492002577,
+      "learning_rate": 4.461554609424884e-06,
+      "loss": 0.7207,
+      "step": 8567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2009118573779975,
+      "learning_rate": 4.460524161382582e-06,
+      "loss": 0.7893,
+      "step": 8568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0176855062591197,
+      "learning_rate": 4.459493736522759e-06,
+      "loss": 0.7425,
+      "step": 8569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.139061854877776,
+      "learning_rate": 4.4584633348896945e-06,
+      "loss": 0.6432,
+      "step": 8570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6750413538334754,
+      "learning_rate": 4.457432956527665e-06,
+      "loss": 0.8433,
+      "step": 8571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5343264223930808,
+      "learning_rate": 4.456402601480949e-06,
+      "loss": 0.7429,
+      "step": 8572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8142906375265575,
+      "learning_rate": 4.455372269793826e-06,
+      "loss": 0.6823,
+      "step": 8573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.642428872663452,
+      "learning_rate": 4.4543419615105685e-06,
+      "loss": 0.693,
+      "step": 8574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6088164171469408,
+      "learning_rate": 4.453311676675453e-06,
+      "loss": 0.6639,
+      "step": 8575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6148695883998077,
+      "learning_rate": 4.452281415332751e-06,
+      "loss": 0.6949,
+      "step": 8576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5138714292045803,
+      "learning_rate": 4.451251177526738e-06,
+      "loss": 0.6697,
+      "step": 8577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5076844322734195,
+      "learning_rate": 4.450220963301683e-06,
+      "loss": 0.7215,
+      "step": 8578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.201995291963954,
+      "learning_rate": 4.449190772701857e-06,
+      "loss": 0.6273,
+      "step": 8579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2114818691568705,
+      "learning_rate": 4.44816060577153e-06,
+      "loss": 0.7384,
+      "step": 8580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8826566535483182,
+      "learning_rate": 4.447130462554974e-06,
+      "loss": 0.718,
+      "step": 8581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0355243140153076,
+      "learning_rate": 4.44610034309645e-06,
+      "loss": 0.7652,
+      "step": 8582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5589109229655216,
+      "learning_rate": 4.4450702474402295e-06,
+      "loss": 0.6381,
+      "step": 8583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7781096132227991,
+      "learning_rate": 4.444040175630577e-06,
+      "loss": 0.8123,
+      "step": 8584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.56012771208772,
+      "learning_rate": 4.443010127711755e-06,
+      "loss": 0.6935,
+      "step": 8585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6964479685817218,
+      "learning_rate": 4.441980103728027e-06,
+      "loss": 0.7452,
+      "step": 8586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9959331368382927,
+      "learning_rate": 4.440950103723658e-06,
+      "loss": 0.8578,
+      "step": 8587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9513938131038482,
+      "learning_rate": 4.439920127742909e-06,
+      "loss": 0.6965,
+      "step": 8588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9149038598030501,
+      "learning_rate": 4.438890175830039e-06,
+      "loss": 0.7656,
+      "step": 8589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5232950869872275,
+      "learning_rate": 4.437860248029307e-06,
+      "loss": 0.8349,
+      "step": 8590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8954530494503405,
+      "learning_rate": 4.4368303443849735e-06,
+      "loss": 0.721,
+      "step": 8591
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6959000609723638,
+      "learning_rate": 4.435800464941292e-06,
+      "loss": 0.7219,
+      "step": 8592
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6072434085534217,
+      "learning_rate": 4.434770609742523e-06,
+      "loss": 0.7165,
+      "step": 8593
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0970762154208002,
+      "learning_rate": 4.433740778832919e-06,
+      "loss": 0.6856,
+      "step": 8594
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.393784888297536,
+      "learning_rate": 4.432710972256737e-06,
+      "loss": 0.7018,
+      "step": 8595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7339978814134407,
+      "learning_rate": 4.431681190058224e-06,
+      "loss": 0.6722,
+      "step": 8596
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.737415725434665,
+      "learning_rate": 4.430651432281639e-06,
+      "loss": 0.817,
+      "step": 8597
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1646736830144993,
+      "learning_rate": 4.429621698971228e-06,
+      "loss": 0.7988,
+      "step": 8598
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7003153849875583,
+      "learning_rate": 4.428591990171246e-06,
+      "loss": 0.6665,
+      "step": 8599
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7025562578653208,
+      "learning_rate": 4.4275623059259355e-06,
+      "loss": 0.837,
+      "step": 8600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4441240897141663,
+      "learning_rate": 4.426532646279548e-06,
+      "loss": 0.7236,
+      "step": 8601
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2341245949579682,
+      "learning_rate": 4.425503011276332e-06,
+      "loss": 0.5327,
+      "step": 8602
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2663668635705796,
+      "learning_rate": 4.42447340096053e-06,
+      "loss": 0.7597,
+      "step": 8603
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7642604598354774,
+      "learning_rate": 4.423443815376387e-06,
+      "loss": 0.7975,
+      "step": 8604
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6019061094268132,
+      "learning_rate": 4.422414254568147e-06,
+      "loss": 0.6622,
+      "step": 8605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2684537839353265,
+      "learning_rate": 4.421384718580055e-06,
+      "loss": 0.6987,
+      "step": 8606
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.723069526431452,
+      "learning_rate": 4.420355207456349e-06,
+      "loss": 0.6907,
+      "step": 8607
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.111308587350219,
+      "learning_rate": 4.4193257212412695e-06,
+      "loss": 0.7035,
+      "step": 8608
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.877352398159395,
+      "learning_rate": 4.41829625997906e-06,
+      "loss": 0.7352,
+      "step": 8609
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.823056442590313,
+      "learning_rate": 4.417266823713953e-06,
+      "loss": 0.776,
+      "step": 8610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.355086871934228,
+      "learning_rate": 4.416237412490189e-06,
+      "loss": 0.796,
+      "step": 8611
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3755718024794286,
+      "learning_rate": 4.415208026352003e-06,
+      "loss": 0.797,
+      "step": 8612
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8745134602683549,
+      "learning_rate": 4.414178665343633e-06,
+      "loss": 0.7595,
+      "step": 8613
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.609155593266224,
+      "learning_rate": 4.413149329509307e-06,
+      "loss": 0.7378,
+      "step": 8614
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9172217076335434,
+      "learning_rate": 4.412120018893263e-06,
+      "loss": 0.7657,
+      "step": 8615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6695823176196605,
+      "learning_rate": 4.411090733539731e-06,
+      "loss": 0.7991,
+      "step": 8616
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8819881522223896,
+      "learning_rate": 4.410061473492943e-06,
+      "loss": 0.7348,
+      "step": 8617
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7147473246747784,
+      "learning_rate": 4.409032238797125e-06,
+      "loss": 0.6664,
+      "step": 8618
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9069344828225583,
+      "learning_rate": 4.4080030294965085e-06,
+      "loss": 0.7603,
+      "step": 8619
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2687075937573602,
+      "learning_rate": 4.406973845635322e-06,
+      "loss": 0.6745,
+      "step": 8620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4530174477598465,
+      "learning_rate": 4.405944687257789e-06,
+      "loss": 0.8074,
+      "step": 8621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5720449515462092,
+      "learning_rate": 4.4049155544081355e-06,
+      "loss": 0.7473,
+      "step": 8622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2591565188813556,
+      "learning_rate": 4.4038864471305845e-06,
+      "loss": 0.7855,
+      "step": 8623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6231999627765368,
+      "learning_rate": 4.402857365469364e-06,
+      "loss": 0.704,
+      "step": 8624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9882714247615303,
+      "learning_rate": 4.40182830946869e-06,
+      "loss": 0.7986,
+      "step": 8625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8431405665964995,
+      "learning_rate": 4.400799279172786e-06,
+      "loss": 0.7846,
+      "step": 8626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9770687113731327,
+      "learning_rate": 4.399770274625871e-06,
+      "loss": 0.7238,
+      "step": 8627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3865017959141386,
+      "learning_rate": 4.3987412958721664e-06,
+      "loss": 0.7146,
+      "step": 8628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.110370742794967,
+      "learning_rate": 4.397712342955885e-06,
+      "loss": 0.813,
+      "step": 8629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5564498466135226,
+      "learning_rate": 4.396683415921247e-06,
+      "loss": 0.68,
+      "step": 8630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8376007661852223,
+      "learning_rate": 4.3956545148124665e-06,
+      "loss": 0.6459,
+      "step": 8631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7491758474629922,
+      "learning_rate": 4.394625639673756e-06,
+      "loss": 0.7406,
+      "step": 8632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2356431891812845,
+      "learning_rate": 4.39359679054933e-06,
+      "loss": 0.5924,
+      "step": 8633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.083463143238993,
+      "learning_rate": 4.392567967483401e-06,
+      "loss": 0.7039,
+      "step": 8634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0105827118193833,
+      "learning_rate": 4.3915391705201805e-06,
+      "loss": 0.7749,
+      "step": 8635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7349721128842857,
+      "learning_rate": 4.390510399703875e-06,
+      "loss": 0.7277,
+      "step": 8636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8165982560304705,
+      "learning_rate": 4.389481655078695e-06,
+      "loss": 0.7103,
+      "step": 8637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.548733435780029,
+      "learning_rate": 4.38845293668885e-06,
+      "loss": 0.7324,
+      "step": 8638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6621674091612935,
+      "learning_rate": 4.387424244578543e-06,
+      "loss": 0.6639,
+      "step": 8639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1007234081869115,
+      "learning_rate": 4.386395578791981e-06,
+      "loss": 0.6835,
+      "step": 8640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6200048495163106,
+      "learning_rate": 4.385366939373365e-06,
+      "loss": 0.7098,
+      "step": 8641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.868047606825379,
+      "learning_rate": 4.384338326366905e-06,
+      "loss": 0.7463,
+      "step": 8642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6264414012120711,
+      "learning_rate": 4.383309739816795e-06,
+      "loss": 0.6898,
+      "step": 8643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9608863567130874,
+      "learning_rate": 4.38228117976724e-06,
+      "loss": 0.7171,
+      "step": 8644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3397946804193732,
+      "learning_rate": 4.381252646262437e-06,
+      "loss": 0.6709,
+      "step": 8645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0979257911122,
+      "learning_rate": 4.3802241393465885e-06,
+      "loss": 0.7618,
+      "step": 8646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7290902639209074,
+      "learning_rate": 4.3791956590638866e-06,
+      "loss": 0.6021,
+      "step": 8647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7773725172689043,
+      "learning_rate": 4.378167205458531e-06,
+      "loss": 0.6164,
+      "step": 8648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0997220975043547,
+      "learning_rate": 4.377138778574716e-06,
+      "loss": 0.6191,
+      "step": 8649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9352863251240686,
+      "learning_rate": 4.376110378456634e-06,
+      "loss": 0.6879,
+      "step": 8650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5890094207788368,
+      "learning_rate": 4.375082005148479e-06,
+      "loss": 0.7501,
+      "step": 8651
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7591159648742014,
+      "learning_rate": 4.37405365869444e-06,
+      "loss": 0.7152,
+      "step": 8652
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8161800307209897,
+      "learning_rate": 4.373025339138713e-06,
+      "loss": 0.7017,
+      "step": 8653
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8286562526332333,
+      "learning_rate": 4.371997046525481e-06,
+      "loss": 0.7325,
+      "step": 8654
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6666749432766903,
+      "learning_rate": 4.370968780898935e-06,
+      "loss": 0.7648,
+      "step": 8655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7833624970647568,
+      "learning_rate": 4.3699405423032625e-06,
+      "loss": 0.6919,
+      "step": 8656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.632765063586775,
+      "learning_rate": 4.368912330782647e-06,
+      "loss": 0.7506,
+      "step": 8657
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5787969346716395,
+      "learning_rate": 4.367884146381274e-06,
+      "loss": 0.6229,
+      "step": 8658
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.754008242206463,
+      "learning_rate": 4.366855989143326e-06,
+      "loss": 0.6124,
+      "step": 8659
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0450876071394521,
+      "learning_rate": 4.365827859112989e-06,
+      "loss": 0.6724,
+      "step": 8660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3266432117876972,
+      "learning_rate": 4.364799756334439e-06,
+      "loss": 0.7843,
+      "step": 8661
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6385190284550528,
+      "learning_rate": 4.3637716808518596e-06,
+      "loss": 0.7278,
+      "step": 8662
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6520080117251748,
+      "learning_rate": 4.362743632709426e-06,
+      "loss": 0.6913,
+      "step": 8663
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0230680236870486,
+      "learning_rate": 4.3617156119513206e-06,
+      "loss": 0.7617,
+      "step": 8664
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.676621250496029,
+      "learning_rate": 4.360687618621715e-06,
+      "loss": 0.7837,
+      "step": 8665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1952489848387513,
+      "learning_rate": 4.359659652764786e-06,
+      "loss": 0.5902,
+      "step": 8666
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7483154997583572,
+      "learning_rate": 4.35863171442471e-06,
+      "loss": 0.8297,
+      "step": 8667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9221604652281369,
+      "learning_rate": 4.357603803645657e-06,
+      "loss": 0.5323,
+      "step": 8668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4849571879074521,
+      "learning_rate": 4.356575920471796e-06,
+      "loss": 0.7391,
+      "step": 8669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.666958220252395,
+      "learning_rate": 4.355548064947303e-06,
+      "loss": 0.8004,
+      "step": 8670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6583311876203746,
+      "learning_rate": 4.354520237116347e-06,
+      "loss": 0.7032,
+      "step": 8671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6264169348625863,
+      "learning_rate": 4.353492437023092e-06,
+      "loss": 0.723,
+      "step": 8672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9207142085129014,
+      "learning_rate": 4.352464664711706e-06,
+      "loss": 0.7708,
+      "step": 8673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.688721437554655,
+      "learning_rate": 4.351436920226357e-06,
+      "loss": 0.8165,
+      "step": 8674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.634385501934032,
+      "learning_rate": 4.350409203611207e-06,
+      "loss": 0.7365,
+      "step": 8675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9267346601989594,
+      "learning_rate": 4.34938151491042e-06,
+      "loss": 0.8199,
+      "step": 8676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7925583620543264,
+      "learning_rate": 4.348353854168158e-06,
+      "loss": 0.7262,
+      "step": 8677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7795472468255842,
+      "learning_rate": 4.347326221428585e-06,
+      "loss": 0.7392,
+      "step": 8678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.047625362841439,
+      "learning_rate": 4.346298616735855e-06,
+      "loss": 0.8663,
+      "step": 8679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5800359330773757,
+      "learning_rate": 4.345271040134129e-06,
+      "loss": 0.6551,
+      "step": 8680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1516491068557864,
+      "learning_rate": 4.344243491667566e-06,
+      "loss": 0.8386,
+      "step": 8681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4801812300933577,
+      "learning_rate": 4.343215971380323e-06,
+      "loss": 0.8037,
+      "step": 8682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1065577383461975,
+      "learning_rate": 4.342188479316548e-06,
+      "loss": 0.7272,
+      "step": 8683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9324183348053696,
+      "learning_rate": 4.3411610155204e-06,
+      "loss": 0.7906,
+      "step": 8684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9666907163980931,
+      "learning_rate": 4.340133580036033e-06,
+      "loss": 0.8048,
+      "step": 8685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0441586488571348,
+      "learning_rate": 4.339106172907594e-06,
+      "loss": 0.6387,
+      "step": 8686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2464329811465584,
+      "learning_rate": 4.338078794179234e-06,
+      "loss": 0.7432,
+      "step": 8687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.970917471346122,
+      "learning_rate": 4.337051443895102e-06,
+      "loss": 0.7363,
+      "step": 8688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7140171462836573,
+      "learning_rate": 4.336024122099348e-06,
+      "loss": 0.6337,
+      "step": 8689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0310543802644805,
+      "learning_rate": 4.334996828836115e-06,
+      "loss": 0.7665,
+      "step": 8690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8197574197275892,
+      "learning_rate": 4.333969564149549e-06,
+      "loss": 0.8306,
+      "step": 8691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.960410350423029,
+      "learning_rate": 4.332942328083796e-06,
+      "loss": 0.7169,
+      "step": 8692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8807704404666867,
+      "learning_rate": 4.3319151206829955e-06,
+      "loss": 0.6718,
+      "step": 8693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6541718900376507,
+      "learning_rate": 4.330887941991288e-06,
+      "loss": 0.702,
+      "step": 8694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.667953260783383,
+      "learning_rate": 4.329860792052819e-06,
+      "loss": 0.795,
+      "step": 8695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6937832467022558,
+      "learning_rate": 4.3288336709117246e-06,
+      "loss": 0.7857,
+      "step": 8696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9802942819853897,
+      "learning_rate": 4.32780657861214e-06,
+      "loss": 0.6952,
+      "step": 8697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6933432380942177,
+      "learning_rate": 4.326779515198203e-06,
+      "loss": 0.6804,
+      "step": 8698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.557825106578164,
+      "learning_rate": 4.325752480714052e-06,
+      "loss": 0.6552,
+      "step": 8699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4635667493695952,
+      "learning_rate": 4.324725475203818e-06,
+      "loss": 0.5877,
+      "step": 8700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.813090497811831,
+      "learning_rate": 4.323698498711634e-06,
+      "loss": 0.715,
+      "step": 8701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7491190415892242,
+      "learning_rate": 4.32267155128163e-06,
+      "loss": 0.7848,
+      "step": 8702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8816980328225306,
+      "learning_rate": 4.321644632957941e-06,
+      "loss": 0.7808,
+      "step": 8703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6309638606586878,
+      "learning_rate": 4.320617743784691e-06,
+      "loss": 0.6893,
+      "step": 8704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6089057893500935,
+      "learning_rate": 4.31959088380601e-06,
+      "loss": 0.7305,
+      "step": 8705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9647333843233867,
+      "learning_rate": 4.318564053066023e-06,
+      "loss": 0.6548,
+      "step": 8706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5729339289985709,
+      "learning_rate": 4.317537251608859e-06,
+      "loss": 0.6436,
+      "step": 8707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.815831413037446,
+      "learning_rate": 4.316510479478636e-06,
+      "loss": 0.7019,
+      "step": 8708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5410445106235333,
+      "learning_rate": 4.315483736719482e-06,
+      "loss": 0.8545,
+      "step": 8709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5528385431606,
+      "learning_rate": 4.314457023375517e-06,
+      "loss": 0.6616,
+      "step": 8710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9473112782508549,
+      "learning_rate": 4.313430339490859e-06,
+      "loss": 0.6501,
+      "step": 8711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6961088700648455,
+      "learning_rate": 4.312403685109627e-06,
+      "loss": 0.7138,
+      "step": 8712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1934526042542983,
+      "learning_rate": 4.311377060275942e-06,
+      "loss": 0.606,
+      "step": 8713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5555267021558643,
+      "learning_rate": 4.310350465033919e-06,
+      "loss": 0.6783,
+      "step": 8714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6057342754381443,
+      "learning_rate": 4.309323899427671e-06,
+      "loss": 0.7268,
+      "step": 8715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5166335252919552,
+      "learning_rate": 4.308297363501314e-06,
+      "loss": 0.7748,
+      "step": 8716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.685711300367988,
+      "learning_rate": 4.3072708572989585e-06,
+      "loss": 0.5975,
+      "step": 8717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5230395518263276,
+      "learning_rate": 4.306244380864719e-06,
+      "loss": 0.762,
+      "step": 8718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4305520526890858,
+      "learning_rate": 4.305217934242703e-06,
+      "loss": 0.659,
+      "step": 8719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6533446891817918,
+      "learning_rate": 4.304191517477019e-06,
+      "loss": 0.7126,
+      "step": 8720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8052974993416422,
+      "learning_rate": 4.303165130611776e-06,
+      "loss": 0.7143,
+      "step": 8721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.096958501350236,
+      "learning_rate": 4.302138773691079e-06,
+      "loss": 0.7132,
+      "step": 8722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.016132452909418,
+      "learning_rate": 4.301112446759033e-06,
+      "loss": 0.6712,
+      "step": 8723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0736584686468953,
+      "learning_rate": 4.30008614985974e-06,
+      "loss": 0.5672,
+      "step": 8724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8424788347775636,
+      "learning_rate": 4.299059883037308e-06,
+      "loss": 0.7637,
+      "step": 8725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5873179235479937,
+      "learning_rate": 4.2980336463358296e-06,
+      "loss": 0.6028,
+      "step": 8726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.92766505691209,
+      "learning_rate": 4.29700743979941e-06,
+      "loss": 0.7848,
+      "step": 8727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.266640654776384,
+      "learning_rate": 4.295981263472148e-06,
+      "loss": 0.7396,
+      "step": 8728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6634976383063724,
+      "learning_rate": 4.294955117398139e-06,
+      "loss": 0.8268,
+      "step": 8729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9638933445785953,
+      "learning_rate": 4.293929001621477e-06,
+      "loss": 0.7957,
+      "step": 8730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8028800550364865,
+      "learning_rate": 4.2929029161862575e-06,
+      "loss": 0.7501,
+      "step": 8731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7727869013934647,
+      "learning_rate": 4.291876861136578e-06,
+      "loss": 0.6728,
+      "step": 8732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0749220239156927,
+      "learning_rate": 4.290850836516526e-06,
+      "loss": 0.6734,
+      "step": 8733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.57521730474748,
+      "learning_rate": 4.28982484237019e-06,
+      "loss": 0.6968,
+      "step": 8734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8438698385634478,
+      "learning_rate": 4.288798878741664e-06,
+      "loss": 0.7081,
+      "step": 8735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0805323293553413,
+      "learning_rate": 4.287772945675035e-06,
+      "loss": 0.6722,
+      "step": 8736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5336810880990779,
+      "learning_rate": 4.286747043214388e-06,
+      "loss": 0.7008,
+      "step": 8737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5857302071717496,
+      "learning_rate": 4.285721171403809e-06,
+      "loss": 0.6329,
+      "step": 8738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5855098899093898,
+      "learning_rate": 4.284695330287383e-06,
+      "loss": 0.6253,
+      "step": 8739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7590512636825253,
+      "learning_rate": 4.283669519909189e-06,
+      "loss": 0.7685,
+      "step": 8740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.995769122847355,
+      "learning_rate": 4.282643740313312e-06,
+      "loss": 0.6377,
+      "step": 8741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7019823478153593,
+      "learning_rate": 4.281617991543832e-06,
+      "loss": 0.8798,
+      "step": 8742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6180464699865507,
+      "learning_rate": 4.280592273644829e-06,
+      "loss": 0.6467,
+      "step": 8743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1938612030099711,
+      "learning_rate": 4.279566586660375e-06,
+      "loss": 0.6364,
+      "step": 8744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6921400382542233,
+      "learning_rate": 4.278540930634549e-06,
+      "loss": 0.709,
+      "step": 8745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7384217164306457,
+      "learning_rate": 4.277515305611427e-06,
+      "loss": 0.7277,
+      "step": 8746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8195631981015912,
+      "learning_rate": 4.276489711635083e-06,
+      "loss": 0.7256,
+      "step": 8747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.754743499848437,
+      "learning_rate": 4.275464148749585e-06,
+      "loss": 0.7355,
+      "step": 8748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0810722861239404,
+      "learning_rate": 4.274438616999007e-06,
+      "loss": 0.6004,
+      "step": 8749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.201118750552473,
+      "learning_rate": 4.273413116427419e-06,
+      "loss": 0.7081,
+      "step": 8750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4925496675208165,
+      "learning_rate": 4.272387647078885e-06,
+      "loss": 0.6125,
+      "step": 8751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5717060489446586,
+      "learning_rate": 4.271362208997476e-06,
+      "loss": 0.6365,
+      "step": 8752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0278679854381165,
+      "learning_rate": 4.270336802227255e-06,
+      "loss": 0.7618,
+      "step": 8753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8079100221644429,
+      "learning_rate": 4.269311426812287e-06,
+      "loss": 0.7172,
+      "step": 8754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4613877868849101,
+      "learning_rate": 4.268286082796634e-06,
+      "loss": 0.7817,
+      "step": 8755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5956326880450795,
+      "learning_rate": 4.267260770224358e-06,
+      "loss": 0.5978,
+      "step": 8756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8059730689665259,
+      "learning_rate": 4.26623548913952e-06,
+      "loss": 0.7233,
+      "step": 8757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5025674796866355,
+      "learning_rate": 4.2652102395861746e-06,
+      "loss": 0.6788,
+      "step": 8758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7073111035389397,
+      "learning_rate": 4.264185021608382e-06,
+      "loss": 0.6501,
+      "step": 8759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.833261689360021,
+      "learning_rate": 4.263159835250199e-06,
+      "loss": 0.7436,
+      "step": 8760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7425588942550259,
+      "learning_rate": 4.26213468055568e-06,
+      "loss": 0.6746,
+      "step": 8761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8053973831935404,
+      "learning_rate": 4.2611095575688745e-06,
+      "loss": 0.7447,
+      "step": 8762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.792945288267956,
+      "learning_rate": 4.260084466333837e-06,
+      "loss": 0.6509,
+      "step": 8763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1292040112241297,
+      "learning_rate": 4.259059406894619e-06,
+      "loss": 0.7447,
+      "step": 8764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.142631719556764,
+      "learning_rate": 4.258034379295268e-06,
+      "loss": 0.7031,
+      "step": 8765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7146950746236491,
+      "learning_rate": 4.257009383579833e-06,
+      "loss": 0.788,
+      "step": 8766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.567078247672029,
+      "learning_rate": 4.255984419792358e-06,
+      "loss": 0.6861,
+      "step": 8767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5752694387598136,
+      "learning_rate": 4.254959487976892e-06,
+      "loss": 0.6767,
+      "step": 8768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.907117468904459,
+      "learning_rate": 4.253934588177473e-06,
+      "loss": 0.8136,
+      "step": 8769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8381304800139435,
+      "learning_rate": 4.252909720438149e-06,
+      "loss": 0.6956,
+      "step": 8770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4526427029302689,
+      "learning_rate": 4.251884884802956e-06,
+      "loss": 0.7412,
+      "step": 8771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.030564652166688,
+      "learning_rate": 4.250860081315937e-06,
+      "loss": 0.6367,
+      "step": 8772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5903050237680705,
+      "learning_rate": 4.2498353100211275e-06,
+      "loss": 0.6242,
+      "step": 8773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.20563647597976,
+      "learning_rate": 4.248810570962567e-06,
+      "loss": 0.7127,
+      "step": 8774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4685486363517628,
+      "learning_rate": 4.247785864184289e-06,
+      "loss": 0.7661,
+      "step": 8775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7358737326747027,
+      "learning_rate": 4.246761189730327e-06,
+      "loss": 0.6716,
+      "step": 8776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6416607798037701,
+      "learning_rate": 4.245736547644714e-06,
+      "loss": 0.7503,
+      "step": 8777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1027659353590202,
+      "learning_rate": 4.2447119379714805e-06,
+      "loss": 0.5838,
+      "step": 8778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5000392417232766,
+      "learning_rate": 4.24368736075466e-06,
+      "loss": 0.5643,
+      "step": 8779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5971512532367662,
+      "learning_rate": 4.242662816038276e-06,
+      "loss": 0.7457,
+      "step": 8780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4571176022662595,
+      "learning_rate": 4.241638303866358e-06,
+      "loss": 0.6749,
+      "step": 8781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7484167803502741,
+      "learning_rate": 4.24061382428293e-06,
+      "loss": 0.7799,
+      "step": 8782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8162248264771859,
+      "learning_rate": 4.23958937733202e-06,
+      "loss": 0.674,
+      "step": 8783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1645990758283868,
+      "learning_rate": 4.238564963057646e-06,
+      "loss": 0.6195,
+      "step": 8784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5735545150235348,
+      "learning_rate": 4.237540581503831e-06,
+      "loss": 0.7169,
+      "step": 8785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.711971733604921,
+      "learning_rate": 4.236516232714597e-06,
+      "loss": 0.7614,
+      "step": 8786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4905300396956287,
+      "learning_rate": 4.235491916733959e-06,
+      "loss": 0.7263,
+      "step": 8787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.960379080679407,
+      "learning_rate": 4.234467633605937e-06,
+      "loss": 0.6194,
+      "step": 8788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7544724157169773,
+      "learning_rate": 4.233443383374545e-06,
+      "loss": 0.6828,
+      "step": 8789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7454249762187084,
+      "learning_rate": 4.232419166083799e-06,
+      "loss": 0.7105,
+      "step": 8790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7975750537026016,
+      "learning_rate": 4.23139498177771e-06,
+      "loss": 0.7279,
+      "step": 8791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5930155846418934,
+      "learning_rate": 4.23037083050029e-06,
+      "loss": 0.6738,
+      "step": 8792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5165765414793915,
+      "learning_rate": 4.229346712295551e-06,
+      "loss": 0.636,
+      "step": 8793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6999313658081252,
+      "learning_rate": 4.228322627207499e-06,
+      "loss": 0.7976,
+      "step": 8794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.943721152342403,
+      "learning_rate": 4.227298575280142e-06,
+      "loss": 0.8409,
+      "step": 8795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6667834452512882,
+      "learning_rate": 4.226274556557487e-06,
+      "loss": 0.6856,
+      "step": 8796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7020831627133637,
+      "learning_rate": 4.225250571083538e-06,
+      "loss": 0.6589,
+      "step": 8797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.916500147210714,
+      "learning_rate": 4.2242266189022975e-06,
+      "loss": 0.8586,
+      "step": 8798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2131800640913677,
+      "learning_rate": 4.223202700057765e-06,
+      "loss": 0.8005,
+      "step": 8799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2122872058398297,
+      "learning_rate": 4.2221788145939425e-06,
+      "loss": 0.6962,
+      "step": 8800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.665451787833109,
+      "learning_rate": 4.221154962554831e-06,
+      "loss": 0.7831,
+      "step": 8801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0266075229908802,
+      "learning_rate": 4.220131143984424e-06,
+      "loss": 0.5755,
+      "step": 8802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9021279842141077,
+      "learning_rate": 4.219107358926718e-06,
+      "loss": 0.5632,
+      "step": 8803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.701686595004378,
+      "learning_rate": 4.21808360742571e-06,
+      "loss": 0.7406,
+      "step": 8804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0712584288352494,
+      "learning_rate": 4.217059889525389e-06,
+      "loss": 0.7084,
+      "step": 8805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.011759573362137,
+      "learning_rate": 4.216036205269748e-06,
+      "loss": 0.7142,
+      "step": 8806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9783132519747813,
+      "learning_rate": 4.215012554702778e-06,
+      "loss": 0.6646,
+      "step": 8807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.85035424238554,
+      "learning_rate": 4.213988937868469e-06,
+      "loss": 0.6069,
+      "step": 8808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4930205155324285,
+      "learning_rate": 4.212965354810802e-06,
+      "loss": 0.6687,
+      "step": 8809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7421917039207264,
+      "learning_rate": 4.211941805573767e-06,
+      "loss": 0.7239,
+      "step": 8810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6187460613679603,
+      "learning_rate": 4.21091829020135e-06,
+      "loss": 0.6347,
+      "step": 8811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.688418133132079,
+      "learning_rate": 4.209894808737531e-06,
+      "loss": 0.6603,
+      "step": 8812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8879681843696428,
+      "learning_rate": 4.20887136122629e-06,
+      "loss": 0.7067,
+      "step": 8813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6984207683572337,
+      "learning_rate": 4.207847947711609e-06,
+      "loss": 0.8447,
+      "step": 8814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2402836293057817,
+      "learning_rate": 4.206824568237468e-06,
+      "loss": 0.5764,
+      "step": 8815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6968832655075496,
+      "learning_rate": 4.205801222847839e-06,
+      "loss": 0.7271,
+      "step": 8816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8485238338292276,
+      "learning_rate": 4.204777911586702e-06,
+      "loss": 0.6821,
+      "step": 8817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7986980795827159,
+      "learning_rate": 4.203754634498027e-06,
+      "loss": 0.7915,
+      "step": 8818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.716107999722378,
+      "learning_rate": 4.202731391625793e-06,
+      "loss": 0.7225,
+      "step": 8819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.071588423481024,
+      "learning_rate": 4.201708183013963e-06,
+      "loss": 0.9038,
+      "step": 8820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8357832928166702,
+      "learning_rate": 4.200685008706511e-06,
+      "loss": 0.8064,
+      "step": 8821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7499008896735107,
+      "learning_rate": 4.199661868747406e-06,
+      "loss": 0.782,
+      "step": 8822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5366071411737325,
+      "learning_rate": 4.198638763180611e-06,
+      "loss": 0.7481,
+      "step": 8823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6412837829655325,
+      "learning_rate": 4.1976156920500935e-06,
+      "loss": 0.7738,
+      "step": 8824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7489562812842556,
+      "learning_rate": 4.196592655399816e-06,
+      "loss": 0.6946,
+      "step": 8825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8826363130685082,
+      "learning_rate": 4.195569653273743e-06,
+      "loss": 0.8262,
+      "step": 8826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4993833705803274,
+      "learning_rate": 4.1945466857158336e-06,
+      "loss": 0.668,
+      "step": 8827
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3308132442105984,
+      "learning_rate": 4.193523752770045e-06,
+      "loss": 0.8787,
+      "step": 8828
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2921659069804101,
+      "learning_rate": 4.19250085448034e-06,
+      "loss": 0.7093,
+      "step": 8829
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9773802668230387,
+      "learning_rate": 4.191477990890668e-06,
+      "loss": 0.7318,
+      "step": 8830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5221223755107067,
+      "learning_rate": 4.19045516204499e-06,
+      "loss": 0.7353,
+      "step": 8831
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7987329266855818,
+      "learning_rate": 4.189432367987255e-06,
+      "loss": 0.7155,
+      "step": 8832
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8430341990492864,
+      "learning_rate": 4.188409608761419e-06,
+      "loss": 0.7901,
+      "step": 8833
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0567397914584795,
+      "learning_rate": 4.187386884411426e-06,
+      "loss": 0.6269,
+      "step": 8834
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.608758302275363,
+      "learning_rate": 4.186364194981231e-06,
+      "loss": 0.6704,
+      "step": 8835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.744336352165245,
+      "learning_rate": 4.1853415405147765e-06,
+      "loss": 0.8149,
+      "step": 8836
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.880806381001084,
+      "learning_rate": 4.184318921056013e-06,
+      "loss": 0.8019,
+      "step": 8837
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5748053992609015,
+      "learning_rate": 4.18329633664888e-06,
+      "loss": 0.6882,
+      "step": 8838
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.814026333091498,
+      "learning_rate": 4.182273787337323e-06,
+      "loss": 0.691,
+      "step": 8839
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0034753748282417,
+      "learning_rate": 4.181251273165283e-06,
+      "loss": 0.5708,
+      "step": 8840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6533707742883532,
+      "learning_rate": 4.1802287941767e-06,
+      "loss": 0.9252,
+      "step": 8841
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8642334224981045,
+      "learning_rate": 4.179206350415509e-06,
+      "loss": 0.6397,
+      "step": 8842
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8022063919814697,
+      "learning_rate": 4.17818394192565e-06,
+      "loss": 0.7224,
+      "step": 8843
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.442453600705675,
+      "learning_rate": 4.177161568751058e-06,
+      "loss": 0.7161,
+      "step": 8844
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0728202775676605,
+      "learning_rate": 4.176139230935666e-06,
+      "loss": 0.5567,
+      "step": 8845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.544941250140119,
+      "learning_rate": 4.175116928523405e-06,
+      "loss": 0.6493,
+      "step": 8846
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0851964347014795,
+      "learning_rate": 4.174094661558209e-06,
+      "loss": 0.7889,
+      "step": 8847
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0929875705257914,
+      "learning_rate": 4.173072430084002e-06,
+      "loss": 0.8239,
+      "step": 8848
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8487861830969488,
+      "learning_rate": 4.172050234144716e-06,
+      "loss": 0.8421,
+      "step": 8849
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5688571720889641,
+      "learning_rate": 4.171028073784274e-06,
+      "loss": 0.6265,
+      "step": 8850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7392601896690854,
+      "learning_rate": 4.170005949046604e-06,
+      "loss": 0.6622,
+      "step": 8851
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9059327393472278,
+      "learning_rate": 4.168983859975625e-06,
+      "loss": 0.6795,
+      "step": 8852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7509103024873163,
+      "learning_rate": 4.16796180661526e-06,
+      "loss": 0.7984,
+      "step": 8853
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6878044755542205,
+      "learning_rate": 4.16693978900943e-06,
+      "loss": 0.6725,
+      "step": 8854
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6999642274098763,
+      "learning_rate": 4.165917807202055e-06,
+      "loss": 0.7493,
+      "step": 8855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5833228848805692,
+      "learning_rate": 4.164895861237046e-06,
+      "loss": 0.6647,
+      "step": 8856
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.313475007695267,
+      "learning_rate": 4.1638739511583224e-06,
+      "loss": 0.6888,
+      "step": 8857
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8990461768815778,
+      "learning_rate": 4.1628520770097994e-06,
+      "loss": 0.7301,
+      "step": 8858
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6951273449609212,
+      "learning_rate": 4.161830238835386e-06,
+      "loss": 0.8494,
+      "step": 8859
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7522529817812034,
+      "learning_rate": 4.160808436678992e-06,
+      "loss": 0.6967,
+      "step": 8860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8891271623445867,
+      "learning_rate": 4.1597866705845295e-06,
+      "loss": 0.7532,
+      "step": 8861
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2317450878392813,
+      "learning_rate": 4.1587649405959065e-06,
+      "loss": 0.5841,
+      "step": 8862
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.799837077540664,
+      "learning_rate": 4.157743246757026e-06,
+      "loss": 0.6918,
+      "step": 8863
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.608688866417239,
+      "learning_rate": 4.156721589111794e-06,
+      "loss": 0.7428,
+      "step": 8864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5019865312764369,
+      "learning_rate": 4.155699967704113e-06,
+      "loss": 0.7601,
+      "step": 8865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.976820778297411,
+      "learning_rate": 4.154678382577887e-06,
+      "loss": 0.7215,
+      "step": 8866
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0639126521223243,
+      "learning_rate": 4.1536568337770114e-06,
+      "loss": 0.8396,
+      "step": 8867
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5872761557181725,
+      "learning_rate": 4.152635321345387e-06,
+      "loss": 0.7432,
+      "step": 8868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9748723060490214,
+      "learning_rate": 4.151613845326912e-06,
+      "loss": 0.8143,
+      "step": 8869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.450099580191211,
+      "learning_rate": 4.150592405765476e-06,
+      "loss": 0.6537,
+      "step": 8870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.283077174808581,
+      "learning_rate": 4.1495710027049776e-06,
+      "loss": 0.8025,
+      "step": 8871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.293125721767891,
+      "learning_rate": 4.148549636189308e-06,
+      "loss": 0.7817,
+      "step": 8872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4523423368548014,
+      "learning_rate": 4.147528306262357e-06,
+      "loss": 0.6005,
+      "step": 8873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.228964320091807,
+      "learning_rate": 4.146507012968013e-06,
+      "loss": 0.5732,
+      "step": 8874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9598605167032503,
+      "learning_rate": 4.145485756350162e-06,
+      "loss": 0.7132,
+      "step": 8875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7048624330162432,
+      "learning_rate": 4.144464536452693e-06,
+      "loss": 0.6917,
+      "step": 8876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7518744270682973,
+      "learning_rate": 4.1434433533194865e-06,
+      "loss": 0.7257,
+      "step": 8877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1965419164300342,
+      "learning_rate": 4.142422206994428e-06,
+      "loss": 0.7627,
+      "step": 8878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4858112706567812,
+      "learning_rate": 4.141401097521396e-06,
+      "loss": 0.6396,
+      "step": 8879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7507439011255892,
+      "learning_rate": 4.140380024944273e-06,
+      "loss": 0.7428,
+      "step": 8880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7629635083067912,
+      "learning_rate": 4.139358989306933e-06,
+      "loss": 0.6911,
+      "step": 8881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1353639050400175,
+      "learning_rate": 4.138337990653255e-06,
+      "loss": 0.6642,
+      "step": 8882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0806290779714627,
+      "learning_rate": 4.137317029027111e-06,
+      "loss": 0.6374,
+      "step": 8883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8964947288943041,
+      "learning_rate": 4.136296104472378e-06,
+      "loss": 0.8563,
+      "step": 8884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8825918140381601,
+      "learning_rate": 4.1352752170329236e-06,
+      "loss": 0.7137,
+      "step": 8885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8387239108713078,
+      "learning_rate": 4.1342543667526195e-06,
+      "loss": 0.7124,
+      "step": 8886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9323026749812138,
+      "learning_rate": 4.133233553675334e-06,
+      "loss": 0.6884,
+      "step": 8887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.770419583878973,
+      "learning_rate": 4.1322127778449315e-06,
+      "loss": 0.7516,
+      "step": 8888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1950894204914209,
+      "learning_rate": 4.131192039305278e-06,
+      "loss": 0.6877,
+      "step": 8889
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7158015251613323,
+      "learning_rate": 4.1301713381002394e-06,
+      "loss": 0.7595,
+      "step": 8890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5490253800357248,
+      "learning_rate": 4.129150674273675e-06,
+      "loss": 0.7997,
+      "step": 8891
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5916818698681252,
+      "learning_rate": 4.128130047869446e-06,
+      "loss": 0.6922,
+      "step": 8892
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.155621912558901,
+      "learning_rate": 4.12710945893141e-06,
+      "loss": 0.7377,
+      "step": 8893
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.703637793860835,
+      "learning_rate": 4.1260889075034254e-06,
+      "loss": 0.674,
+      "step": 8894
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6632081025083234,
+      "learning_rate": 4.125068393629346e-06,
+      "loss": 0.7449,
+      "step": 8895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.623699830621374,
+      "learning_rate": 4.124047917353026e-06,
+      "loss": 0.5701,
+      "step": 8896
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1357707395915644,
+      "learning_rate": 4.123027478718318e-06,
+      "loss": 0.7947,
+      "step": 8897
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5439493500816361,
+      "learning_rate": 4.1220070777690736e-06,
+      "loss": 0.7116,
+      "step": 8898
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0671720874122292,
+      "learning_rate": 4.120986714549139e-06,
+      "loss": 0.7,
+      "step": 8899
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.729038875479124,
+      "learning_rate": 4.119966389102363e-06,
+      "loss": 0.6648,
+      "step": 8900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.191524448439295,
+      "learning_rate": 4.118946101472591e-06,
+      "loss": 0.6977,
+      "step": 8901
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4825518596049578,
+      "learning_rate": 4.117925851703669e-06,
+      "loss": 0.7112,
+      "step": 8902
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6225141755158428,
+      "learning_rate": 4.116905639839436e-06,
+      "loss": 0.6697,
+      "step": 8903
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8987707129403604,
+      "learning_rate": 4.115885465923734e-06,
+      "loss": 0.7578,
+      "step": 8904
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7105779431156964,
+      "learning_rate": 4.114865330000405e-06,
+      "loss": 0.7482,
+      "step": 8905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0410321614553637,
+      "learning_rate": 4.113845232113282e-06,
+      "loss": 0.6604,
+      "step": 8906
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.275461945230651,
+      "learning_rate": 4.112825172306203e-06,
+      "loss": 0.8334,
+      "step": 8907
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6729916529676572,
+      "learning_rate": 4.111805150623002e-06,
+      "loss": 0.7647,
+      "step": 8908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7693430691230227,
+      "learning_rate": 4.110785167107514e-06,
+      "loss": 0.7532,
+      "step": 8909
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7728150441052752,
+      "learning_rate": 4.1097652218035666e-06,
+      "loss": 0.6536,
+      "step": 8910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6517762458813063,
+      "learning_rate": 4.108745314754989e-06,
+      "loss": 0.8469,
+      "step": 8911
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5475120962623745,
+      "learning_rate": 4.107725446005612e-06,
+      "loss": 0.6853,
+      "step": 8912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.681078269465921,
+      "learning_rate": 4.106705615599258e-06,
+      "loss": 0.7502,
+      "step": 8913
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9730862974044,
+      "learning_rate": 4.1056858235797545e-06,
+      "loss": 0.7543,
+      "step": 8914
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6685398696650602,
+      "learning_rate": 4.104666069990921e-06,
+      "loss": 0.6693,
+      "step": 8915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8195936773916335,
+      "learning_rate": 4.103646354876584e-06,
+      "loss": 0.6717,
+      "step": 8916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5698772691054828,
+      "learning_rate": 4.102626678280556e-06,
+      "loss": 0.6756,
+      "step": 8917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0961935880636255,
+      "learning_rate": 4.101607040246659e-06,
+      "loss": 0.6258,
+      "step": 8918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9886593217298876,
+      "learning_rate": 4.100587440818709e-06,
+      "loss": 0.6355,
+      "step": 8919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7518431111276525,
+      "learning_rate": 4.099567880040521e-06,
+      "loss": 0.7107,
+      "step": 8920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9732845030021084,
+      "learning_rate": 4.098548357955904e-06,
+      "loss": 0.6651,
+      "step": 8921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7250389016102794,
+      "learning_rate": 4.0975288746086724e-06,
+      "loss": 0.7595,
+      "step": 8922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5778339453938546,
+      "learning_rate": 4.0965094300426364e-06,
+      "loss": 0.733,
+      "step": 8923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7039917372841056,
+      "learning_rate": 4.0954900243016016e-06,
+      "loss": 0.7235,
+      "step": 8924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5773473333356345,
+      "learning_rate": 4.094470657429374e-06,
+      "loss": 0.6707,
+      "step": 8925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3642011819941624,
+      "learning_rate": 4.093451329469758e-06,
+      "loss": 0.71,
+      "step": 8926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9986887609450334,
+      "learning_rate": 4.092432040466561e-06,
+      "loss": 0.7564,
+      "step": 8927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9507756553507163,
+      "learning_rate": 4.091412790463578e-06,
+      "loss": 0.6762,
+      "step": 8928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5815605484224795,
+      "learning_rate": 4.090393579504612e-06,
+      "loss": 0.8061,
+      "step": 8929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.250100630372114,
+      "learning_rate": 4.0893744076334606e-06,
+      "loss": 0.7628,
+      "step": 8930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7207639943278392,
+      "learning_rate": 4.088355274893916e-06,
+      "loss": 0.6589,
+      "step": 8931
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.641282728333402,
+      "learning_rate": 4.087336181329777e-06,
+      "loss": 0.7334,
+      "step": 8932
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7851265648607313,
+      "learning_rate": 4.086317126984835e-06,
+      "loss": 0.7076,
+      "step": 8933
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7789549447290232,
+      "learning_rate": 4.085298111902882e-06,
+      "loss": 0.8609,
+      "step": 8934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9304604593095513,
+      "learning_rate": 4.084279136127704e-06,
+      "loss": 0.6559,
+      "step": 8935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9030987598163864,
+      "learning_rate": 4.083260199703092e-06,
+      "loss": 0.7654,
+      "step": 8936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4935932418555888,
+      "learning_rate": 4.082241302672831e-06,
+      "loss": 0.6797,
+      "step": 8937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.571183142171023,
+      "learning_rate": 4.081222445080707e-06,
+      "loss": 0.7146,
+      "step": 8938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5781728836901274,
+      "learning_rate": 4.080203626970498e-06,
+      "loss": 0.6194,
+      "step": 8939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8586759625529545,
+      "learning_rate": 4.079184848385988e-06,
+      "loss": 0.6192,
+      "step": 8940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.161942275910268,
+      "learning_rate": 4.078166109370957e-06,
+      "loss": 0.7182,
+      "step": 8941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5497115668045354,
+      "learning_rate": 4.077147409969181e-06,
+      "loss": 0.6956,
+      "step": 8942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0402734189466067,
+      "learning_rate": 4.076128750224436e-06,
+      "loss": 0.7745,
+      "step": 8943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8206481552087632,
+      "learning_rate": 4.0751101301804945e-06,
+      "loss": 0.6636,
+      "step": 8944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8592896403101373,
+      "learning_rate": 4.074091549881134e-06,
+      "loss": 0.7042,
+      "step": 8945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6630738713395226,
+      "learning_rate": 4.0730730093701185e-06,
+      "loss": 0.61,
+      "step": 8946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.570855885883635,
+      "learning_rate": 4.072054508691222e-06,
+      "loss": 0.6249,
+      "step": 8947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8037538424046673,
+      "learning_rate": 4.07103604788821e-06,
+      "loss": 0.7446,
+      "step": 8948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2490484719259778,
+      "learning_rate": 4.070017627004847e-06,
+      "loss": 0.67,
+      "step": 8949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6828021906314978,
+      "learning_rate": 4.068999246084898e-06,
+      "loss": 0.6816,
+      "step": 8950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0100021890248443,
+      "learning_rate": 4.067980905172125e-06,
+      "loss": 0.6574,
+      "step": 8951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9274504906704557,
+      "learning_rate": 4.066962604310288e-06,
+      "loss": 0.6847,
+      "step": 8952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7730377043012904,
+      "learning_rate": 4.065944343543146e-06,
+      "loss": 0.8015,
+      "step": 8953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5853620819660181,
+      "learning_rate": 4.0649261229144554e-06,
+      "loss": 0.7302,
+      "step": 8954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8182094182218214,
+      "learning_rate": 4.063907942467971e-06,
+      "loss": 0.721,
+      "step": 8955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8579657401154401,
+      "learning_rate": 4.06288980224745e-06,
+      "loss": 0.8653,
+      "step": 8956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6823590970205495,
+      "learning_rate": 4.0618717022966395e-06,
+      "loss": 0.7839,
+      "step": 8957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6453610661461922,
+      "learning_rate": 4.060853642659291e-06,
+      "loss": 0.6852,
+      "step": 8958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7559750007385286,
+      "learning_rate": 4.059835623379155e-06,
+      "loss": 0.641,
+      "step": 8959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.537307273363078,
+      "learning_rate": 4.058817644499973e-06,
+      "loss": 0.7147,
+      "step": 8960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.970296508950716,
+      "learning_rate": 4.057799706065495e-06,
+      "loss": 0.7905,
+      "step": 8961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.787378217202663,
+      "learning_rate": 4.056781808119461e-06,
+      "loss": 0.7927,
+      "step": 8962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0883782712660333,
+      "learning_rate": 4.055763950705615e-06,
+      "loss": 0.6739,
+      "step": 8963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5147748840098556,
+      "learning_rate": 4.054746133867693e-06,
+      "loss": 0.588,
+      "step": 8964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1736803902128046,
+      "learning_rate": 4.053728357649436e-06,
+      "loss": 0.6938,
+      "step": 8965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.676792307172639,
+      "learning_rate": 4.052710622094579e-06,
+      "loss": 0.6725,
+      "step": 8966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5506773133011238,
+      "learning_rate": 4.051692927246857e-06,
+      "loss": 0.6823,
+      "step": 8967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6394156805519948,
+      "learning_rate": 4.050675273149999e-06,
+      "loss": 0.6936,
+      "step": 8968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.617455300809906,
+      "learning_rate": 4.0496576598477396e-06,
+      "loss": 0.7125,
+      "step": 8969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.600568909774209,
+      "learning_rate": 4.048640087383809e-06,
+      "loss": 0.7259,
+      "step": 8970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6124794283588115,
+      "learning_rate": 4.0476225558019324e-06,
+      "loss": 0.7361,
+      "step": 8971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7930693805802176,
+      "learning_rate": 4.046605065145834e-06,
+      "loss": 0.7461,
+      "step": 8972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6264711912040413,
+      "learning_rate": 4.04558761545924e-06,
+      "loss": 0.6856,
+      "step": 8973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3485500151782244,
+      "learning_rate": 4.044570206785874e-06,
+      "loss": 0.8163,
+      "step": 8974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6590903926792022,
+      "learning_rate": 4.043552839169451e-06,
+      "loss": 0.7572,
+      "step": 8975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6075195582313955,
+      "learning_rate": 4.042535512653693e-06,
+      "loss": 0.7113,
+      "step": 8976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.796204540493019,
+      "learning_rate": 4.041518227282319e-06,
+      "loss": 0.6858,
+      "step": 8977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.617050950695864,
+      "learning_rate": 4.040500983099038e-06,
+      "loss": 0.6496,
+      "step": 8978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4606466465584047,
+      "learning_rate": 4.039483780147568e-06,
+      "loss": 0.6618,
+      "step": 8979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.799548170336794,
+      "learning_rate": 4.038466618471619e-06,
+      "loss": 0.817,
+      "step": 8980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.106528783542908,
+      "learning_rate": 4.037449498114903e-06,
+      "loss": 0.6824,
+      "step": 8981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7514327706476973,
+      "learning_rate": 4.036432419121123e-06,
+      "loss": 0.7082,
+      "step": 8982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5516509787727415,
+      "learning_rate": 4.035415381533988e-06,
+      "loss": 0.7035,
+      "step": 8983
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7893937771535218,
+      "learning_rate": 4.0343983853972045e-06,
+      "loss": 0.7283,
+      "step": 8984
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5215752019177184,
+      "learning_rate": 4.0333814307544726e-06,
+      "loss": 0.806,
+      "step": 8985
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.691977505145663,
+      "learning_rate": 4.032364517649491e-06,
+      "loss": 0.699,
+      "step": 8986
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.531810140872411,
+      "learning_rate": 4.0313476461259615e-06,
+      "loss": 0.6694,
+      "step": 8987
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5870321739577522,
+      "learning_rate": 4.0303308162275835e-06,
+      "loss": 0.8368,
+      "step": 8988
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5845212565393791,
+      "learning_rate": 4.029314027998049e-06,
+      "loss": 0.7389,
+      "step": 8989
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6338625301248726,
+      "learning_rate": 4.028297281481051e-06,
+      "loss": 0.7058,
+      "step": 8990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1596355036973915,
+      "learning_rate": 4.027280576720282e-06,
+      "loss": 0.7672,
+      "step": 8991
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9615900268118955,
+      "learning_rate": 4.026263913759437e-06,
+      "loss": 0.6977,
+      "step": 8992
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.955493402088102,
+      "learning_rate": 4.025247292642197e-06,
+      "loss": 0.8662,
+      "step": 8993
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.57573924058569,
+      "learning_rate": 4.024230713412253e-06,
+      "loss": 0.7182,
+      "step": 8994
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.165267529086785,
+      "learning_rate": 4.0232141761132894e-06,
+      "loss": 0.6546,
+      "step": 8995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6200460450735386,
+      "learning_rate": 4.022197680788986e-06,
+      "loss": 0.5881,
+      "step": 8996
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0421633466537887,
+      "learning_rate": 4.021181227483026e-06,
+      "loss": 0.551,
+      "step": 8997
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7254998936735397,
+      "learning_rate": 4.02016481623909e-06,
+      "loss": 0.7241,
+      "step": 8998
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7057877845037575,
+      "learning_rate": 4.019148447100855e-06,
+      "loss": 0.7239,
+      "step": 8999
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9952709327090924,
+      "learning_rate": 4.018132120111992e-06,
+      "loss": 0.6779,
+      "step": 9000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9573184592493715,
+      "learning_rate": 4.0171158353161805e-06,
+      "loss": 0.6749,
+      "step": 9001
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7701498117846168,
+      "learning_rate": 4.016099592757091e-06,
+      "loss": 0.8045,
+      "step": 9002
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9076248431446374,
+      "learning_rate": 4.015083392478393e-06,
+      "loss": 0.63,
+      "step": 9003
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9388037249299117,
+      "learning_rate": 4.014067234523756e-06,
+      "loss": 0.7664,
+      "step": 9004
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9189161976870959,
+      "learning_rate": 4.0130511189368435e-06,
+      "loss": 0.5869,
+      "step": 9005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.190313865207998,
+      "learning_rate": 4.012035045761326e-06,
+      "loss": 0.7377,
+      "step": 9006
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7205347882542443,
+      "learning_rate": 4.0110190150408594e-06,
+      "loss": 0.7127,
+      "step": 9007
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8916613510433995,
+      "learning_rate": 4.01000302681911e-06,
+      "loss": 0.7087,
+      "step": 9008
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5882127027335549,
+      "learning_rate": 4.008987081139734e-06,
+      "loss": 0.7394,
+      "step": 9009
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.067657140745371,
+      "learning_rate": 4.007971178046393e-06,
+      "loss": 0.7156,
+      "step": 9010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5169929877925237,
+      "learning_rate": 4.006955317582737e-06,
+      "loss": 0.6537,
+      "step": 9011
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9946074822870961,
+      "learning_rate": 4.005939499792424e-06,
+      "loss": 0.6305,
+      "step": 9012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6950525631741862,
+      "learning_rate": 4.004923724719106e-06,
+      "loss": 0.6959,
+      "step": 9013
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.791261326821622,
+      "learning_rate": 4.0039079924064285e-06,
+      "loss": 0.7787,
+      "step": 9014
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6638837987318869,
+      "learning_rate": 4.0028923028980435e-06,
+      "loss": 0.68,
+      "step": 9015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4192529968538565,
+      "learning_rate": 4.0018766562375984e-06,
+      "loss": 0.7902,
+      "step": 9016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.777080814250431,
+      "learning_rate": 4.000861052468736e-06,
+      "loss": 0.8148,
+      "step": 9017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.832190585653593,
+      "learning_rate": 3.9998454916350996e-06,
+      "loss": 0.7272,
+      "step": 9018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0857118317127268,
+      "learning_rate": 3.998829973780329e-06,
+      "loss": 0.6211,
+      "step": 9019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2973543379556298,
+      "learning_rate": 3.997814498948064e-06,
+      "loss": 0.7707,
+      "step": 9020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4783673510379842,
+      "learning_rate": 3.996799067181943e-06,
+      "loss": 0.6621,
+      "step": 9021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6700281298433328,
+      "learning_rate": 3.995783678525601e-06,
+      "loss": 0.7731,
+      "step": 9022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3696412244493035,
+      "learning_rate": 3.994768333022669e-06,
+      "loss": 0.627,
+      "step": 9023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1725554012775397,
+      "learning_rate": 3.993753030716783e-06,
+      "loss": 0.7853,
+      "step": 9024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.693321015913711,
+      "learning_rate": 3.992737771651567e-06,
+      "loss": 0.7919,
+      "step": 9025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6530435484114734,
+      "learning_rate": 3.991722555870655e-06,
+      "loss": 0.6402,
+      "step": 9026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3885817820854,
+      "learning_rate": 3.990707383417668e-06,
+      "loss": 0.8095,
+      "step": 9027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8939388930469299,
+      "learning_rate": 3.989692254336235e-06,
+      "loss": 0.7255,
+      "step": 9028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0791760842008484,
+      "learning_rate": 3.988677168669974e-06,
+      "loss": 0.7571,
+      "step": 9029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.656057775910834,
+      "learning_rate": 3.987662126462507e-06,
+      "loss": 0.6737,
+      "step": 9030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.235421018309674,
+      "learning_rate": 3.9866471277574554e-06,
+      "loss": 0.7271,
+      "step": 9031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.148939637765672,
+      "learning_rate": 3.985632172598433e-06,
+      "loss": 0.7504,
+      "step": 9032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7426689709995733,
+      "learning_rate": 3.984617261029054e-06,
+      "loss": 0.6279,
+      "step": 9033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3215778417926958,
+      "learning_rate": 3.983602393092931e-06,
+      "loss": 0.6108,
+      "step": 9034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1045871613719234,
+      "learning_rate": 3.982587568833681e-06,
+      "loss": 0.7661,
+      "step": 9035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.740952803912405,
+      "learning_rate": 3.981572788294907e-06,
+      "loss": 0.7836,
+      "step": 9036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7277754592530405,
+      "learning_rate": 3.980558051520218e-06,
+      "loss": 0.7594,
+      "step": 9037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9882512227534304,
+      "learning_rate": 3.97954335855322e-06,
+      "loss": 0.698,
+      "step": 9038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6407374439350506,
+      "learning_rate": 3.978528709437518e-06,
+      "loss": 0.6808,
+      "step": 9039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6786910490515266,
+      "learning_rate": 3.977514104216711e-06,
+      "loss": 0.6742,
+      "step": 9040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.487646895352071,
+      "learning_rate": 3.9764995429344e-06,
+      "loss": 0.5926,
+      "step": 9041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.751101848798728,
+      "learning_rate": 3.975485025634185e-06,
+      "loss": 0.7614,
+      "step": 9042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6753323169712802,
+      "learning_rate": 3.9744705523596575e-06,
+      "loss": 0.7986,
+      "step": 9043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8671166860918018,
+      "learning_rate": 3.973456123154415e-06,
+      "loss": 0.6145,
+      "step": 9044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6789620916230283,
+      "learning_rate": 3.972441738062049e-06,
+      "loss": 0.7316,
+      "step": 9045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5791544238480602,
+      "learning_rate": 3.971427397126153e-06,
+      "loss": 0.7006,
+      "step": 9046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.240241270028661,
+      "learning_rate": 3.970413100390309e-06,
+      "loss": 0.6744,
+      "step": 9047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8270777730552008,
+      "learning_rate": 3.9693988478981076e-06,
+      "loss": 0.7126,
+      "step": 9048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7482118554574912,
+      "learning_rate": 3.9683846396931345e-06,
+      "loss": 0.727,
+      "step": 9049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5197588366676023,
+      "learning_rate": 3.967370475818971e-06,
+      "loss": 0.6761,
+      "step": 9050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6857819941362733,
+      "learning_rate": 3.966356356319196e-06,
+      "loss": 0.8589,
+      "step": 9051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9407903403321602,
+      "learning_rate": 3.965342281237391e-06,
+      "loss": 0.789,
+      "step": 9052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4933352638501827,
+      "learning_rate": 3.964328250617134e-06,
+      "loss": 0.6579,
+      "step": 9053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6066821496584647,
+      "learning_rate": 3.9633142645019965e-06,
+      "loss": 0.754,
+      "step": 9054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2366598774908675,
+      "learning_rate": 3.962300322935556e-06,
+      "loss": 0.7135,
+      "step": 9055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6775586149641513,
+      "learning_rate": 3.961286425961379e-06,
+      "loss": 0.6867,
+      "step": 9056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6663016424219204,
+      "learning_rate": 3.9602725736230405e-06,
+      "loss": 0.7388,
+      "step": 9057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6689347824561527,
+      "learning_rate": 3.959258765964104e-06,
+      "loss": 0.7141,
+      "step": 9058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2004531314049216,
+      "learning_rate": 3.958245003028136e-06,
+      "loss": 0.6805,
+      "step": 9059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.155837218914689,
+      "learning_rate": 3.957231284858701e-06,
+      "loss": 0.7232,
+      "step": 9060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9082894041464464,
+      "learning_rate": 3.956217611499359e-06,
+      "loss": 0.7743,
+      "step": 9061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.649706577151289,
+      "learning_rate": 3.9552039829936705e-06,
+      "loss": 0.8036,
+      "step": 9062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8206683657592124,
+      "learning_rate": 3.954190399385195e-06,
+      "loss": 0.7865,
+      "step": 9063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1296570979911025,
+      "learning_rate": 3.953176860717488e-06,
+      "loss": 0.637,
+      "step": 9064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.071956191703273,
+      "learning_rate": 3.9521633670341005e-06,
+      "loss": 0.7031,
+      "step": 9065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5848015668086868,
+      "learning_rate": 3.9511499183785875e-06,
+      "loss": 0.7508,
+      "step": 9066
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5246930921494262,
+      "learning_rate": 3.950136514794499e-06,
+      "loss": 0.6393,
+      "step": 9067
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3169128400577632,
+      "learning_rate": 3.9491231563253815e-06,
+      "loss": 0.6648,
+      "step": 9068
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.748530273028535,
+      "learning_rate": 3.948109843014784e-06,
+      "loss": 0.6509,
+      "step": 9069
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8423332098857401,
+      "learning_rate": 3.947096574906248e-06,
+      "loss": 0.6857,
+      "step": 9070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8775678674393979,
+      "learning_rate": 3.946083352043318e-06,
+      "loss": 0.6758,
+      "step": 9071
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0523057768272157,
+      "learning_rate": 3.9450701744695325e-06,
+      "loss": 0.6408,
+      "step": 9072
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4212772882756806,
+      "learning_rate": 3.944057042228432e-06,
+      "loss": 0.6192,
+      "step": 9073
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6493677250770402,
+      "learning_rate": 3.9430439553635504e-06,
+      "loss": 0.6134,
+      "step": 9074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8619658086031636,
+      "learning_rate": 3.942030913918427e-06,
+      "loss": 0.8423,
+      "step": 9075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8198512041823034,
+      "learning_rate": 3.9410179179365895e-06,
+      "loss": 0.766,
+      "step": 9076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0906177370751937,
+      "learning_rate": 3.940004967461571e-06,
+      "loss": 0.6768,
+      "step": 9077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6139510791266212,
+      "learning_rate": 3.9389920625368996e-06,
+      "loss": 0.6409,
+      "step": 9078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.195320650259466,
+      "learning_rate": 3.937979203206103e-06,
+      "loss": 0.7117,
+      "step": 9079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.471101519882911,
+      "learning_rate": 3.936966389512703e-06,
+      "loss": 0.7188,
+      "step": 9080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5984766487320203,
+      "learning_rate": 3.935953621500226e-06,
+      "loss": 0.8166,
+      "step": 9081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7326146079086777,
+      "learning_rate": 3.934940899212193e-06,
+      "loss": 0.8064,
+      "step": 9082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4224937950193102,
+      "learning_rate": 3.9339282226921215e-06,
+      "loss": 0.7853,
+      "step": 9083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1232138837181604,
+      "learning_rate": 3.932915591983526e-06,
+      "loss": 0.7817,
+      "step": 9084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6336239730178208,
+      "learning_rate": 3.931903007129927e-06,
+      "loss": 0.8418,
+      "step": 9085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.681379540630682,
+      "learning_rate": 3.930890468174833e-06,
+      "loss": 0.6894,
+      "step": 9086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1187696740587818,
+      "learning_rate": 3.929877975161757e-06,
+      "loss": 0.7038,
+      "step": 9087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5224268491231918,
+      "learning_rate": 3.928865528134206e-06,
+      "loss": 0.7216,
+      "step": 9088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.48815955168306,
+      "learning_rate": 3.927853127135692e-06,
+      "loss": 0.9206,
+      "step": 9089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5522714314244825,
+      "learning_rate": 3.926840772209714e-06,
+      "loss": 0.7828,
+      "step": 9090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6767118269617451,
+      "learning_rate": 3.92582846339978e-06,
+      "loss": 0.7018,
+      "step": 9091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2596714171247334,
+      "learning_rate": 3.924816200749387e-06,
+      "loss": 0.6087,
+      "step": 9092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4878974267832412,
+      "learning_rate": 3.92380398430204e-06,
+      "loss": 0.7229,
+      "step": 9093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0062391566074347,
+      "learning_rate": 3.92279181410123e-06,
+      "loss": 0.6618,
+      "step": 9094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6472648256976383,
+      "learning_rate": 3.921779690190454e-06,
+      "loss": 0.6423,
+      "step": 9095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5704999184452668,
+      "learning_rate": 3.920767612613209e-06,
+      "loss": 0.7276,
+      "step": 9096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.106570046635567,
+      "learning_rate": 3.919755581412982e-06,
+      "loss": 0.7785,
+      "step": 9097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6498434534965385,
+      "learning_rate": 3.918743596633262e-06,
+      "loss": 0.6785,
+      "step": 9098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5608318981260911,
+      "learning_rate": 3.917731658317538e-06,
+      "loss": 0.6914,
+      "step": 9099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6048846139776534,
+      "learning_rate": 3.916719766509297e-06,
+      "loss": 0.7498,
+      "step": 9100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.847837631558765,
+      "learning_rate": 3.91570792125202e-06,
+      "loss": 0.8784,
+      "step": 9101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.188142627836074,
+      "learning_rate": 3.914696122589187e-06,
+      "loss": 0.7565,
+      "step": 9102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5514001938741353,
+      "learning_rate": 3.91368437056428e-06,
+      "loss": 0.666,
+      "step": 9103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6273424979924842,
+      "learning_rate": 3.912672665220773e-06,
+      "loss": 0.7238,
+      "step": 9104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2265512675013381,
+      "learning_rate": 3.911661006602144e-06,
+      "loss": 0.6582,
+      "step": 9105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7043127853601696,
+      "learning_rate": 3.910649394751867e-06,
+      "loss": 0.8273,
+      "step": 9106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9748310272763134,
+      "learning_rate": 3.9096378297134115e-06,
+      "loss": 0.8913,
+      "step": 9107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.912495306698274,
+      "learning_rate": 3.908626311530245e-06,
+      "loss": 0.7917,
+      "step": 9108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.632111733565036,
+      "learning_rate": 3.907614840245836e-06,
+      "loss": 0.6775,
+      "step": 9109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6732905873246398,
+      "learning_rate": 3.906603415903652e-06,
+      "loss": 0.7837,
+      "step": 9110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7757891708852098,
+      "learning_rate": 3.905592038547155e-06,
+      "loss": 0.6943,
+      "step": 9111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6506493577879748,
+      "learning_rate": 3.904580708219804e-06,
+      "loss": 0.8417,
+      "step": 9112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0846168342879687,
+      "learning_rate": 3.903569424965059e-06,
+      "loss": 0.6569,
+      "step": 9113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6856045359600555,
+      "learning_rate": 3.90255818882638e-06,
+      "loss": 0.7724,
+      "step": 9114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9473758570753366,
+      "learning_rate": 3.901546999847217e-06,
+      "loss": 0.7344,
+      "step": 9115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.497519892570605,
+      "learning_rate": 3.9005358580710275e-06,
+      "loss": 0.694,
+      "step": 9116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3778056645970667,
+      "learning_rate": 3.89952476354126e-06,
+      "loss": 0.8247,
+      "step": 9117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.786344888583225,
+      "learning_rate": 3.898513716301366e-06,
+      "loss": 0.7617,
+      "step": 9118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3553237508831912,
+      "learning_rate": 3.897502716394789e-06,
+      "loss": 0.5834,
+      "step": 9119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7213003370886968,
+      "learning_rate": 3.896491763864976e-06,
+      "loss": 0.6787,
+      "step": 9120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9414242700519404,
+      "learning_rate": 3.89548085875537e-06,
+      "loss": 0.6627,
+      "step": 9121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5562470189890318,
+      "learning_rate": 3.894470001109412e-06,
+      "loss": 0.686,
+      "step": 9122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7688164578865726,
+      "learning_rate": 3.8934591909705385e-06,
+      "loss": 0.6803,
+      "step": 9123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6123680441139905,
+      "learning_rate": 3.892448428382189e-06,
+      "loss": 0.6622,
+      "step": 9124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5286856633412915,
+      "learning_rate": 3.8914377133877976e-06,
+      "loss": 0.6804,
+      "step": 9125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.752296013912401,
+      "learning_rate": 3.890427046030796e-06,
+      "loss": 0.777,
+      "step": 9126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6671675440303335,
+      "learning_rate": 3.889416426354614e-06,
+      "loss": 0.7392,
+      "step": 9127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.612217926151794,
+      "learning_rate": 3.888405854402684e-06,
+      "loss": 0.6843,
+      "step": 9128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0335575866981535,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.6921,
+      "step": 9129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.211637885457524,
+      "learning_rate": 3.886384853845274e-06,
+      "loss": 0.64,
+      "step": 9130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5092555209610954,
+      "learning_rate": 3.885374425326642e-06,
+      "loss": 0.7414,
+      "step": 9131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.69130146019382,
+      "learning_rate": 3.884364044705955e-06,
+      "loss": 0.7094,
+      "step": 9132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6147292839765617,
+      "learning_rate": 3.883353712026629e-06,
+      "loss": 0.7203,
+      "step": 9133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8141300581122983,
+      "learning_rate": 3.8823434273320794e-06,
+      "loss": 0.7476,
+      "step": 9134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.532518986690597,
+      "learning_rate": 3.881333190665723e-06,
+      "loss": 0.6712,
+      "step": 9135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6019462158746791,
+      "learning_rate": 3.880323002070971e-06,
+      "loss": 0.7584,
+      "step": 9136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.693322297062379,
+      "learning_rate": 3.879312861591233e-06,
+      "loss": 0.7825,
+      "step": 9137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8366338898566767,
+      "learning_rate": 3.878302769269917e-06,
+      "loss": 0.8209,
+      "step": 9138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0191030324418695,
+      "learning_rate": 3.877292725150429e-06,
+      "loss": 0.881,
+      "step": 9139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9106695041219262,
+      "learning_rate": 3.876282729276174e-06,
+      "loss": 0.7939,
+      "step": 9140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.678890208565541,
+      "learning_rate": 3.875272781690552e-06,
+      "loss": 0.6829,
+      "step": 9141
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7919678370036203,
+      "learning_rate": 3.8742628824369624e-06,
+      "loss": 0.7039,
+      "step": 9142
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4117620910937585,
+      "learning_rate": 3.8732530315588065e-06,
+      "loss": 0.6084,
+      "step": 9143
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8662554660243973,
+      "learning_rate": 3.872243229099476e-06,
+      "loss": 0.7501,
+      "step": 9144
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6266534316004393,
+      "learning_rate": 3.871233475102365e-06,
+      "loss": 0.8092,
+      "step": 9145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8818047710980488,
+      "learning_rate": 3.870223769610865e-06,
+      "loss": 0.7471,
+      "step": 9146
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0514572494462975,
+      "learning_rate": 3.869214112668368e-06,
+      "loss": 0.841,
+      "step": 9147
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.59072957954758,
+      "learning_rate": 3.868204504318258e-06,
+      "loss": 0.6729,
+      "step": 9148
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9673817199804213,
+      "learning_rate": 3.86719494460392e-06,
+      "loss": 0.6847,
+      "step": 9149
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.758230908019945,
+      "learning_rate": 3.866185433568741e-06,
+      "loss": 0.7716,
+      "step": 9150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6863136282175393,
+      "learning_rate": 3.865175971256095e-06,
+      "loss": 0.7713,
+      "step": 9151
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0846738009487409,
+      "learning_rate": 3.864166557709368e-06,
+      "loss": 0.7074,
+      "step": 9152
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0315743064557417,
+      "learning_rate": 3.863157192971932e-06,
+      "loss": 0.8774,
+      "step": 9153
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.586497312426587,
+      "learning_rate": 3.8621478770871645e-06,
+      "loss": 0.7023,
+      "step": 9154
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6651716620785537,
+      "learning_rate": 3.861138610098435e-06,
+      "loss": 0.7292,
+      "step": 9155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4947350480088777,
+      "learning_rate": 3.8601293920491165e-06,
+      "loss": 0.666,
+      "step": 9156
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9720948098494837,
+      "learning_rate": 3.859120222982578e-06,
+      "loss": 0.7287,
+      "step": 9157
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.780743944083532,
+      "learning_rate": 3.858111102942184e-06,
+      "loss": 0.7932,
+      "step": 9158
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8425196600013916,
+      "learning_rate": 3.857102031971298e-06,
+      "loss": 0.7593,
+      "step": 9159
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8570644971472268,
+      "learning_rate": 3.856093010113282e-06,
+      "loss": 0.6613,
+      "step": 9160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7479889320217148,
+      "learning_rate": 3.855084037411501e-06,
+      "loss": 0.8708,
+      "step": 9161
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1151629660429934,
+      "learning_rate": 3.854075113909307e-06,
+      "loss": 0.8593,
+      "step": 9162
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6438911961542177,
+      "learning_rate": 3.853066239650055e-06,
+      "loss": 0.6871,
+      "step": 9163
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.891666512251198,
+      "learning_rate": 3.852057414677102e-06,
+      "loss": 0.8024,
+      "step": 9164
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7176333104397843,
+      "learning_rate": 3.851048639033801e-06,
+      "loss": 0.694,
+      "step": 9165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7965970467429793,
+      "learning_rate": 3.850039912763496e-06,
+      "loss": 0.6981,
+      "step": 9166
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7586366103437334,
+      "learning_rate": 3.849031235909539e-06,
+      "loss": 0.8038,
+      "step": 9167
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.643626535779331,
+      "learning_rate": 3.848022608515273e-06,
+      "loss": 0.7177,
+      "step": 9168
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6488860866032509,
+      "learning_rate": 3.84701403062404e-06,
+      "loss": 0.8107,
+      "step": 9169
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7128588152124873,
+      "learning_rate": 3.846005502279182e-06,
+      "loss": 0.5619,
+      "step": 9170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5322862094909961,
+      "learning_rate": 3.844997023524038e-06,
+      "loss": 0.6756,
+      "step": 9171
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7502043474964732,
+      "learning_rate": 3.843988594401946e-06,
+      "loss": 0.7544,
+      "step": 9172
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6452870386376768,
+      "learning_rate": 3.842980214956236e-06,
+      "loss": 0.7934,
+      "step": 9173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.623757327610563,
+      "learning_rate": 3.841971885230243e-06,
+      "loss": 0.6514,
+      "step": 9174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1900984976698978,
+      "learning_rate": 3.840963605267298e-06,
+      "loss": 0.6616,
+      "step": 9175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5652670385654874,
+      "learning_rate": 3.83995537511073e-06,
+      "loss": 0.7385,
+      "step": 9176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5085349837078026,
+      "learning_rate": 3.83894719480386e-06,
+      "loss": 0.6287,
+      "step": 9177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6198878738265952,
+      "learning_rate": 3.837939064390016e-06,
+      "loss": 0.7595,
+      "step": 9178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8037118074809926,
+      "learning_rate": 3.83693098391252e-06,
+      "loss": 0.6678,
+      "step": 9179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0703458543434752,
+      "learning_rate": 3.835922953414688e-06,
+      "loss": 0.687,
+      "step": 9180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8255312174500338,
+      "learning_rate": 3.83491497293984e-06,
+      "loss": 0.7423,
+      "step": 9181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.647199622110169,
+      "learning_rate": 3.833907042531289e-06,
+      "loss": 0.7463,
+      "step": 9182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5406126600763947,
+      "learning_rate": 3.832899162232352e-06,
+      "loss": 0.7561,
+      "step": 9183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6454986372658869,
+      "learning_rate": 3.8318913320863355e-06,
+      "loss": 0.6408,
+      "step": 9184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7169642424871663,
+      "learning_rate": 3.830883552136551e-06,
+      "loss": 0.6525,
+      "step": 9185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9860653631695786,
+      "learning_rate": 3.829875822426304e-06,
+      "loss": 0.6503,
+      "step": 9186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5117644669264563,
+      "learning_rate": 3.828868142998898e-06,
+      "loss": 0.6725,
+      "step": 9187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.768131576357141,
+      "learning_rate": 3.827860513897635e-06,
+      "loss": 0.7018,
+      "step": 9188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9586806759901674,
+      "learning_rate": 3.826852935165818e-06,
+      "loss": 0.7058,
+      "step": 9189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8260363294424595,
+      "learning_rate": 3.825845406846743e-06,
+      "loss": 0.7132,
+      "step": 9190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.696445206856576,
+      "learning_rate": 3.8248379289837065e-06,
+      "loss": 0.6433,
+      "step": 9191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.126818945716699,
+      "learning_rate": 3.8238305016199995e-06,
+      "loss": 0.7852,
+      "step": 9192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6263577707119081,
+      "learning_rate": 3.822823124798915e-06,
+      "loss": 0.7896,
+      "step": 9193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4974309810727189,
+      "learning_rate": 3.8218157985637465e-06,
+      "loss": 0.6999,
+      "step": 9194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7932160142781508,
+      "learning_rate": 3.820808522957774e-06,
+      "loss": 0.7562,
+      "step": 9195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.546559737247508,
+      "learning_rate": 3.819801298024286e-06,
+      "loss": 0.691,
+      "step": 9196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4270377573948791,
+      "learning_rate": 3.818794123806565e-06,
+      "loss": 0.6254,
+      "step": 9197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6186170896116394,
+      "learning_rate": 3.81778700034789e-06,
+      "loss": 0.773,
+      "step": 9198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6641332509925204,
+      "learning_rate": 3.816779927691542e-06,
+      "loss": 0.647,
+      "step": 9199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9898404480349559,
+      "learning_rate": 3.815772905880794e-06,
+      "loss": 0.669,
+      "step": 9200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.807941500272252,
+      "learning_rate": 3.814765934958924e-06,
+      "loss": 0.7118,
+      "step": 9201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.644193671753846,
+      "learning_rate": 3.8137590149691992e-06,
+      "loss": 0.8035,
+      "step": 9202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7579610783093247,
+      "learning_rate": 3.8127521459548934e-06,
+      "loss": 0.7074,
+      "step": 9203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.105686980086701,
+      "learning_rate": 3.811745327959271e-06,
+      "loss": 0.7167,
+      "step": 9204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1018728736501726,
+      "learning_rate": 3.810738561025599e-06,
+      "loss": 0.6421,
+      "step": 9205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0182001622920493,
+      "learning_rate": 3.8097318451971375e-06,
+      "loss": 0.6586,
+      "step": 9206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7028542333469505,
+      "learning_rate": 3.80872518051715e-06,
+      "loss": 0.6865,
+      "step": 9207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0012198459667054,
+      "learning_rate": 3.807718567028897e-06,
+      "loss": 0.6941,
+      "step": 9208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5683650183399729,
+      "learning_rate": 3.8067120047756313e-06,
+      "loss": 0.6709,
+      "step": 9209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7163258156356758,
+      "learning_rate": 3.8057054938006072e-06,
+      "loss": 0.683,
+      "step": 9210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.523243743446187,
+      "learning_rate": 3.804699034147078e-06,
+      "loss": 0.7029,
+      "step": 9211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6196605229722787,
+      "learning_rate": 3.803692625858295e-06,
+      "loss": 0.7579,
+      "step": 9212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.640874467009652,
+      "learning_rate": 3.8026862689775033e-06,
+      "loss": 0.6725,
+      "step": 9213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6753012576794002,
+      "learning_rate": 3.801679963547949e-06,
+      "loss": 0.6875,
+      "step": 9214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4993387075289537,
+      "learning_rate": 3.8006737096128763e-06,
+      "loss": 0.664,
+      "step": 9215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5334475691463338,
+      "learning_rate": 3.7996675072155243e-06,
+      "loss": 0.7119,
+      "step": 9216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4685159112357113,
+      "learning_rate": 3.7986613563991327e-06,
+      "loss": 0.6302,
+      "step": 9217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6718839227633233,
+      "learning_rate": 3.7976552572069385e-06,
+      "loss": 0.7018,
+      "step": 9218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.643659253249025,
+      "learning_rate": 3.7966492096821773e-06,
+      "loss": 0.7521,
+      "step": 9219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7004051301464216,
+      "learning_rate": 3.795643213868077e-06,
+      "loss": 0.682,
+      "step": 9220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8413979250337031,
+      "learning_rate": 3.7946372698078704e-06,
+      "loss": 0.8082,
+      "step": 9221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.585092595516021,
+      "learning_rate": 3.793631377544787e-06,
+      "loss": 0.7459,
+      "step": 9222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5785303239730513,
+      "learning_rate": 3.7926255371220487e-06,
+      "loss": 0.7527,
+      "step": 9223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6207383365062678,
+      "learning_rate": 3.7916197485828793e-06,
+      "loss": 0.6742,
+      "step": 9224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.640151900617584,
+      "learning_rate": 3.7906140119705006e-06,
+      "loss": 0.6242,
+      "step": 9225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5208383506619527,
+      "learning_rate": 3.7896083273281324e-06,
+      "loss": 0.7239,
+      "step": 9226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7167536207214389,
+      "learning_rate": 3.7886026946989896e-06,
+      "loss": 0.6337,
+      "step": 9227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3684491673379227,
+      "learning_rate": 3.787597114126286e-06,
+      "loss": 0.8364,
+      "step": 9228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7851994017497712,
+      "learning_rate": 3.786591585653235e-06,
+      "loss": 0.719,
+      "step": 9229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5727320104904838,
+      "learning_rate": 3.785586109323047e-06,
+      "loss": 0.762,
+      "step": 9230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.528344403616548,
+      "learning_rate": 3.7845806851789275e-06,
+      "loss": 0.6631,
+      "step": 9231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.643710491730836,
+      "learning_rate": 3.783575313264083e-06,
+      "loss": 0.6401,
+      "step": 9232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.658415704708341,
+      "learning_rate": 3.7825699936217183e-06,
+      "loss": 0.753,
+      "step": 9233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.646163240464668,
+      "learning_rate": 3.7815647262950293e-06,
+      "loss": 0.627,
+      "step": 9234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7809713447819453,
+      "learning_rate": 3.7805595113272186e-06,
+      "loss": 0.7289,
+      "step": 9235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1588796525329141,
+      "learning_rate": 3.7795543487614815e-06,
+      "loss": 0.6785,
+      "step": 9236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6922876999578051,
+      "learning_rate": 3.7785492386410138e-06,
+      "loss": 0.7548,
+      "step": 9237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.549341168576637,
+      "learning_rate": 3.7775441810090043e-06,
+      "loss": 0.7528,
+      "step": 9238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9817543155277988,
+      "learning_rate": 3.7765391759086424e-06,
+      "loss": 0.6986,
+      "step": 9239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7684461649817906,
+      "learning_rate": 3.7755342233831188e-06,
+      "loss": 0.7665,
+      "step": 9240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.660836599410642,
+      "learning_rate": 3.7745293234756164e-06,
+      "loss": 0.7247,
+      "step": 9241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7237151934250017,
+      "learning_rate": 3.7735244762293183e-06,
+      "loss": 0.6876,
+      "step": 9242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.026962034696045,
+      "learning_rate": 3.7725196816874045e-06,
+      "loss": 0.6901,
+      "step": 9243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.436296845703334,
+      "learning_rate": 3.771514939893055e-06,
+      "loss": 0.7442,
+      "step": 9244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1452864285684992,
+      "learning_rate": 3.7705102508894425e-06,
+      "loss": 0.635,
+      "step": 9245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3107016615396296,
+      "learning_rate": 3.769505614719744e-06,
+      "loss": 0.6839,
+      "step": 9246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.567301034841197,
+      "learning_rate": 3.7685010314271287e-06,
+      "loss": 0.67,
+      "step": 9247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.790381329575214,
+      "learning_rate": 3.7674965010547686e-06,
+      "loss": 0.7206,
+      "step": 9248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0511136863526578,
+      "learning_rate": 3.766492023645827e-06,
+      "loss": 0.6236,
+      "step": 9249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.795632791799927,
+      "learning_rate": 3.7654875992434714e-06,
+      "loss": 0.8041,
+      "step": 9250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6476880379850023,
+      "learning_rate": 3.764483227890864e-06,
+      "loss": 0.7204,
+      "step": 9251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9728683512078813,
+      "learning_rate": 3.763478909631162e-06,
+      "loss": 0.7017,
+      "step": 9252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8934523909062924,
+      "learning_rate": 3.7624746445075243e-06,
+      "loss": 0.668,
+      "step": 9253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6786154747561914,
+      "learning_rate": 3.761470432563109e-06,
+      "loss": 0.6719,
+      "step": 9254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.771573310069557,
+      "learning_rate": 3.760466273841068e-06,
+      "loss": 0.7099,
+      "step": 9255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.882857042972653,
+      "learning_rate": 3.759462168384551e-06,
+      "loss": 0.6873,
+      "step": 9256
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8041169662054324,
+      "learning_rate": 3.758458116236707e-06,
+      "loss": 0.7737,
+      "step": 9257
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.641243471290868,
+      "learning_rate": 3.757454117440682e-06,
+      "loss": 0.7725,
+      "step": 9258
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9024183907520806,
+      "learning_rate": 3.7564501720396242e-06,
+      "loss": 0.8135,
+      "step": 9259
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7473784184033254,
+      "learning_rate": 3.75544628007667e-06,
+      "loss": 0.8917,
+      "step": 9260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.672002785846439,
+      "learning_rate": 3.75444244159496e-06,
+      "loss": 0.7228,
+      "step": 9261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1030967542212309,
+      "learning_rate": 3.7534386566376345e-06,
+      "loss": 0.7468,
+      "step": 9262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.071315296231957,
+      "learning_rate": 3.7524349252478243e-06,
+      "loss": 0.6903,
+      "step": 9263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.301554688351584,
+      "learning_rate": 3.7514312474686643e-06,
+      "loss": 0.6165,
+      "step": 9264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7512371575777022,
+      "learning_rate": 3.7504276233432835e-06,
+      "loss": 0.7415,
+      "step": 9265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9558403949142569,
+      "learning_rate": 3.749424052914813e-06,
+      "loss": 0.7702,
+      "step": 9266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3092597874615066,
+      "learning_rate": 3.7484205362263737e-06,
+      "loss": 0.6265,
+      "step": 9267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7554724709042824,
+      "learning_rate": 3.747417073321092e-06,
+      "loss": 0.7242,
+      "step": 9268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0257128065013719,
+      "learning_rate": 3.74641366424209e-06,
+      "loss": 0.5599,
+      "step": 9269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6061950151179518,
+      "learning_rate": 3.7454103090324835e-06,
+      "loss": 0.6521,
+      "step": 9270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2057699883834716,
+      "learning_rate": 3.7444070077353896e-06,
+      "loss": 0.7174,
+      "step": 9271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6124077187305266,
+      "learning_rate": 3.743403760393922e-06,
+      "loss": 0.8357,
+      "step": 9272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6108800281250206,
+      "learning_rate": 3.7424005670511964e-06,
+      "loss": 0.7012,
+      "step": 9273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.782418183754326,
+      "learning_rate": 3.7413974277503183e-06,
+      "loss": 0.8137,
+      "step": 9274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7386292799509528,
+      "learning_rate": 3.740394342534394e-06,
+      "loss": 0.6479,
+      "step": 9275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8255683585269633,
+      "learning_rate": 3.739391311446531e-06,
+      "loss": 0.7808,
+      "step": 9276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8472801289018432,
+      "learning_rate": 3.738388334529832e-06,
+      "loss": 0.8093,
+      "step": 9277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8295383163656829,
+      "learning_rate": 3.7373854118273955e-06,
+      "loss": 0.7251,
+      "step": 9278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1640726544203375,
+      "learning_rate": 3.7363825433823187e-06,
+      "loss": 0.8054,
+      "step": 9279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8535188839739916,
+      "learning_rate": 3.7353797292377002e-06,
+      "loss": 0.6258,
+      "step": 9280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.777698620453768,
+      "learning_rate": 3.7343769694366294e-06,
+      "loss": 0.7165,
+      "step": 9281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9363459438253288,
+      "learning_rate": 3.7333742640221994e-06,
+      "loss": 0.725,
+      "step": 9282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7166493231667486,
+      "learning_rate": 3.732371613037498e-06,
+      "loss": 0.8074,
+      "step": 9283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7626593825507424,
+      "learning_rate": 3.7313690165256134e-06,
+      "loss": 0.7491,
+      "step": 9284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6561718037545974,
+      "learning_rate": 3.730366474529625e-06,
+      "loss": 0.7677,
+      "step": 9285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0752337729198564,
+      "learning_rate": 3.7293639870926173e-06,
+      "loss": 0.6972,
+      "step": 9286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.532431701108721,
+      "learning_rate": 3.728361554257671e-06,
+      "loss": 0.7341,
+      "step": 9287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7406753908521908,
+      "learning_rate": 3.7273591760678594e-06,
+      "loss": 0.7767,
+      "step": 9288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.677844047656967,
+      "learning_rate": 3.7263568525662574e-06,
+      "loss": 0.6837,
+      "step": 9289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6167821240944245,
+      "learning_rate": 3.7253545837959383e-06,
+      "loss": 0.7247,
+      "step": 9290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.137717027519479,
+      "learning_rate": 3.7243523697999733e-06,
+      "loss": 0.7494,
+      "step": 9291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0693188649937924,
+      "learning_rate": 3.723350210621426e-06,
+      "loss": 0.7164,
+      "step": 9292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7441917566165783,
+      "learning_rate": 3.722348106303364e-06,
+      "loss": 0.7459,
+      "step": 9293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6855596915059994,
+      "learning_rate": 3.7213460568888493e-06,
+      "loss": 0.6961,
+      "step": 9294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1320084433631696,
+      "learning_rate": 3.7203440624209438e-06,
+      "loss": 0.7706,
+      "step": 9295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7397276480856032,
+      "learning_rate": 3.7193421229427017e-06,
+      "loss": 0.7717,
+      "step": 9296
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2617491116337436,
+      "learning_rate": 3.7183402384971818e-06,
+      "loss": 0.6378,
+      "step": 9297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7778396071712277,
+      "learning_rate": 3.7173384091274372e-06,
+      "loss": 0.7351,
+      "step": 9298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7585272087433923,
+      "learning_rate": 3.716336634876516e-06,
+      "loss": 0.6957,
+      "step": 9299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.629597270196318,
+      "learning_rate": 3.7153349157874687e-06,
+      "loss": 0.619,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2505344473679894,
+      "learning_rate": 3.7143332519033413e-06,
+      "loss": 0.7821,
+      "step": 9301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1462644645552253,
+      "learning_rate": 3.7133316432671794e-06,
+      "loss": 0.7708,
+      "step": 9302
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8550216484007813,
+      "learning_rate": 3.7123300899220193e-06,
+      "loss": 0.7325,
+      "step": 9303
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7515873539468616,
+      "learning_rate": 3.711328591910904e-06,
+      "loss": 0.839,
+      "step": 9304
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7769523662572058,
+      "learning_rate": 3.710327149276871e-06,
+      "loss": 0.7007,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7156312011923203,
+      "learning_rate": 3.70932576206295e-06,
+      "loss": 0.7572,
+      "step": 9306
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1363215375249585,
+      "learning_rate": 3.708324430312177e-06,
+      "loss": 0.6124,
+      "step": 9307
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6328404143615236,
+      "learning_rate": 3.7073231540675792e-06,
+      "loss": 0.7584,
+      "step": 9308
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.451442010483691,
+      "learning_rate": 3.7063219333721857e-06,
+      "loss": 0.6272,
+      "step": 9309
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6679304028305761,
+      "learning_rate": 3.7053207682690184e-06,
+      "loss": 0.6887,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5277084108727141,
+      "learning_rate": 3.704319658801102e-06,
+      "loss": 0.6421,
+      "step": 9311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8708854501723473,
+      "learning_rate": 3.7033186050114547e-06,
+      "loss": 0.7126,
+      "step": 9312
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9092347117895405,
+      "learning_rate": 3.7023176069430965e-06,
+      "loss": 0.8209,
+      "step": 9313
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9250632368846843,
+      "learning_rate": 3.7013166646390384e-06,
+      "loss": 0.779,
+      "step": 9314
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6774675673211148,
+      "learning_rate": 3.700315778142297e-06,
+      "loss": 0.7647,
+      "step": 9315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1321924330143274,
+      "learning_rate": 3.699314947495881e-06,
+      "loss": 0.7445,
+      "step": 9316
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.787464267785241,
+      "learning_rate": 3.698314172742799e-06,
+      "loss": 0.7007,
+      "step": 9317
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6463119228354957,
+      "learning_rate": 3.6973134539260545e-06,
+      "loss": 0.6785,
+      "step": 9318
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6354188530371787,
+      "learning_rate": 3.6963127910886526e-06,
+      "loss": 0.7582,
+      "step": 9319
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5951666990479565,
+      "learning_rate": 3.695312184273595e-06,
+      "loss": 0.655,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7256861395762488,
+      "learning_rate": 3.694311633523878e-06,
+      "loss": 0.6462,
+      "step": 9321
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.840555174612165,
+      "learning_rate": 3.693311138882497e-06,
+      "loss": 0.685,
+      "step": 9322
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2214505548197478,
+      "learning_rate": 3.6923107003924486e-06,
+      "loss": 0.6321,
+      "step": 9323
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8182705929004548,
+      "learning_rate": 3.691310318096719e-06,
+      "loss": 0.6906,
+      "step": 9324
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6314886593545261,
+      "learning_rate": 3.690309992038301e-06,
+      "loss": 0.6371,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.755107957247885,
+      "learning_rate": 3.689309722260179e-06,
+      "loss": 0.7346,
+      "step": 9326
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.536755444135288,
+      "learning_rate": 3.6883095088053387e-06,
+      "loss": 0.6117,
+      "step": 9327
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0262997429410907,
+      "learning_rate": 3.6873093517167584e-06,
+      "loss": 0.7973,
+      "step": 9328
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0093974648668502,
+      "learning_rate": 3.6863092510374198e-06,
+      "loss": 0.6793,
+      "step": 9329
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8747804903524823,
+      "learning_rate": 3.685309206810297e-06,
+      "loss": 0.7223,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6000828175509854,
+      "learning_rate": 3.684309219078368e-06,
+      "loss": 0.7467,
+      "step": 9331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.871942675944839,
+      "learning_rate": 3.6833092878846e-06,
+      "loss": 0.6009,
+      "step": 9332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6415284562386487,
+      "learning_rate": 3.6823094132719645e-06,
+      "loss": 0.7473,
+      "step": 9333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8829599931885777,
+      "learning_rate": 3.68130959528343e-06,
+      "loss": 0.7991,
+      "step": 9334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.573873084413931,
+      "learning_rate": 3.680309833961958e-06,
+      "loss": 0.6104,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6624697243376887,
+      "learning_rate": 3.6793101293505106e-06,
+      "loss": 0.7348,
+      "step": 9336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9140745040596732,
+      "learning_rate": 3.678310481492048e-06,
+      "loss": 0.7829,
+      "step": 9337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6041945516960459,
+      "learning_rate": 3.6773108904295294e-06,
+      "loss": 0.7152,
+      "step": 9338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1195938273397903,
+      "learning_rate": 3.6763113562059077e-06,
+      "loss": 0.5738,
+      "step": 9339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6245045616624743,
+      "learning_rate": 3.6753118788641327e-06,
+      "loss": 0.683,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7170185390091048,
+      "learning_rate": 3.67431245844716e-06,
+      "loss": 0.8667,
+      "step": 9341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7850455604336677,
+      "learning_rate": 3.67331309499793e-06,
+      "loss": 0.7818,
+      "step": 9342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6787338217573822,
+      "learning_rate": 3.672313788559392e-06,
+      "loss": 0.7163,
+      "step": 9343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3918441012027167,
+      "learning_rate": 3.6713145391744877e-06,
+      "loss": 0.6228,
+      "step": 9344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6780583183373385,
+      "learning_rate": 3.6703153468861585e-06,
+      "loss": 0.7913,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.528364377180614,
+      "learning_rate": 3.6693162117373377e-06,
+      "loss": 0.6742,
+      "step": 9346
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8218146371588504,
+      "learning_rate": 3.668317133770963e-06,
+      "loss": 0.6501,
+      "step": 9347
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9816872206050313,
+      "learning_rate": 3.667318113029968e-06,
+      "loss": 0.6928,
+      "step": 9348
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8386816941419872,
+      "learning_rate": 3.6663191495572827e-06,
+      "loss": 0.7986,
+      "step": 9349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9980079990389024,
+      "learning_rate": 3.6653202433958313e-06,
+      "loss": 0.8192,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4075885774448105,
+      "learning_rate": 3.664321394588542e-06,
+      "loss": 0.6414,
+      "step": 9351
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6554115439769232,
+      "learning_rate": 3.663322603178339e-06,
+      "loss": 0.7353,
+      "step": 9352
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6498133467601623,
+      "learning_rate": 3.66232386920814e-06,
+      "loss": 0.6962,
+      "step": 9353
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.768058275460197,
+      "learning_rate": 3.661325192720862e-06,
+      "loss": 0.617,
+      "step": 9354
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.583086048189437,
+      "learning_rate": 3.660326573759422e-06,
+      "loss": 0.8862,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0066215227238298,
+      "learning_rate": 3.659328012366735e-06,
+      "loss": 0.6914,
+      "step": 9356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5440158277604437,
+      "learning_rate": 3.658329508585707e-06,
+      "loss": 0.7775,
+      "step": 9357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.125743659453796,
+      "learning_rate": 3.657331062459249e-06,
+      "loss": 0.6196,
+      "step": 9358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6084912430615852,
+      "learning_rate": 3.6563326740302664e-06,
+      "loss": 0.7209,
+      "step": 9359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8435886056842548,
+      "learning_rate": 3.6553343433416598e-06,
+      "loss": 0.6636,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7693608099524252,
+      "learning_rate": 3.6543360704363305e-06,
+      "loss": 0.7405,
+      "step": 9361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.82326927614921,
+      "learning_rate": 3.6533378553571786e-06,
+      "loss": 0.6781,
+      "step": 9362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.75269349383698,
+      "learning_rate": 3.6523396981470995e-06,
+      "loss": 0.6794,
+      "step": 9363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0514838201096017,
+      "learning_rate": 3.6513415988489824e-06,
+      "loss": 0.8326,
+      "step": 9364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7263044938441705,
+      "learning_rate": 3.6503435575057205e-06,
+      "loss": 0.7174,
+      "step": 9365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.775957940253427,
+      "learning_rate": 3.6493455741602035e-06,
+      "loss": 0.6021,
+      "step": 9366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9122231903140936,
+      "learning_rate": 3.6483476488553143e-06,
+      "loss": 0.6796,
+      "step": 9367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.658096245521902,
+      "learning_rate": 3.6473497816339375e-06,
+      "loss": 0.7704,
+      "step": 9368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1873578826772255,
+      "learning_rate": 3.6463519725389516e-06,
+      "loss": 0.7195,
+      "step": 9369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9508463951542465,
+      "learning_rate": 3.645354221613239e-06,
+      "loss": 0.7679,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6303180546095464,
+      "learning_rate": 3.6443565288996698e-06,
+      "loss": 0.6519,
+      "step": 9371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7146527599501662,
+      "learning_rate": 3.6433588944411207e-06,
+      "loss": 0.7287,
+      "step": 9372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.735212108251585,
+      "learning_rate": 3.642361318280461e-06,
+      "loss": 0.8471,
+      "step": 9373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7243795556158654,
+      "learning_rate": 3.64136380046056e-06,
+      "loss": 0.6354,
+      "step": 9374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.684421491668609,
+      "learning_rate": 3.640366341024281e-06,
+      "loss": 0.7454,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7616474807136344,
+      "learning_rate": 3.639368940014489e-06,
+      "loss": 0.7838,
+      "step": 9376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.81349202664025,
+      "learning_rate": 3.638371597474044e-06,
+      "loss": 0.7383,
+      "step": 9377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8457144331599642,
+      "learning_rate": 3.637374313445805e-06,
+      "loss": 0.707,
+      "step": 9378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5916819008514107,
+      "learning_rate": 3.6363770879726247e-06,
+      "loss": 0.6464,
+      "step": 9379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.382656230194411,
+      "learning_rate": 3.635379921097359e-06,
+      "loss": 0.5794,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7392579964396975,
+      "learning_rate": 3.634382812862857e-06,
+      "loss": 0.7303,
+      "step": 9381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.827409811139042,
+      "learning_rate": 3.6333857633119673e-06,
+      "loss": 0.7471,
+      "step": 9382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5176512455040314,
+      "learning_rate": 3.632388772487534e-06,
+      "loss": 0.6724,
+      "step": 9383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.628019566389735,
+      "learning_rate": 3.6313918404324e-06,
+      "loss": 0.65,
+      "step": 9384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5360374397990748,
+      "learning_rate": 3.63039496718941e-06,
+      "loss": 0.765,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8289428014419296,
+      "learning_rate": 3.6293981528013965e-06,
+      "loss": 0.7076,
+      "step": 9386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6146864084703112,
+      "learning_rate": 3.6284013973111962e-06,
+      "loss": 0.6874,
+      "step": 9387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0183439211694556,
+      "learning_rate": 3.627404700761645e-06,
+      "loss": 0.7455,
+      "step": 9388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.822698231357024,
+      "learning_rate": 3.6264080631955683e-06,
+      "loss": 0.7554,
+      "step": 9389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2595330589925506,
+      "learning_rate": 3.6254114846557977e-06,
+      "loss": 0.6683,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.160014910786491,
+      "learning_rate": 3.624414965185156e-06,
+      "loss": 0.7281,
+      "step": 9391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4387164267784531,
+      "learning_rate": 3.623418504826468e-06,
+      "loss": 0.7446,
+      "step": 9392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6632607574423814,
+      "learning_rate": 3.622422103622551e-06,
+      "loss": 0.6575,
+      "step": 9393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.753520943502266,
+      "learning_rate": 3.621425761616224e-06,
+      "loss": 0.6931,
+      "step": 9394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7444688630877678,
+      "learning_rate": 3.6204294788503034e-06,
+      "loss": 0.7675,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.789915950783939,
+      "learning_rate": 3.6194332553676026e-06,
+      "loss": 0.6962,
+      "step": 9396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7615634717095734,
+      "learning_rate": 3.618437091210927e-06,
+      "loss": 0.7231,
+      "step": 9397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5124418139061235,
+      "learning_rate": 3.617440986423086e-06,
+      "loss": 0.7456,
+      "step": 9398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4406651433401794,
+      "learning_rate": 3.616444941046887e-06,
+      "loss": 0.7254,
+      "step": 9399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5561690787135947,
+      "learning_rate": 3.615448955125129e-06,
+      "loss": 0.7136,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.179900583265979,
+      "learning_rate": 3.614453028700613e-06,
+      "loss": 0.7779,
+      "step": 9401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.855486892377156,
+      "learning_rate": 3.613457161816135e-06,
+      "loss": 0.7158,
+      "step": 9402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9013504033911313,
+      "learning_rate": 3.6124613545144937e-06,
+      "loss": 0.7324,
+      "step": 9403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0481013116872995,
+      "learning_rate": 3.6114656068384767e-06,
+      "loss": 0.6799,
+      "step": 9404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9524727268340174,
+      "learning_rate": 3.610469918830874e-06,
+      "loss": 0.8522,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.548478742051711,
+      "learning_rate": 3.6094742905344753e-06,
+      "loss": 0.8304,
+      "step": 9406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7173495552189382,
+      "learning_rate": 3.6084787219920616e-06,
+      "loss": 0.7163,
+      "step": 9407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4829638570447081,
+      "learning_rate": 3.6074832132464165e-06,
+      "loss": 0.7418,
+      "step": 9408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.887929906485753,
+      "learning_rate": 3.6064877643403194e-06,
+      "loss": 0.7672,
+      "step": 9409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.394193224681791,
+      "learning_rate": 3.605492375316548e-06,
+      "loss": 0.7043,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.55166481425126,
+      "learning_rate": 3.604497046217873e-06,
+      "loss": 0.7326,
+      "step": 9411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7233271409401318,
+      "learning_rate": 3.6035017770870672e-06,
+      "loss": 0.7069,
+      "step": 9412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.67071615433743,
+      "learning_rate": 3.6025065679669025e-06,
+      "loss": 0.6725,
+      "step": 9413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7862486981973487,
+      "learning_rate": 3.601511418900143e-06,
+      "loss": 0.8439,
+      "step": 9414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4582974630768375,
+      "learning_rate": 3.600516329929551e-06,
+      "loss": 0.7234,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.62105718003997,
+      "learning_rate": 3.599521301097888e-06,
+      "loss": 0.6708,
+      "step": 9416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5857242556141922,
+      "learning_rate": 3.598526332447917e-06,
+      "loss": 0.8058,
+      "step": 9417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.115894799611276,
+      "learning_rate": 3.597531424022388e-06,
+      "loss": 0.6879,
+      "step": 9418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6107942068562024,
+      "learning_rate": 3.5965365758640587e-06,
+      "loss": 0.6106,
+      "step": 9419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.468450726264781,
+      "learning_rate": 3.5955417880156766e-06,
+      "loss": 0.6326,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9637847192324938,
+      "learning_rate": 3.594547060519995e-06,
+      "loss": 0.7637,
+      "step": 9421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2462808412580797,
+      "learning_rate": 3.5935523934197537e-06,
+      "loss": 0.6004,
+      "step": 9422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7759071364719878,
+      "learning_rate": 3.592557786757699e-06,
+      "loss": 0.7167,
+      "step": 9423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.696120261752832,
+      "learning_rate": 3.591563240576572e-06,
+      "loss": 0.6954,
+      "step": 9424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5253389725576516,
+      "learning_rate": 3.590568754919107e-06,
+      "loss": 0.7207,
+      "step": 9425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6969271491933533,
+      "learning_rate": 3.5895743298280423e-06,
+      "loss": 0.8149,
+      "step": 9426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0559926772787052,
+      "learning_rate": 3.5885799653461107e-06,
+      "loss": 0.6095,
+      "step": 9427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6812735568299728,
+      "learning_rate": 3.587585661516042e-06,
+      "loss": 0.793,
+      "step": 9428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.836698836409467,
+      "learning_rate": 3.5865914183805606e-06,
+      "loss": 0.7337,
+      "step": 9429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8949897606901671,
+      "learning_rate": 3.585597235982394e-06,
+      "loss": 0.7728,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5104843293578605,
+      "learning_rate": 3.584603114364266e-06,
+      "loss": 0.6906,
+      "step": 9431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6846863883691419,
+      "learning_rate": 3.5836090535688944e-06,
+      "loss": 0.7644,
+      "step": 9432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6413922392914952,
+      "learning_rate": 3.582615053638996e-06,
+      "loss": 0.6365,
+      "step": 9433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1026437385083416,
+      "learning_rate": 3.581621114617284e-06,
+      "loss": 0.7207,
+      "step": 9434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7250764675245327,
+      "learning_rate": 3.580627236546474e-06,
+      "loss": 0.8688,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4427239545754775,
+      "learning_rate": 3.5796334194692704e-06,
+      "loss": 0.7245,
+      "step": 9436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6735821841908503,
+      "learning_rate": 3.578639663428384e-06,
+      "loss": 0.6994,
+      "step": 9437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0972645123621374,
+      "learning_rate": 3.5776459684665146e-06,
+      "loss": 0.6187,
+      "step": 9438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1352685309349042,
+      "learning_rate": 3.5766523346263682e-06,
+      "loss": 0.6696,
+      "step": 9439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6295612372392299,
+      "learning_rate": 3.5756587619506388e-06,
+      "loss": 0.6456,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.824681014411123,
+      "learning_rate": 3.5746652504820263e-06,
+      "loss": 0.7571,
+      "step": 9441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3160150007687694,
+      "learning_rate": 3.5736718002632222e-06,
+      "loss": 0.6973,
+      "step": 9442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9298125268545125,
+      "learning_rate": 3.572678411336916e-06,
+      "loss": 0.8839,
+      "step": 9443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.808905335860812,
+      "learning_rate": 3.571685083745798e-06,
+      "loss": 0.7192,
+      "step": 9444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8031688370591028,
+      "learning_rate": 3.5706918175325523e-06,
+      "loss": 0.7497,
+      "step": 9445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.977931097735382,
+      "learning_rate": 3.5696986127398646e-06,
+      "loss": 0.7375,
+      "step": 9446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.852189103654961,
+      "learning_rate": 3.5687054694104117e-06,
+      "loss": 0.7194,
+      "step": 9447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6921243652659173,
+      "learning_rate": 3.5677123875868713e-06,
+      "loss": 0.7309,
+      "step": 9448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2334568770204395,
+      "learning_rate": 3.56671936731192e-06,
+      "loss": 0.6745,
+      "step": 9449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6504950170700026,
+      "learning_rate": 3.5657264086282317e-06,
+      "loss": 0.6167,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7559299345509352,
+      "learning_rate": 3.564733511578472e-06,
+      "loss": 0.7446,
+      "step": 9451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0729411250883407,
+      "learning_rate": 3.5637406762053094e-06,
+      "loss": 0.7581,
+      "step": 9452
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6002622753567763,
+      "learning_rate": 3.562747902551411e-06,
+      "loss": 0.658,
+      "step": 9453
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6590851120711687,
+      "learning_rate": 3.561755190659434e-06,
+      "loss": 0.7557,
+      "step": 9454
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6132647954388897,
+      "learning_rate": 3.5607625405720402e-06,
+      "loss": 0.7353,
+      "step": 9455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4509726065038753,
+      "learning_rate": 3.559769952331884e-06,
+      "loss": 0.7361,
+      "step": 9456
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.512645140209209,
+      "learning_rate": 3.5587774259816234e-06,
+      "loss": 0.7491,
+      "step": 9457
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6441578477502343,
+      "learning_rate": 3.557784961563904e-06,
+      "loss": 0.7027,
+      "step": 9458
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7930506293983772,
+      "learning_rate": 3.556792559121377e-06,
+      "loss": 0.6149,
+      "step": 9459
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.585255250569748,
+      "learning_rate": 3.5558002186966906e-06,
+      "loss": 0.6698,
+      "step": 9460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3169376480438064,
+      "learning_rate": 3.554807940332483e-06,
+      "loss": 0.8159,
+      "step": 9461
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5261632106999332,
+      "learning_rate": 3.5538157240713964e-06,
+      "loss": 0.7423,
+      "step": 9462
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7995355522287668,
+      "learning_rate": 3.5528235699560697e-06,
+      "loss": 0.7028,
+      "step": 9463
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.856195900806251,
+      "learning_rate": 3.5518314780291384e-06,
+      "loss": 0.8795,
+      "step": 9464
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5384160139278078,
+      "learning_rate": 3.5508394483332333e-06,
+      "loss": 0.735,
+      "step": 9465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.105371523951052,
+      "learning_rate": 3.549847480910983e-06,
+      "loss": 0.6515,
+      "step": 9466
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0644578385897063,
+      "learning_rate": 3.548855575805016e-06,
+      "loss": 0.6321,
+      "step": 9467
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9745877197957162,
+      "learning_rate": 3.547863733057959e-06,
+      "loss": 0.5935,
+      "step": 9468
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6459317593678904,
+      "learning_rate": 3.5468719527124294e-06,
+      "loss": 0.7231,
+      "step": 9469
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.917967208523159,
+      "learning_rate": 3.5458802348110488e-06,
+      "loss": 0.7292,
+      "step": 9470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.742036180487708,
+      "learning_rate": 3.544888579396435e-06,
+      "loss": 0.6104,
+      "step": 9471
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5043572313006361,
+      "learning_rate": 3.543896986511197e-06,
+      "loss": 0.7529,
+      "step": 9472
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.715094088168675,
+      "learning_rate": 3.5429054561979485e-06,
+      "loss": 0.795,
+      "step": 9473
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.574525768211088,
+      "learning_rate": 3.541913988499299e-06,
+      "loss": 0.7348,
+      "step": 9474
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5264594984726951,
+      "learning_rate": 3.5409225834578536e-06,
+      "loss": 0.6888,
+      "step": 9475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6121672990276406,
+      "learning_rate": 3.5399312411162124e-06,
+      "loss": 0.6368,
+      "step": 9476
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1477586506546584,
+      "learning_rate": 3.5389399615169774e-06,
+      "loss": 0.6319,
+      "step": 9477
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7924594300306713,
+      "learning_rate": 3.5379487447027483e-06,
+      "loss": 0.6433,
+      "step": 9478
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7929988295069266,
+      "learning_rate": 3.5369575907161167e-06,
+      "loss": 0.707,
+      "step": 9479
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.532770830057696,
+      "learning_rate": 3.5359664995996746e-06,
+      "loss": 0.6633,
+      "step": 9480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7689895379980642,
+      "learning_rate": 3.5349754713960126e-06,
+      "loss": 0.7128,
+      "step": 9481
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0008332713099843,
+      "learning_rate": 3.5339845061477195e-06,
+      "loss": 0.7554,
+      "step": 9482
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.638380030219287,
+      "learning_rate": 3.5329936038973756e-06,
+      "loss": 0.957,
+      "step": 9483
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6969771450433508,
+      "learning_rate": 3.5320027646875643e-06,
+      "loss": 0.7547,
+      "step": 9484
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5114304544007395,
+      "learning_rate": 3.5310119885608625e-06,
+      "loss": 0.6473,
+      "step": 9485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7151923245927703,
+      "learning_rate": 3.5300212755598497e-06,
+      "loss": 0.6622,
+      "step": 9486
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.747406512892668,
+      "learning_rate": 3.5290306257270935e-06,
+      "loss": 0.7975,
+      "step": 9487
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9941387397246482,
+      "learning_rate": 3.5280400391051693e-06,
+      "loss": 0.7608,
+      "step": 9488
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3263097933613115,
+      "learning_rate": 3.5270495157366434e-06,
+      "loss": 0.781,
+      "step": 9489
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0953930474145266,
+      "learning_rate": 3.5260590556640783e-06,
+      "loss": 0.4507,
+      "step": 9490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3910252647335943,
+      "learning_rate": 3.5250686589300387e-06,
+      "loss": 0.6715,
+      "step": 9491
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.295939345982505,
+      "learning_rate": 3.524078325577084e-06,
+      "loss": 0.7581,
+      "step": 9492
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6555756636166612,
+      "learning_rate": 3.5230880556477707e-06,
+      "loss": 0.6756,
+      "step": 9493
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8984178758198247,
+      "learning_rate": 3.5220978491846534e-06,
+      "loss": 0.8119,
+      "step": 9494
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8024823166598143,
+      "learning_rate": 3.521107706230282e-06,
+      "loss": 0.7598,
+      "step": 9495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.14127749722859,
+      "learning_rate": 3.5201176268272058e-06,
+      "loss": 0.6957,
+      "step": 9496
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2036357652475194,
+      "learning_rate": 3.519127611017973e-06,
+      "loss": 0.761,
+      "step": 9497
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.654995492955756,
+      "learning_rate": 3.5181376588451244e-06,
+      "loss": 0.7246,
+      "step": 9498
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8793015918804423,
+      "learning_rate": 3.517147770351199e-06,
+      "loss": 0.8122,
+      "step": 9499
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6590864091145356,
+      "learning_rate": 3.5161579455787388e-06,
+      "loss": 0.6921,
+      "step": 9500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7502338404129711,
+      "learning_rate": 3.5151681845702747e-06,
+      "loss": 0.751,
+      "step": 9501
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9115360278278692,
+      "learning_rate": 3.5141784873683418e-06,
+      "loss": 0.707,
+      "step": 9502
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.581239679025022,
+      "learning_rate": 3.513188854015466e-06,
+      "loss": 0.6482,
+      "step": 9503
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7905983715030862,
+      "learning_rate": 3.5121992845541797e-06,
+      "loss": 0.8136,
+      "step": 9504
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8671286290559486,
+      "learning_rate": 3.5112097790270017e-06,
+      "loss": 0.7547,
+      "step": 9505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4444181603730475,
+      "learning_rate": 3.5102203374764555e-06,
+      "loss": 0.6226,
+      "step": 9506
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7303401325763605,
+      "learning_rate": 3.5092309599450596e-06,
+      "loss": 0.6436,
+      "step": 9507
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1814760548643795,
+      "learning_rate": 3.50824164647533e-06,
+      "loss": 0.6264,
+      "step": 9508
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.25519202196152,
+      "learning_rate": 3.507252397109777e-06,
+      "loss": 0.7548,
+      "step": 9509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2175701568705186,
+      "learning_rate": 3.506263211890914e-06,
+      "loss": 0.6022,
+      "step": 9510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7325791758467033,
+      "learning_rate": 3.505274090861249e-06,
+      "loss": 0.6647,
+      "step": 9511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.478403799433111,
+      "learning_rate": 3.504285034063285e-06,
+      "loss": 0.6124,
+      "step": 9512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5102040007567339,
+      "learning_rate": 3.503296041539522e-06,
+      "loss": 0.7324,
+      "step": 9513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4546962702654234,
+      "learning_rate": 3.5023071133324627e-06,
+      "loss": 0.6772,
+      "step": 9514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.779504269415885,
+      "learning_rate": 3.501318249484604e-06,
+      "loss": 0.7585,
+      "step": 9515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7885153513415226,
+      "learning_rate": 3.5003294500384367e-06,
+      "loss": 0.7018,
+      "step": 9516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.128600983678418,
+      "learning_rate": 3.4993407150364513e-06,
+      "loss": 0.7133,
+      "step": 9517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8253782842418143,
+      "learning_rate": 3.4983520445211403e-06,
+      "loss": 0.7451,
+      "step": 9518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5663318747699806,
+      "learning_rate": 3.497363438534984e-06,
+      "loss": 0.7524,
+      "step": 9519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5222769366377686,
+      "learning_rate": 3.496374897120467e-06,
+      "loss": 0.6955,
+      "step": 9520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0228335311193464,
+      "learning_rate": 3.495386420320071e-06,
+      "loss": 0.5616,
+      "step": 9521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5509358915987763,
+      "learning_rate": 3.4943980081762717e-06,
+      "loss": 0.6019,
+      "step": 9522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3719051765894354,
+      "learning_rate": 3.493409660731542e-06,
+      "loss": 0.7763,
+      "step": 9523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4169663107043868,
+      "learning_rate": 3.4924213780283545e-06,
+      "loss": 0.6013,
+      "step": 9524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9112958707818835,
+      "learning_rate": 3.4914331601091793e-06,
+      "loss": 0.7629,
+      "step": 9525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8526671269650332,
+      "learning_rate": 3.4904450070164803e-06,
+      "loss": 0.7483,
+      "step": 9526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.721514418416335,
+      "learning_rate": 3.4894569187927204e-06,
+      "loss": 0.5925,
+      "step": 9527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6514910934432985,
+      "learning_rate": 3.48846889548036e-06,
+      "loss": 0.7493,
+      "step": 9528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6121168571862283,
+      "learning_rate": 3.4874809371218608e-06,
+      "loss": 0.7216,
+      "step": 9529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8975708065840489,
+      "learning_rate": 3.486493043759673e-06,
+      "loss": 0.7351,
+      "step": 9530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5769645960170213,
+      "learning_rate": 3.4855052154362485e-06,
+      "loss": 0.6991,
+      "step": 9531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8412364949191038,
+      "learning_rate": 3.4845174521940384e-06,
+      "loss": 0.7708,
+      "step": 9532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7319982117890513,
+      "learning_rate": 3.4835297540754896e-06,
+      "loss": 0.6502,
+      "step": 9533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.779396954132934,
+      "learning_rate": 3.4825421211230437e-06,
+      "loss": 0.6817,
+      "step": 9534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.918858319823319,
+      "learning_rate": 3.481554553379143e-06,
+      "loss": 0.8933,
+      "step": 9535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7017534382196473,
+      "learning_rate": 3.4805670508862255e-06,
+      "loss": 0.6173,
+      "step": 9536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8674961324995922,
+      "learning_rate": 3.4795796136867237e-06,
+      "loss": 0.7194,
+      "step": 9537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7382204754834805,
+      "learning_rate": 3.4785922418230722e-06,
+      "loss": 0.6366,
+      "step": 9538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1176775692258105,
+      "learning_rate": 3.4776049353377016e-06,
+      "loss": 0.6965,
+      "step": 9539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.985919385007966,
+      "learning_rate": 3.476617694273039e-06,
+      "loss": 0.6874,
+      "step": 9540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.683243060561715,
+      "learning_rate": 3.4756305186715046e-06,
+      "loss": 0.6848,
+      "step": 9541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.740980088997226,
+      "learning_rate": 3.474643408575521e-06,
+      "loss": 0.7407,
+      "step": 9542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3922349386316006,
+      "learning_rate": 3.47365636402751e-06,
+      "loss": 0.7693,
+      "step": 9543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9524244049233257,
+      "learning_rate": 3.4726693850698824e-06,
+      "loss": 0.927,
+      "step": 9544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5918871515531614,
+      "learning_rate": 3.4716824717450538e-06,
+      "loss": 0.644,
+      "step": 9545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8154508098096396,
+      "learning_rate": 3.4706956240954317e-06,
+      "loss": 0.7239,
+      "step": 9546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5883604020739746,
+      "learning_rate": 3.4697088421634266e-06,
+      "loss": 0.705,
+      "step": 9547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6611522502363831,
+      "learning_rate": 3.4687221259914394e-06,
+      "loss": 0.6163,
+      "step": 9548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8447079822799972,
+      "learning_rate": 3.467735475621873e-06,
+      "loss": 0.7694,
+      "step": 9549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5630443550140998,
+      "learning_rate": 3.4667488910971252e-06,
+      "loss": 0.7071,
+      "step": 9550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.529071061194679,
+      "learning_rate": 3.4657623724595944e-06,
+      "loss": 0.7848,
+      "step": 9551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.614221744592092,
+      "learning_rate": 3.4647759197516687e-06,
+      "loss": 0.6472,
+      "step": 9552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7816493089888688,
+      "learning_rate": 3.4637895330157423e-06,
+      "loss": 0.7443,
+      "step": 9553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.747428295523427,
+      "learning_rate": 3.4628032122942024e-06,
+      "loss": 0.7868,
+      "step": 9554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7340532192176779,
+      "learning_rate": 3.461816957629429e-06,
+      "loss": 0.7349,
+      "step": 9555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7223651257043877,
+      "learning_rate": 3.4608307690638075e-06,
+      "loss": 0.5339,
+      "step": 9556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7211474357016137,
+      "learning_rate": 3.459844646639717e-06,
+      "loss": 0.7124,
+      "step": 9557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.571557300589324,
+      "learning_rate": 3.4588585903995314e-06,
+      "loss": 0.6786,
+      "step": 9558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2700851113134068,
+      "learning_rate": 3.4578726003856245e-06,
+      "loss": 0.5585,
+      "step": 9559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6635501920265925,
+      "learning_rate": 3.4568866766403665e-06,
+      "loss": 0.7942,
+      "step": 9560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7211878150270585,
+      "learning_rate": 3.4559008192061254e-06,
+      "loss": 0.7151,
+      "step": 9561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7331532350081846,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.6689,
+      "step": 9562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.671381303570133,
+      "learning_rate": 3.453929303440145e-06,
+      "loss": 0.6493,
+      "step": 9563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.803868268609347,
+      "learning_rate": 3.4529436451931263e-06,
+      "loss": 0.7191,
+      "step": 9564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0118915703398397,
+      "learning_rate": 3.4519580534265674e-06,
+      "loss": 0.8336,
+      "step": 9565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6683172382371971,
+      "learning_rate": 3.4509725281828155e-06,
+      "loss": 0.8267,
+      "step": 9566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6968884191384115,
+      "learning_rate": 3.449987069504226e-06,
+      "loss": 0.6703,
+      "step": 9567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.511078493372515,
+      "learning_rate": 3.4490016774331426e-06,
+      "loss": 0.6513,
+      "step": 9568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.697210621472526,
+      "learning_rate": 3.448016352011914e-06,
+      "loss": 0.6669,
+      "step": 9569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2020617106023432,
+      "learning_rate": 3.4470310932828766e-06,
+      "loss": 0.6592,
+      "step": 9570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5273270942570907,
+      "learning_rate": 3.4460459012883716e-06,
+      "loss": 0.6879,
+      "step": 9571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3140771412358332,
+      "learning_rate": 3.4450607760707373e-06,
+      "loss": 0.7251,
+      "step": 9572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2084765655135374,
+      "learning_rate": 3.4440757176723037e-06,
+      "loss": 0.8857,
+      "step": 9573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4403995956383524,
+      "learning_rate": 3.4430907261354e-06,
+      "loss": 0.6352,
+      "step": 9574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6385297329836594,
+      "learning_rate": 3.442105801502356e-06,
+      "loss": 0.6229,
+      "step": 9575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7178903577023537,
+      "learning_rate": 3.441120943815497e-06,
+      "loss": 0.7296,
+      "step": 9576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.667278296947427,
+      "learning_rate": 3.440136153117142e-06,
+      "loss": 0.9098,
+      "step": 9577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7129467318333016,
+      "learning_rate": 3.4391514294496097e-06,
+      "loss": 0.6699,
+      "step": 9578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6662471827955285,
+      "learning_rate": 3.438166772855218e-06,
+      "loss": 0.6757,
+      "step": 9579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5623419222909498,
+      "learning_rate": 3.4371821833762776e-06,
+      "loss": 0.8102,
+      "step": 9580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6785581454505105,
+      "learning_rate": 3.4361976610550996e-06,
+      "loss": 0.7069,
+      "step": 9581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.638362847148322,
+      "learning_rate": 3.43521320593399e-06,
+      "loss": 0.5477,
+      "step": 9582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7354738565155858,
+      "learning_rate": 3.4342288180552556e-06,
+      "loss": 0.731,
+      "step": 9583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7454245676275553,
+      "learning_rate": 3.4332444974611946e-06,
+      "loss": 0.6946,
+      "step": 9584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7319800927483098,
+      "learning_rate": 3.4322602441941064e-06,
+      "loss": 0.7004,
+      "step": 9585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5448955076931392,
+      "learning_rate": 3.4312760582962877e-06,
+      "loss": 0.6463,
+      "step": 9586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7240609448418749,
+      "learning_rate": 3.4302919398100325e-06,
+      "loss": 0.7533,
+      "step": 9587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6733717971565516,
+      "learning_rate": 3.4293078887776264e-06,
+      "loss": 0.7375,
+      "step": 9588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6674931091810812,
+      "learning_rate": 3.428323905241358e-06,
+      "loss": 0.6876,
+      "step": 9589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5191939354081496,
+      "learning_rate": 3.427339989243514e-06,
+      "loss": 0.5766,
+      "step": 9590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0490341892348174,
+      "learning_rate": 3.426356140826372e-06,
+      "loss": 0.5849,
+      "step": 9591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7934863712004732,
+      "learning_rate": 3.4253723600322105e-06,
+      "loss": 0.733,
+      "step": 9592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1020986324986175,
+      "learning_rate": 3.424388646903305e-06,
+      "loss": 0.7016,
+      "step": 9593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6756318143243818,
+      "learning_rate": 3.4234050014819308e-06,
+      "loss": 0.6984,
+      "step": 9594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7570280144463852,
+      "learning_rate": 3.422421423810354e-06,
+      "loss": 0.7042,
+      "step": 9595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.598960963870466,
+      "learning_rate": 3.421437913930842e-06,
+      "loss": 0.641,
+      "step": 9596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3283892401512696,
+      "learning_rate": 3.420454471885659e-06,
+      "loss": 0.7383,
+      "step": 9597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.767799548237936,
+      "learning_rate": 3.4194710977170635e-06,
+      "loss": 0.841,
+      "step": 9598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7168379794613584,
+      "learning_rate": 3.4184877914673155e-06,
+      "loss": 0.7559,
+      "step": 9599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.859933702665036,
+      "learning_rate": 3.41750455317867e-06,
+      "loss": 0.8009,
+      "step": 9600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9463489455766367,
+      "learning_rate": 3.4165213828933797e-06,
+      "loss": 0.6511,
+      "step": 9601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9785734165019949,
+      "learning_rate": 3.4155382806536897e-06,
+      "loss": 0.5094,
+      "step": 9602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8114823469830592,
+      "learning_rate": 3.4145552465018483e-06,
+      "loss": 0.8767,
+      "step": 9603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.275843830648734,
+      "learning_rate": 3.4135722804801004e-06,
+      "loss": 0.7343,
+      "step": 9604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5354844895083857,
+      "learning_rate": 3.4125893826306855e-06,
+      "loss": 0.6533,
+      "step": 9605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7700551296609515,
+      "learning_rate": 3.411606552995838e-06,
+      "loss": 0.6624,
+      "step": 9606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.095221164339161,
+      "learning_rate": 3.4106237916177954e-06,
+      "loss": 0.6359,
+      "step": 9607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7405546089291337,
+      "learning_rate": 3.4096410985387898e-06,
+      "loss": 0.6762,
+      "step": 9608
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6098154245414835,
+      "learning_rate": 3.4086584738010455e-06,
+      "loss": 0.7139,
+      "step": 9609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9553180816646891,
+      "learning_rate": 3.4076759174467923e-06,
+      "loss": 0.6771,
+      "step": 9610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0380915332084966,
+      "learning_rate": 3.4066934295182496e-06,
+      "loss": 0.7429,
+      "step": 9611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6496548810605387,
+      "learning_rate": 3.4057110100576408e-06,
+      "loss": 0.8535,
+      "step": 9612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.771184618530829,
+      "learning_rate": 3.4047286591071794e-06,
+      "loss": 0.6907,
+      "step": 9613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8021059654193916,
+      "learning_rate": 3.4037463767090807e-06,
+      "loss": 0.6223,
+      "step": 9614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2610095117937545,
+      "learning_rate": 3.402764162905554e-06,
+      "loss": 0.6721,
+      "step": 9615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6243704252176694,
+      "learning_rate": 3.401782017738811e-06,
+      "loss": 0.7002,
+      "step": 9616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6863460883807606,
+      "learning_rate": 3.400799941251052e-06,
+      "loss": 0.7037,
+      "step": 9617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5174598608181116,
+      "learning_rate": 3.3998179334844823e-06,
+      "loss": 0.6366,
+      "step": 9618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5564863287130282,
+      "learning_rate": 3.3988359944812997e-06,
+      "loss": 0.746,
+      "step": 9619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7334611613806232,
+      "learning_rate": 3.3978541242837003e-06,
+      "loss": 0.7057,
+      "step": 9620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5722697808350272,
+      "learning_rate": 3.396872322933876e-06,
+      "loss": 0.6692,
+      "step": 9621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5760234661445203,
+      "learning_rate": 3.395890590474019e-06,
+      "loss": 0.6971,
+      "step": 9622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4692155751817104,
+      "learning_rate": 3.3949089269463178e-06,
+      "loss": 0.626,
+      "step": 9623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7712534556763946,
+      "learning_rate": 3.3939273323929533e-06,
+      "loss": 0.7318,
+      "step": 9624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5435793240915823,
+      "learning_rate": 3.3929458068561073e-06,
+      "loss": 0.7622,
+      "step": 9625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9774432996777025,
+      "learning_rate": 3.391964350377961e-06,
+      "loss": 0.6581,
+      "step": 9626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.152158468571261,
+      "learning_rate": 3.3909829630006865e-06,
+      "loss": 0.6851,
+      "step": 9627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7961963273520334,
+      "learning_rate": 3.390001644766458e-06,
+      "loss": 0.6523,
+      "step": 9628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7230948792242118,
+      "learning_rate": 3.3890203957174437e-06,
+      "loss": 0.7932,
+      "step": 9629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6026171790048118,
+      "learning_rate": 3.388039215895813e-06,
+      "loss": 0.6782,
+      "step": 9630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5885317339949352,
+      "learning_rate": 3.3870581053437244e-06,
+      "loss": 0.7234,
+      "step": 9631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5416734960567635,
+      "learning_rate": 3.3860770641033417e-06,
+      "loss": 0.6397,
+      "step": 9632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7282948333857324,
+      "learning_rate": 3.385096092216823e-06,
+      "loss": 0.7725,
+      "step": 9633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8332363722671894,
+      "learning_rate": 3.3841151897263234e-06,
+      "loss": 0.741,
+      "step": 9634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7579005537594685,
+      "learning_rate": 3.3831343566739904e-06,
+      "loss": 0.7937,
+      "step": 9635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8133549377690543,
+      "learning_rate": 3.3821535931019744e-06,
+      "loss": 0.8079,
+      "step": 9636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7441243997631113,
+      "learning_rate": 3.3811728990524245e-06,
+      "loss": 0.8112,
+      "step": 9637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.62223241719708,
+      "learning_rate": 3.380192274567479e-06,
+      "loss": 0.7038,
+      "step": 9638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3978370118048926,
+      "learning_rate": 3.379211719689278e-06,
+      "loss": 0.6631,
+      "step": 9639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4443576867918948,
+      "learning_rate": 3.3782312344599598e-06,
+      "loss": 0.7386,
+      "step": 9640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.712368817742478,
+      "learning_rate": 3.3772508189216586e-06,
+      "loss": 0.6961,
+      "step": 9641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9878686124680667,
+      "learning_rate": 3.376270473116503e-06,
+      "loss": 0.8252,
+      "step": 9642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.627682818700191,
+      "learning_rate": 3.37529019708662e-06,
+      "loss": 0.7267,
+      "step": 9643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1234315439681417,
+      "learning_rate": 3.3743099908741385e-06,
+      "loss": 0.8201,
+      "step": 9644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.756162039505631,
+      "learning_rate": 3.3733298545211748e-06,
+      "loss": 0.7697,
+      "step": 9645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4738630137402535,
+      "learning_rate": 3.37234978806985e-06,
+      "loss": 0.6616,
+      "step": 9646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6979099653483223,
+      "learning_rate": 3.3713697915622806e-06,
+      "loss": 0.731,
+      "step": 9647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6285901365515647,
+      "learning_rate": 3.3703898650405796e-06,
+      "loss": 0.6978,
+      "step": 9648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0426478053996338,
+      "learning_rate": 3.3694100085468535e-06,
+      "loss": 0.684,
+      "step": 9649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8098819078246673,
+      "learning_rate": 3.3684302221232106e-06,
+      "loss": 0.7845,
+      "step": 9650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.770674605934243,
+      "learning_rate": 3.3674505058117556e-06,
+      "loss": 0.8448,
+      "step": 9651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5754103411259146,
+      "learning_rate": 3.36647085965459e-06,
+      "loss": 0.6691,
+      "step": 9652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6753228374263298,
+      "learning_rate": 3.365491283693807e-06,
+      "loss": 0.7136,
+      "step": 9653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8124528954636936,
+      "learning_rate": 3.364511777971504e-06,
+      "loss": 0.9076,
+      "step": 9654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9020299757259544,
+      "learning_rate": 3.363532342529775e-06,
+      "loss": 0.8501,
+      "step": 9655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.023349968081234,
+      "learning_rate": 3.3625529774107028e-06,
+      "loss": 0.6953,
+      "step": 9656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.674076639485243,
+      "learning_rate": 3.361573682656377e-06,
+      "loss": 0.7325,
+      "step": 9657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6657967998745489,
+      "learning_rate": 3.360594458308879e-06,
+      "loss": 0.7267,
+      "step": 9658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1295719482604674,
+      "learning_rate": 3.3596153044102897e-06,
+      "loss": 0.7352,
+      "step": 9659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5655793359277153,
+      "learning_rate": 3.358636221002682e-06,
+      "loss": 0.8144,
+      "step": 9660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0234867273944535,
+      "learning_rate": 3.357657208128132e-06,
+      "loss": 0.6945,
+      "step": 9661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.602716414854916,
+      "learning_rate": 3.356678265828711e-06,
+      "loss": 0.7811,
+      "step": 9662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.574868965604673,
+      "learning_rate": 3.3556993941464834e-06,
+      "loss": 0.699,
+      "step": 9663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.588235826849425,
+      "learning_rate": 3.354720593123514e-06,
+      "loss": 0.6416,
+      "step": 9664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7202454627164483,
+      "learning_rate": 3.353741862801866e-06,
+      "loss": 0.7755,
+      "step": 9665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5482264636005951,
+      "learning_rate": 3.352763203223598e-06,
+      "loss": 0.6739,
+      "step": 9666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6820525889212057,
+      "learning_rate": 3.351784614430761e-06,
+      "loss": 0.8066,
+      "step": 9667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7853634732967214,
+      "learning_rate": 3.3508060964654106e-06,
+      "loss": 0.742,
+      "step": 9668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8671394029635502,
+      "learning_rate": 3.349827649369596e-06,
+      "loss": 0.7574,
+      "step": 9669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.12908079319576,
+      "learning_rate": 3.348849273185362e-06,
+      "loss": 0.6176,
+      "step": 9670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8354898302583253,
+      "learning_rate": 3.3478709679547527e-06,
+      "loss": 0.7857,
+      "step": 9671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9577241873159223,
+      "learning_rate": 3.346892733719806e-06,
+      "loss": 0.7296,
+      "step": 9672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6737285747795827,
+      "learning_rate": 3.3459145705225627e-06,
+      "loss": 0.6673,
+      "step": 9673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.794866362051434,
+      "learning_rate": 3.3449364784050515e-06,
+      "loss": 0.7384,
+      "step": 9674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9043865588829503,
+      "learning_rate": 3.343958457409307e-06,
+      "loss": 0.7116,
+      "step": 9675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4190688399293585,
+      "learning_rate": 3.3429805075773547e-06,
+      "loss": 0.8082,
+      "step": 9676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.616961869978645,
+      "learning_rate": 3.3420026289512232e-06,
+      "loss": 0.8862,
+      "step": 9677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9742339160154405,
+      "learning_rate": 3.341024821572928e-06,
+      "loss": 0.8536,
+      "step": 9678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7803164665371691,
+      "learning_rate": 3.3400470854844925e-06,
+      "loss": 0.6545,
+      "step": 9679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6865743163874476,
+      "learning_rate": 3.3390694207279308e-06,
+      "loss": 0.7755,
+      "step": 9680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6908794694814533,
+      "learning_rate": 3.3380918273452557e-06,
+      "loss": 0.6419,
+      "step": 9681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4947492085495262,
+      "learning_rate": 3.3371143053784737e-06,
+      "loss": 0.6538,
+      "step": 9682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4551210881755405,
+      "learning_rate": 3.336136854869595e-06,
+      "loss": 0.7171,
+      "step": 9683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0963572346731454,
+      "learning_rate": 3.3351594758606222e-06,
+      "loss": 0.6072,
+      "step": 9684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8145518764199213,
+      "learning_rate": 3.334182168393554e-06,
+      "loss": 0.7143,
+      "step": 9685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4866718700388764,
+      "learning_rate": 3.3332049325103867e-06,
+      "loss": 0.7357,
+      "step": 9686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5934983672857514,
+      "learning_rate": 3.332227768253116e-06,
+      "loss": 0.6952,
+      "step": 9687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.266384617103505,
+      "learning_rate": 3.3312506756637343e-06,
+      "loss": 0.6963,
+      "step": 9688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7298514743759856,
+      "learning_rate": 3.3302736547842263e-06,
+      "loss": 0.7027,
+      "step": 9689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5991263603150137,
+      "learning_rate": 3.3292967056565766e-06,
+      "loss": 0.7356,
+      "step": 9690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3721096461121387,
+      "learning_rate": 3.3283198283227715e-06,
+      "loss": 0.6396,
+      "step": 9691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5362459582568895,
+      "learning_rate": 3.3273430228247834e-06,
+      "loss": 0.8229,
+      "step": 9692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.438196452150864,
+      "learning_rate": 3.3263662892045924e-06,
+      "loss": 0.6923,
+      "step": 9693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7515611984912438,
+      "learning_rate": 3.3253896275041677e-06,
+      "loss": 0.7018,
+      "step": 9694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5446585817614182,
+      "learning_rate": 3.324413037765483e-06,
+      "loss": 0.5963,
+      "step": 9695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9317892583804102,
+      "learning_rate": 3.3234365200304997e-06,
+      "loss": 0.8548,
+      "step": 9696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5689268093500843,
+      "learning_rate": 3.322460074341183e-06,
+      "loss": 0.5951,
+      "step": 9697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9004301888832593,
+      "learning_rate": 3.321483700739494e-06,
+      "loss": 0.76,
+      "step": 9698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.962229907207434,
+      "learning_rate": 3.3205073992673885e-06,
+      "loss": 0.6237,
+      "step": 9699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4356412590146768,
+      "learning_rate": 3.31953116996682e-06,
+      "loss": 0.6551,
+      "step": 9700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.036256853459456,
+      "learning_rate": 3.3185550128797387e-06,
+      "loss": 0.6781,
+      "step": 9701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7149844407330623,
+      "learning_rate": 3.317578928048096e-06,
+      "loss": 0.7921,
+      "step": 9702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2598732938267616,
+      "learning_rate": 3.3166029155138324e-06,
+      "loss": 0.6589,
+      "step": 9703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0477667435641649,
+      "learning_rate": 3.3156269753188895e-06,
+      "loss": 0.6104,
+      "step": 9704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1819674102499635,
+      "learning_rate": 3.314651107505207e-06,
+      "loss": 0.623,
+      "step": 9705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.202145327723715,
+      "learning_rate": 3.313675312114722e-06,
+      "loss": 0.6559,
+      "step": 9706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.497435088553448,
+      "learning_rate": 3.312699589189362e-06,
+      "loss": 0.7051,
+      "step": 9707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.743223519227618,
+      "learning_rate": 3.31172393877106e-06,
+      "loss": 0.791,
+      "step": 9708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.082864119468584,
+      "learning_rate": 3.310748360901741e-06,
+      "loss": 0.731,
+      "step": 9709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.215655781509481,
+      "learning_rate": 3.3097728556233256e-06,
+      "loss": 0.6285,
+      "step": 9710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.800578892850952,
+      "learning_rate": 3.308797422977734e-06,
+      "loss": 0.7313,
+      "step": 9711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6861853398662296,
+      "learning_rate": 3.307822063006886e-06,
+      "loss": 0.6886,
+      "step": 9712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.007058933727683,
+      "learning_rate": 3.3068467757526927e-06,
+      "loss": 0.8765,
+      "step": 9713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7825667701619303,
+      "learning_rate": 3.3058715612570623e-06,
+      "loss": 0.6306,
+      "step": 9714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6481537323036735,
+      "learning_rate": 3.3048964195619044e-06,
+      "loss": 0.8077,
+      "step": 9715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6399878250999478,
+      "learning_rate": 3.303921350709124e-06,
+      "loss": 0.839,
+      "step": 9716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.546062201192303,
+      "learning_rate": 3.3029463547406198e-06,
+      "loss": 0.7993,
+      "step": 9717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5655649247947285,
+      "learning_rate": 3.30197143169829e-06,
+      "loss": 0.687,
+      "step": 9718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5931209543537526,
+      "learning_rate": 3.300996581624028e-06,
+      "loss": 0.9669,
+      "step": 9719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6352417690228445,
+      "learning_rate": 3.30002180455973e-06,
+      "loss": 0.6745,
+      "step": 9720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8073910192668499,
+      "learning_rate": 3.299047100547278e-06,
+      "loss": 0.6665,
+      "step": 9721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7883429754658513,
+      "learning_rate": 3.2980724696285626e-06,
+      "loss": 0.7729,
+      "step": 9722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6916884978898106,
+      "learning_rate": 3.2970979118454616e-06,
+      "loss": 0.6918,
+      "step": 9723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1249387323511595,
+      "learning_rate": 3.2961234272398578e-06,
+      "loss": 0.5805,
+      "step": 9724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.526030860077004,
+      "learning_rate": 3.2951490158536238e-06,
+      "loss": 0.7348,
+      "step": 9725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4563445414115015,
+      "learning_rate": 3.2941746777286344e-06,
+      "loss": 0.659,
+      "step": 9726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6397512101392022,
+      "learning_rate": 3.2932004129067592e-06,
+      "loss": 0.6266,
+      "step": 9727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5675495539612463,
+      "learning_rate": 3.2922262214298613e-06,
+      "loss": 0.7339,
+      "step": 9728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4425844670057275,
+      "learning_rate": 3.291252103339806e-06,
+      "loss": 0.7061,
+      "step": 9729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8759634842183324,
+      "learning_rate": 3.2902780586784542e-06,
+      "loss": 0.7282,
+      "step": 9730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6269388502245394,
+      "learning_rate": 3.2893040874876624e-06,
+      "loss": 0.7139,
+      "step": 9731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6528889627622447,
+      "learning_rate": 3.2883301898092833e-06,
+      "loss": 0.7305,
+      "step": 9732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0370412615659743,
+      "learning_rate": 3.287356365685168e-06,
+      "loss": 0.6494,
+      "step": 9733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0952751877800895,
+      "learning_rate": 3.2863826151571654e-06,
+      "loss": 0.7252,
+      "step": 9734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7284622997252612,
+      "learning_rate": 3.285408938267116e-06,
+      "loss": 0.8379,
+      "step": 9735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.81932324189115,
+      "learning_rate": 3.284435335056865e-06,
+      "loss": 0.7422,
+      "step": 9736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5746877101716739,
+      "learning_rate": 3.283461805568246e-06,
+      "loss": 0.7887,
+      "step": 9737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8634699670790322,
+      "learning_rate": 3.2824883498430992e-06,
+      "loss": 0.7095,
+      "step": 9738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3154965989350837,
+      "learning_rate": 3.2815149679232507e-06,
+      "loss": 0.6323,
+      "step": 9739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9883456350379676,
+      "learning_rate": 3.2805416598505323e-06,
+      "loss": 0.6808,
+      "step": 9740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5011911917855167,
+      "learning_rate": 3.2795684256667674e-06,
+      "loss": 0.6817,
+      "step": 9741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5492860632690657,
+      "learning_rate": 3.278595265413781e-06,
+      "loss": 0.6896,
+      "step": 9742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.702773315621557,
+      "learning_rate": 3.277622179133387e-06,
+      "loss": 0.5776,
+      "step": 9743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9568651376397495,
+      "learning_rate": 3.276649166867406e-06,
+      "loss": 0.7148,
+      "step": 9744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8744085898774057,
+      "learning_rate": 3.2756762286576482e-06,
+      "loss": 0.8329,
+      "step": 9745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7712688474223381,
+      "learning_rate": 3.2747033645459226e-06,
+      "loss": 0.8174,
+      "step": 9746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7842560279459214,
+      "learning_rate": 3.2737305745740353e-06,
+      "loss": 0.7198,
+      "step": 9747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8791960778230212,
+      "learning_rate": 3.2727578587837893e-06,
+      "loss": 0.6773,
+      "step": 9748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8989285662400286,
+      "learning_rate": 3.271785217216987e-06,
+      "loss": 0.7202,
+      "step": 9749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5488991134586898,
+      "learning_rate": 3.2708126499154225e-06,
+      "loss": 0.6759,
+      "step": 9750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1782821027904573,
+      "learning_rate": 3.2698401569208883e-06,
+      "loss": 0.792,
+      "step": 9751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.997247018941902,
+      "learning_rate": 3.2688677382751754e-06,
+      "loss": 0.7608,
+      "step": 9752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5000742603505777,
+      "learning_rate": 3.2678953940200744e-06,
+      "loss": 0.7239,
+      "step": 9753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4371325960727757,
+      "learning_rate": 3.266923124197363e-06,
+      "loss": 0.6703,
+      "step": 9754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7201673253687675,
+      "learning_rate": 3.265950928848825e-06,
+      "loss": 0.8,
+      "step": 9755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5273102218400454,
+      "learning_rate": 3.264978808016239e-06,
+      "loss": 0.7738,
+      "step": 9756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4548215537169096,
+      "learning_rate": 3.264006761741376e-06,
+      "loss": 0.6832,
+      "step": 9757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1700223703300197,
+      "learning_rate": 3.2630347900660094e-06,
+      "loss": 0.6461,
+      "step": 9758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.598597772054794,
+      "learning_rate": 3.2620628930319065e-06,
+      "loss": 0.7973,
+      "step": 9759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8135352641890903,
+      "learning_rate": 3.261091070680833e-06,
+      "loss": 0.7064,
+      "step": 9760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3451321567078163,
+      "learning_rate": 3.260119323054546e-06,
+      "loss": 0.7642,
+      "step": 9761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.790936606106109,
+      "learning_rate": 3.2591476501948075e-06,
+      "loss": 0.7397,
+      "step": 9762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5152071396485187,
+      "learning_rate": 3.258176052143374e-06,
+      "loss": 0.6928,
+      "step": 9763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5844504398800103,
+      "learning_rate": 3.257204528941993e-06,
+      "loss": 0.6172,
+      "step": 9764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6818171602886578,
+      "learning_rate": 3.256233080632414e-06,
+      "loss": 0.813,
+      "step": 9765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5053861251286955,
+      "learning_rate": 3.255261707256383e-06,
+      "loss": 0.6866,
+      "step": 9766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7644063306583324,
+      "learning_rate": 3.2542904088556437e-06,
+      "loss": 0.717,
+      "step": 9767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5391731790330396,
+      "learning_rate": 3.2533191854719336e-06,
+      "loss": 0.5607,
+      "step": 9768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7425827533646314,
+      "learning_rate": 3.2523480371469863e-06,
+      "loss": 0.6863,
+      "step": 9769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7114999907393238,
+      "learning_rate": 3.251376963922537e-06,
+      "loss": 0.8746,
+      "step": 9770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4452473199841032,
+      "learning_rate": 3.250405965840315e-06,
+      "loss": 0.6648,
+      "step": 9771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.493458752140425,
+      "learning_rate": 3.249435042942043e-06,
+      "loss": 0.7298,
+      "step": 9772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.911627369409465,
+      "learning_rate": 3.2484641952694473e-06,
+      "loss": 0.7338,
+      "step": 9773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5164706495917597,
+      "learning_rate": 3.2474934228642475e-06,
+      "loss": 0.7122,
+      "step": 9774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8890919069450856,
+      "learning_rate": 3.246522725768156e-06,
+      "loss": 0.7086,
+      "step": 9775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.226086001913698,
+      "learning_rate": 3.2455521040228883e-06,
+      "loss": 0.8112,
+      "step": 9776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7691256045384005,
+      "learning_rate": 3.2445815576701544e-06,
+      "loss": 0.7876,
+      "step": 9777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6612868007296486,
+      "learning_rate": 3.2436110867516625e-06,
+      "loss": 0.8146,
+      "step": 9778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9487609741000005,
+      "learning_rate": 3.242640691309111e-06,
+      "loss": 0.7632,
+      "step": 9779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5894984302687132,
+      "learning_rate": 3.2416703713842036e-06,
+      "loss": 0.6845,
+      "step": 9780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6793883372886564,
+      "learning_rate": 3.240700127018638e-06,
+      "loss": 0.6796,
+      "step": 9781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1279769674531703,
+      "learning_rate": 3.239729958254104e-06,
+      "loss": 0.5896,
+      "step": 9782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8141926474429018,
+      "learning_rate": 3.238759865132295e-06,
+      "loss": 0.7351,
+      "step": 9783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7419786035151916,
+      "learning_rate": 3.2377898476948964e-06,
+      "loss": 0.6755,
+      "step": 9784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.454635775575221,
+      "learning_rate": 3.2368199059835947e-06,
+      "loss": 0.6519,
+      "step": 9785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.785682027373229,
+      "learning_rate": 3.235850040040066e-06,
+      "loss": 0.8369,
+      "step": 9786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8034191288100758,
+      "learning_rate": 3.2348802499059915e-06,
+      "loss": 0.8126,
+      "step": 9787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4747988723484673,
+      "learning_rate": 3.2339105356230426e-06,
+      "loss": 0.6984,
+      "step": 9788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6491654640080702,
+      "learning_rate": 3.2329408972328934e-06,
+      "loss": 0.6929,
+      "step": 9789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6589517240259013,
+      "learning_rate": 3.2319713347772075e-06,
+      "loss": 0.5882,
+      "step": 9790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6110670585419664,
+      "learning_rate": 3.231001848297651e-06,
+      "loss": 0.681,
+      "step": 9791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6463269399454676,
+      "learning_rate": 3.230032437835887e-06,
+      "loss": 0.8011,
+      "step": 9792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6801021496483708,
+      "learning_rate": 3.2290631034335684e-06,
+      "loss": 0.8611,
+      "step": 9793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6498048120413638,
+      "learning_rate": 3.2280938451323524e-06,
+      "loss": 0.6155,
+      "step": 9794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9970944934753159,
+      "learning_rate": 3.2271246629738917e-06,
+      "loss": 0.6325,
+      "step": 9795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.881180253048101,
+      "learning_rate": 3.2261555569998316e-06,
+      "loss": 0.6712,
+      "step": 9796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6840095868054186,
+      "learning_rate": 3.2251865272518187e-06,
+      "loss": 0.6896,
+      "step": 9797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6072077288700612,
+      "learning_rate": 3.224217573771492e-06,
+      "loss": 0.6912,
+      "step": 9798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.634674798369735,
+      "learning_rate": 3.223248696600493e-06,
+      "loss": 0.6889,
+      "step": 9799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7763359948196231,
+      "learning_rate": 3.2222798957804524e-06,
+      "loss": 0.7664,
+      "step": 9800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7623583438969028,
+      "learning_rate": 3.2213111713530056e-06,
+      "loss": 0.6867,
+      "step": 9801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.512608593918939,
+      "learning_rate": 3.2203425233597773e-06,
+      "loss": 0.5791,
+      "step": 9802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.733229497140722,
+      "learning_rate": 3.2193739518423972e-06,
+      "loss": 0.7867,
+      "step": 9803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8663156203125988,
+      "learning_rate": 3.2184054568424817e-06,
+      "loss": 0.7535,
+      "step": 9804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0678924091765158,
+      "learning_rate": 3.217437038401652e-06,
+      "loss": 0.596,
+      "step": 9805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5848904714667926,
+      "learning_rate": 3.2164686965615223e-06,
+      "loss": 0.7081,
+      "step": 9806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6681374740859247,
+      "learning_rate": 3.215500431363706e-06,
+      "loss": 0.6646,
+      "step": 9807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9571998718641643,
+      "learning_rate": 3.2145322428498093e-06,
+      "loss": 0.7222,
+      "step": 9808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4713431208651104,
+      "learning_rate": 3.2135641310614383e-06,
+      "loss": 0.749,
+      "step": 9809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0745385584225848,
+      "learning_rate": 3.2125960960401974e-06,
+      "loss": 0.5924,
+      "step": 9810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9437961625736018,
+      "learning_rate": 3.211628137827682e-06,
+      "loss": 0.7311,
+      "step": 9811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.614683636513871,
+      "learning_rate": 3.2106602564654877e-06,
+      "loss": 0.609,
+      "step": 9812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6918029319851882,
+      "learning_rate": 3.2096924519952067e-06,
+      "loss": 0.7463,
+      "step": 9813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5924307596648895,
+      "learning_rate": 3.20872472445843e-06,
+      "loss": 0.6321,
+      "step": 9814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6183370014638625,
+      "learning_rate": 3.207757073896741e-06,
+      "loss": 0.7346,
+      "step": 9815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6171055565998746,
+      "learning_rate": 3.2067895003517207e-06,
+      "loss": 0.8462,
+      "step": 9816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.970802989831183,
+      "learning_rate": 3.2058220038649513e-06,
+      "loss": 0.7345,
+      "step": 9817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.663332311899242,
+      "learning_rate": 3.2048545844780044e-06,
+      "loss": 0.7755,
+      "step": 9818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.731397734668209,
+      "learning_rate": 3.203887242232455e-06,
+      "loss": 0.7551,
+      "step": 9819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4577952761549955,
+      "learning_rate": 3.2029199771698693e-06,
+      "loss": 0.7189,
+      "step": 9820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4890085306749719,
+      "learning_rate": 3.2019527893318177e-06,
+      "loss": 0.5904,
+      "step": 9821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0783669079252658,
+      "learning_rate": 3.2009856787598558e-06,
+      "loss": 0.6322,
+      "step": 9822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9726323485466937,
+      "learning_rate": 3.2000186454955474e-06,
+      "loss": 0.821,
+      "step": 9823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5371045567290142,
+      "learning_rate": 3.1990516895804467e-06,
+      "loss": 0.634,
+      "step": 9824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4474762204128475,
+      "learning_rate": 3.198084811056107e-06,
+      "loss": 0.7411,
+      "step": 9825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9065594027689186,
+      "learning_rate": 3.197118009964074e-06,
+      "loss": 0.7062,
+      "step": 9826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1692802683582928,
+      "learning_rate": 3.1961512863458953e-06,
+      "loss": 0.7235,
+      "step": 9827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6694986982780775,
+      "learning_rate": 3.195184640243115e-06,
+      "loss": 0.7808,
+      "step": 9828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1047627699167575,
+      "learning_rate": 3.1942180716972698e-06,
+      "loss": 0.6087,
+      "step": 9829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8940874367401572,
+      "learning_rate": 3.1932515807498945e-06,
+      "loss": 0.7311,
+      "step": 9830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7309885001692598,
+      "learning_rate": 3.1922851674425226e-06,
+      "loss": 0.7386,
+      "step": 9831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.631218916623737,
+      "learning_rate": 3.191318831816685e-06,
+      "loss": 0.8214,
+      "step": 9832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3682364092577968,
+      "learning_rate": 3.190352573913903e-06,
+      "loss": 0.596,
+      "step": 9833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.095377384886487,
+      "learning_rate": 3.189386393775703e-06,
+      "loss": 0.5882,
+      "step": 9834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7709960770983941,
+      "learning_rate": 3.1884202914436024e-06,
+      "loss": 0.8464,
+      "step": 9835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6180763819724613,
+      "learning_rate": 3.1874542669591148e-06,
+      "loss": 0.799,
+      "step": 9836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6206711432639125,
+      "learning_rate": 3.1864883203637544e-06,
+      "loss": 0.7052,
+      "step": 9837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6476120298059809,
+      "learning_rate": 3.18552245169903e-06,
+      "loss": 0.7029,
+      "step": 9838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.645204263191284,
+      "learning_rate": 3.1845566610064487e-06,
+      "loss": 0.7362,
+      "step": 9839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.350528976634715,
+      "learning_rate": 3.1835909483275083e-06,
+      "loss": 0.7043,
+      "step": 9840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9061379884589933,
+      "learning_rate": 3.1826253137037112e-06,
+      "loss": 0.7155,
+      "step": 9841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8054032973531229,
+      "learning_rate": 3.1816597571765517e-06,
+      "loss": 0.7887,
+      "step": 9842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8068065848939467,
+      "learning_rate": 3.180694278787524e-06,
+      "loss": 0.6774,
+      "step": 9843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6168457091117021,
+      "learning_rate": 3.179728878578112e-06,
+      "loss": 0.7672,
+      "step": 9844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5583616414996682,
+      "learning_rate": 3.1787635565898046e-06,
+      "loss": 0.6364,
+      "step": 9845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1353730317403978,
+      "learning_rate": 3.1777983128640855e-06,
+      "loss": 0.594,
+      "step": 9846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7916018439591037,
+      "learning_rate": 3.1768331474424285e-06,
+      "loss": 0.8372,
+      "step": 9847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.737624416118953,
+      "learning_rate": 3.1758680603663127e-06,
+      "loss": 0.8114,
+      "step": 9848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5698273909895513,
+      "learning_rate": 3.1749030516772084e-06,
+      "loss": 0.6686,
+      "step": 9849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2148249079243305,
+      "learning_rate": 3.1739381214165864e-06,
+      "loss": 0.6508,
+      "step": 9850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5610346459722493,
+      "learning_rate": 3.1729732696259085e-06,
+      "loss": 0.7085,
+      "step": 9851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5801187276482662,
+      "learning_rate": 3.172008496346639e-06,
+      "loss": 0.7467,
+      "step": 9852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6603081739766912,
+      "learning_rate": 3.1710438016202362e-06,
+      "loss": 0.7521,
+      "step": 9853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.050340150343746,
+      "learning_rate": 3.170079185488153e-06,
+      "loss": 0.6067,
+      "step": 9854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6963927380988149,
+      "learning_rate": 3.1691146479918423e-06,
+      "loss": 0.7511,
+      "step": 9855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.476550951172742,
+      "learning_rate": 3.168150189172754e-06,
+      "loss": 0.7196,
+      "step": 9856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7630990299361355,
+      "learning_rate": 3.1671858090723315e-06,
+      "loss": 0.8906,
+      "step": 9857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6836105504701773,
+      "learning_rate": 3.1662215077320174e-06,
+      "loss": 0.7228,
+      "step": 9858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.84988076520665,
+      "learning_rate": 3.165257285193248e-06,
+      "loss": 0.7258,
+      "step": 9859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1954726214853308,
+      "learning_rate": 3.1642931414974586e-06,
+      "loss": 0.6057,
+      "step": 9860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5877666743486656,
+      "learning_rate": 3.1633290766860834e-06,
+      "loss": 0.6743,
+      "step": 9861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8376573670053848,
+      "learning_rate": 3.1623650908005476e-06,
+      "loss": 0.915,
+      "step": 9862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6160242044275037,
+      "learning_rate": 3.1614011838822755e-06,
+      "loss": 0.6859,
+      "step": 9863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.930205591731279,
+      "learning_rate": 3.1604373559726915e-06,
+      "loss": 0.7326,
+      "step": 9864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9075246921661193,
+      "learning_rate": 3.1594736071132086e-06,
+      "loss": 0.6869,
+      "step": 9865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4767308140296354,
+      "learning_rate": 3.1585099373452454e-06,
+      "loss": 0.645,
+      "step": 9866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5357440027222413,
+      "learning_rate": 3.15754634671021e-06,
+      "loss": 0.6436,
+      "step": 9867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9213912408640759,
+      "learning_rate": 3.1565828352495144e-06,
+      "loss": 0.8362,
+      "step": 9868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7862956979426632,
+      "learning_rate": 3.1556194030045563e-06,
+      "loss": 0.7627,
+      "step": 9869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.118024626569028,
+      "learning_rate": 3.154656050016742e-06,
+      "loss": 0.5591,
+      "step": 9870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1325367673495097,
+      "learning_rate": 3.153692776327466e-06,
+      "loss": 0.7502,
+      "step": 9871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5816338995633876,
+      "learning_rate": 3.1527295819781256e-06,
+      "loss": 0.6867,
+      "step": 9872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.08428502482623,
+      "learning_rate": 3.151766467010107e-06,
+      "loss": 0.7866,
+      "step": 9873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5165169391022866,
+      "learning_rate": 3.1508034314647994e-06,
+      "loss": 0.7243,
+      "step": 9874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6494190307543006,
+      "learning_rate": 3.149840475383589e-06,
+      "loss": 0.8201,
+      "step": 9875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2278665589709412,
+      "learning_rate": 3.1488775988078525e-06,
+      "loss": 0.6904,
+      "step": 9876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1435665075674104,
+      "learning_rate": 3.1479148017789673e-06,
+      "loss": 0.7013,
+      "step": 9877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.164616634423257,
+      "learning_rate": 3.146952084338308e-06,
+      "loss": 0.704,
+      "step": 9878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6925734992830475,
+      "learning_rate": 3.1459894465272467e-06,
+      "loss": 0.7197,
+      "step": 9879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5143055538552976,
+      "learning_rate": 3.1450268883871467e-06,
+      "loss": 0.6104,
+      "step": 9880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5692012896409535,
+      "learning_rate": 3.1440644099593716e-06,
+      "loss": 0.6484,
+      "step": 9881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5124716238044595,
+      "learning_rate": 3.143102011285284e-06,
+      "loss": 0.7177,
+      "step": 9882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5944613197425868,
+      "learning_rate": 3.142139692406237e-06,
+      "loss": 0.6622,
+      "step": 9883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.542728041732593,
+      "learning_rate": 3.1411774533635854e-06,
+      "loss": 0.8312,
+      "step": 9884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7913420981154782,
+      "learning_rate": 3.1402152941986786e-06,
+      "loss": 0.755,
+      "step": 9885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6664022499092614,
+      "learning_rate": 3.1392532149528643e-06,
+      "loss": 0.662,
+      "step": 9886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.103881049389222,
+      "learning_rate": 3.1382912156674807e-06,
+      "loss": 0.6129,
+      "step": 9887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6372767682285967,
+      "learning_rate": 3.13732929638387e-06,
+      "loss": 0.6611,
+      "step": 9888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2185285823785836,
+      "learning_rate": 3.136367457143369e-06,
+      "loss": 0.7315,
+      "step": 9889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6773775804089148,
+      "learning_rate": 3.1354056979873104e-06,
+      "loss": 0.7507,
+      "step": 9890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7066096152071606,
+      "learning_rate": 3.134444018957019e-06,
+      "loss": 0.7498,
+      "step": 9891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7175051313543295,
+      "learning_rate": 3.1334824200938228e-06,
+      "loss": 0.7599,
+      "step": 9892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8593909013862417,
+      "learning_rate": 3.132520901439047e-06,
+      "loss": 0.7538,
+      "step": 9893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5545042741249464,
+      "learning_rate": 3.1315594630340052e-06,
+      "loss": 0.6666,
+      "step": 9894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7229822206019185,
+      "learning_rate": 3.1305981049200134e-06,
+      "loss": 0.8259,
+      "step": 9895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7967461966847649,
+      "learning_rate": 3.129636827138385e-06,
+      "loss": 0.8203,
+      "step": 9896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.582946591737606,
+      "learning_rate": 3.128675629730429e-06,
+      "loss": 0.7191,
+      "step": 9897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5759094845958364,
+      "learning_rate": 3.1277145127374475e-06,
+      "loss": 0.7084,
+      "step": 9898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7499088006575483,
+      "learning_rate": 3.1267534762007435e-06,
+      "loss": 0.6105,
+      "step": 9899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6481297200023048,
+      "learning_rate": 3.125792520161616e-06,
+      "loss": 0.7533,
+      "step": 9900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0821744566544522,
+      "learning_rate": 3.1248316446613563e-06,
+      "loss": 0.7838,
+      "step": 9901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8060033611090236,
+      "learning_rate": 3.1238708497412566e-06,
+      "loss": 0.6768,
+      "step": 9902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.660667993575307,
+      "learning_rate": 3.1229101354426063e-06,
+      "loss": 0.7534,
+      "step": 9903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5345330220536746,
+      "learning_rate": 3.1219495018066888e-06,
+      "loss": 0.929,
+      "step": 9904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4432497013208967,
+      "learning_rate": 3.1209889488747813e-06,
+      "loss": 0.6693,
+      "step": 9905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.509041046043661,
+      "learning_rate": 3.120028476688164e-06,
+      "loss": 0.7513,
+      "step": 9906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.623128717781504,
+      "learning_rate": 3.119068085288111e-06,
+      "loss": 0.7694,
+      "step": 9907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7266198783044209,
+      "learning_rate": 3.118107774715892e-06,
+      "loss": 0.7164,
+      "step": 9908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8447449442778445,
+      "learning_rate": 3.1171475450127717e-06,
+      "loss": 0.7328,
+      "step": 9909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.010003482541257,
+      "learning_rate": 3.1161873962200146e-06,
+      "loss": 0.7348,
+      "step": 9910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.027960027851861,
+      "learning_rate": 3.1152273283788824e-06,
+      "loss": 0.7107,
+      "step": 9911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6071259630622894,
+      "learning_rate": 3.114267341530627e-06,
+      "loss": 0.6853,
+      "step": 9912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5368941681301567,
+      "learning_rate": 3.1133074357165055e-06,
+      "loss": 0.6407,
+      "step": 9913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.522679345404686,
+      "learning_rate": 3.112347610977764e-06,
+      "loss": 0.7575,
+      "step": 9914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7334877145858054,
+      "learning_rate": 3.111387867355652e-06,
+      "loss": 0.6995,
+      "step": 9915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.665352746285693,
+      "learning_rate": 3.110428204891407e-06,
+      "loss": 0.6915,
+      "step": 9916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6961297729953637,
+      "learning_rate": 3.109468623626272e-06,
+      "loss": 0.6142,
+      "step": 9917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6561659976334668,
+      "learning_rate": 3.1085091236014808e-06,
+      "loss": 0.6994,
+      "step": 9918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.599764123622961,
+      "learning_rate": 3.1075497048582635e-06,
+      "loss": 0.6347,
+      "step": 9919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5042865892436255,
+      "learning_rate": 3.1065903674378507e-06,
+      "loss": 0.7617,
+      "step": 9920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.459831209847208,
+      "learning_rate": 3.1056311113814675e-06,
+      "loss": 0.7044,
+      "step": 9921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4762403192335145,
+      "learning_rate": 3.1046719367303334e-06,
+      "loss": 0.7338,
+      "step": 9922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0179761508227387,
+      "learning_rate": 3.1037128435256688e-06,
+      "loss": 0.6994,
+      "step": 9923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.869626311971399,
+      "learning_rate": 3.102753831808685e-06,
+      "loss": 0.8115,
+      "step": 9924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5242059711589875,
+      "learning_rate": 3.101794901620594e-06,
+      "loss": 0.7411,
+      "step": 9925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0377164474816656,
+      "learning_rate": 3.1008360530026053e-06,
+      "loss": 0.816,
+      "step": 9926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7432239563362706,
+      "learning_rate": 3.099877285995921e-06,
+      "loss": 0.8377,
+      "step": 9927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.623376533796444,
+      "learning_rate": 3.09891860064174e-06,
+      "loss": 0.7424,
+      "step": 9928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5595043585359896,
+      "learning_rate": 3.097959996981263e-06,
+      "loss": 0.7381,
+      "step": 9929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6192678230845776,
+      "learning_rate": 3.0970014750556786e-06,
+      "loss": 0.7691,
+      "step": 9930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6626183744114806,
+      "learning_rate": 3.0960430349061797e-06,
+      "loss": 0.729,
+      "step": 9931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1855921673846472,
+      "learning_rate": 3.0950846765739518e-06,
+      "loss": 0.8701,
+      "step": 9932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0886879500864188,
+      "learning_rate": 3.0941264001001796e-06,
+      "loss": 0.6231,
+      "step": 9933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8025509517224956,
+      "learning_rate": 3.093168205526038e-06,
+      "loss": 0.8421,
+      "step": 9934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1623455709668056,
+      "learning_rate": 3.0922100928927062e-06,
+      "loss": 0.571,
+      "step": 9935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.108513352598046,
+      "learning_rate": 3.0912520622413573e-06,
+      "loss": 0.6646,
+      "step": 9936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2115031749342586,
+      "learning_rate": 3.090294113613157e-06,
+      "loss": 0.6807,
+      "step": 9937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.800466952681871,
+      "learning_rate": 3.0893362470492716e-06,
+      "loss": 0.7053,
+      "step": 9938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.465091465554897,
+      "learning_rate": 3.0883784625908618e-06,
+      "loss": 0.6027,
+      "step": 9939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.526929309913808,
+      "learning_rate": 3.0874207602790895e-06,
+      "loss": 0.6941,
+      "step": 9940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.153948519087093,
+      "learning_rate": 3.086463140155106e-06,
+      "loss": 0.7608,
+      "step": 9941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5911487427508206,
+      "learning_rate": 3.085505602260061e-06,
+      "loss": 0.5745,
+      "step": 9942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0328759380870483,
+      "learning_rate": 3.0845481466351047e-06,
+      "loss": 0.6515,
+      "step": 9943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8062558077005249,
+      "learning_rate": 3.083590773321383e-06,
+      "loss": 0.7081,
+      "step": 9944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6806608879121359,
+      "learning_rate": 3.082633482360032e-06,
+      "loss": 0.66,
+      "step": 9945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6000895436572398,
+      "learning_rate": 3.0816762737921896e-06,
+      "loss": 0.63,
+      "step": 9946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.683050374547006,
+      "learning_rate": 3.0807191476589926e-06,
+      "loss": 0.7155,
+      "step": 9947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8206094465874254,
+      "learning_rate": 3.0797621040015656e-06,
+      "loss": 0.7157,
+      "step": 9948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6860152416602747,
+      "learning_rate": 3.0788051428610377e-06,
+      "loss": 0.8238,
+      "step": 9949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4170766537545494,
+      "learning_rate": 3.077848264278533e-06,
+      "loss": 0.7662,
+      "step": 9950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5562124490461342,
+      "learning_rate": 3.07689146829517e-06,
+      "loss": 0.758,
+      "step": 9951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8390087776327193,
+      "learning_rate": 3.0759347549520614e-06,
+      "loss": 0.8293,
+      "step": 9952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8651839777251538,
+      "learning_rate": 3.074978124290322e-06,
+      "loss": 0.7638,
+      "step": 9953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8355303175897983,
+      "learning_rate": 3.0740215763510617e-06,
+      "loss": 0.7304,
+      "step": 9954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6224645949957455,
+      "learning_rate": 3.0730651111753824e-06,
+      "loss": 0.7397,
+      "step": 9955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5517739862085642,
+      "learning_rate": 3.072108728804386e-06,
+      "loss": 0.6568,
+      "step": 9956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.757382246996966,
+      "learning_rate": 3.0711524292791707e-06,
+      "loss": 0.7219,
+      "step": 9957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7344479540724616,
+      "learning_rate": 3.0701962126408334e-06,
+      "loss": 0.7943,
+      "step": 9958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6837546819939502,
+      "learning_rate": 3.069240078930461e-06,
+      "loss": 0.7019,
+      "step": 9959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6003921941365655,
+      "learning_rate": 3.0682840281891433e-06,
+      "loss": 0.6958,
+      "step": 9960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6722365793648652,
+      "learning_rate": 3.0673280604579623e-06,
+      "loss": 0.7298,
+      "step": 9961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.371237045832826,
+      "learning_rate": 3.0663721757780006e-06,
+      "loss": 0.6457,
+      "step": 9962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.506254099583517,
+      "learning_rate": 3.0654163741903315e-06,
+      "loss": 0.6222,
+      "step": 9963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0628005048979747,
+      "learning_rate": 3.0644606557360303e-06,
+      "loss": 0.6675,
+      "step": 9964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5890126898427075,
+      "learning_rate": 3.0635050204561658e-06,
+      "loss": 0.7789,
+      "step": 9965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7521918757796648,
+      "learning_rate": 3.0625494683918012e-06,
+      "loss": 0.7031,
+      "step": 9966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.716226479616048,
+      "learning_rate": 3.0615939995840017e-06,
+      "loss": 0.7465,
+      "step": 9967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4667828250193475,
+      "learning_rate": 3.0606386140738253e-06,
+      "loss": 0.5949,
+      "step": 9968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1679328237960713,
+      "learning_rate": 3.0596833119023283e-06,
+      "loss": 0.6064,
+      "step": 9969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8379664417324417,
+      "learning_rate": 3.058728093110559e-06,
+      "loss": 0.6767,
+      "step": 9970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6078572685251489,
+      "learning_rate": 3.0577729577395663e-06,
+      "loss": 0.8324,
+      "step": 9971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9270874057752416,
+      "learning_rate": 3.056817905830397e-06,
+      "loss": 0.7602,
+      "step": 9972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8683163590198588,
+      "learning_rate": 3.055862937424089e-06,
+      "loss": 0.7442,
+      "step": 9973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6599323476194716,
+      "learning_rate": 3.054908052561681e-06,
+      "loss": 0.7235,
+      "step": 9974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8253132389488747,
+      "learning_rate": 3.053953251284205e-06,
+      "loss": 0.6237,
+      "step": 9975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0499353076645273,
+      "learning_rate": 3.0529985336326933e-06,
+      "loss": 0.6673,
+      "step": 9976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.638971790190698,
+      "learning_rate": 3.0520438996481694e-06,
+      "loss": 0.6961,
+      "step": 9977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6570659216680803,
+      "learning_rate": 3.0510893493716574e-06,
+      "loss": 0.737,
+      "step": 9978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5465283033253425,
+      "learning_rate": 3.0501348828441767e-06,
+      "loss": 0.7887,
+      "step": 9979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4868359420321802,
+      "learning_rate": 3.0491805001067443e-06,
+      "loss": 0.7049,
+      "step": 9980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7999499676744293,
+      "learning_rate": 3.048226201200369e-06,
+      "loss": 0.6693,
+      "step": 9981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5599529929749283,
+      "learning_rate": 3.047271986166061e-06,
+      "loss": 0.6269,
+      "step": 9982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6784711175311164,
+      "learning_rate": 3.0463178550448246e-06,
+      "loss": 0.7435,
+      "step": 9983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8603452221947794,
+      "learning_rate": 3.0453638078776614e-06,
+      "loss": 0.6856,
+      "step": 9984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8362379131154536,
+      "learning_rate": 3.044409844705568e-06,
+      "loss": 0.8775,
+      "step": 9985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9790661301667467,
+      "learning_rate": 3.043455965569538e-06,
+      "loss": 0.7463,
+      "step": 9986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5977720778508404,
+      "learning_rate": 3.042502170510565e-06,
+      "loss": 0.7113,
+      "step": 9987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.698614213317283,
+      "learning_rate": 3.041548459569633e-06,
+      "loss": 0.663,
+      "step": 9988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7120791618253508,
+      "learning_rate": 3.0405948327877233e-06,
+      "loss": 0.7389,
+      "step": 9989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9253152926188262,
+      "learning_rate": 3.039641290205818e-06,
+      "loss": 0.6926,
+      "step": 9990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5659308844783761,
+      "learning_rate": 3.038687831864895e-06,
+      "loss": 0.6962,
+      "step": 9991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5721400164610764,
+      "learning_rate": 3.037734457805922e-06,
+      "loss": 0.7322,
+      "step": 9992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7413894123885418,
+      "learning_rate": 3.036781168069869e-06,
+      "loss": 0.6917,
+      "step": 9993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0777569439813042,
+      "learning_rate": 3.0358279626977034e-06,
+      "loss": 0.6329,
+      "step": 9994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8244274938731133,
+      "learning_rate": 3.0348748417303826e-06,
+      "loss": 0.8021,
+      "step": 9995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7381421573767433,
+      "learning_rate": 3.033921805208867e-06,
+      "loss": 0.7645,
+      "step": 9996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.135736311101823,
+      "learning_rate": 3.0329688531741097e-06,
+      "loss": 0.6133,
+      "step": 9997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.820383396468158,
+      "learning_rate": 3.032015985667064e-06,
+      "loss": 0.7202,
+      "step": 9998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1091288018563512,
+      "learning_rate": 3.0310632027286717e-06,
+      "loss": 0.7087,
+      "step": 9999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8504373486518255,
+      "learning_rate": 3.030110504399879e-06,
+      "loss": 0.7901,
+      "step": 10000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.527410013362632,
+      "learning_rate": 3.029157890721627e-06,
+      "loss": 0.746,
+      "step": 10001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6963276618519816,
+      "learning_rate": 3.0282053617348485e-06,
+      "loss": 0.6688,
+      "step": 10002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8130673541970044,
+      "learning_rate": 3.027252917480476e-06,
+      "loss": 0.7137,
+      "step": 10003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7596165275323066,
+      "learning_rate": 3.026300557999439e-06,
+      "loss": 0.6926,
+      "step": 10004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2434890735561024,
+      "learning_rate": 3.0253482833326654e-06,
+      "loss": 0.8473,
+      "step": 10005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.544769456401173,
+      "learning_rate": 3.0243960935210726e-06,
+      "loss": 0.7917,
+      "step": 10006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4750843269772245,
+      "learning_rate": 3.023443988605578e-06,
+      "loss": 0.5977,
+      "step": 10007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5825788821941154,
+      "learning_rate": 3.0224919686270982e-06,
+      "loss": 0.7892,
+      "step": 10008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.076631728520331,
+      "learning_rate": 3.021540033626544e-06,
+      "loss": 0.6345,
+      "step": 10009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5213100106677038,
+      "learning_rate": 3.0205881836448186e-06,
+      "loss": 0.7064,
+      "step": 10010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7344834825333462,
+      "learning_rate": 3.0196364187228287e-06,
+      "loss": 0.6879,
+      "step": 10011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7343080186036623,
+      "learning_rate": 3.0186847389014738e-06,
+      "loss": 0.7464,
+      "step": 10012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.633961400238472,
+      "learning_rate": 3.0177331442216455e-06,
+      "loss": 0.7285,
+      "step": 10013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3476235268006813,
+      "learning_rate": 3.0167816347242396e-06,
+      "loss": 0.7525,
+      "step": 10014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6422866658355917,
+      "learning_rate": 3.0158302104501435e-06,
+      "loss": 0.7367,
+      "step": 10015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3055022642649787,
+      "learning_rate": 3.0148788714402444e-06,
+      "loss": 0.6728,
+      "step": 10016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8836411217616558,
+      "learning_rate": 3.0139276177354188e-06,
+      "loss": 0.7258,
+      "step": 10017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9817442541936836,
+      "learning_rate": 3.0129764493765466e-06,
+      "loss": 0.7652,
+      "step": 10018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4569130516733013,
+      "learning_rate": 3.012025366404504e-06,
+      "loss": 0.7062,
+      "step": 10019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0404723508570095,
+      "learning_rate": 3.0110743688601578e-06,
+      "loss": 0.6223,
+      "step": 10020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2142109964598298,
+      "learning_rate": 3.0101234567843747e-06,
+      "loss": 0.5989,
+      "step": 10021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.714109368994043,
+      "learning_rate": 3.009172630218018e-06,
+      "loss": 0.657,
+      "step": 10022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.661670528100128,
+      "learning_rate": 3.0082218892019487e-06,
+      "loss": 0.7329,
+      "step": 10023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9004955165931554,
+      "learning_rate": 3.00727123377702e-06,
+      "loss": 0.7893,
+      "step": 10024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8177756479823937,
+      "learning_rate": 3.0063206639840846e-06,
+      "loss": 0.6956,
+      "step": 10025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.576273321342593,
+      "learning_rate": 3.00537017986399e-06,
+      "loss": 0.6478,
+      "step": 10026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7064499414498737,
+      "learning_rate": 3.0044197814575825e-06,
+      "loss": 0.6761,
+      "step": 10027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7381560262643814,
+      "learning_rate": 3.0034694688056998e-06,
+      "loss": 0.6844,
+      "step": 10028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4941791808644929,
+      "learning_rate": 3.002519241949181e-06,
+      "loss": 0.6963,
+      "step": 10029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5505551275067693,
+      "learning_rate": 3.0015691009288605e-06,
+      "loss": 0.5984,
+      "step": 10030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6534141001124791,
+      "learning_rate": 3.0006190457855643e-06,
+      "loss": 0.6755,
+      "step": 10031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1943694541821883,
+      "learning_rate": 2.99966907656012e-06,
+      "loss": 0.6879,
+      "step": 10032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2957545954322494,
+      "learning_rate": 2.9987191932933523e-06,
+      "loss": 0.7136,
+      "step": 10033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1091932754476352,
+      "learning_rate": 2.997769396026078e-06,
+      "loss": 0.6789,
+      "step": 10034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.676802503484728,
+      "learning_rate": 2.9968196847991117e-06,
+      "loss": 0.7477,
+      "step": 10035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9073170919693996,
+      "learning_rate": 2.995870059653263e-06,
+      "loss": 0.659,
+      "step": 10036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2312315250248873,
+      "learning_rate": 2.994920520629344e-06,
+      "loss": 0.7468,
+      "step": 10037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.74964437136644,
+      "learning_rate": 2.9939710677681545e-06,
+      "loss": 0.6887,
+      "step": 10038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2164442425618052,
+      "learning_rate": 2.9930217011104957e-06,
+      "loss": 0.708,
+      "step": 10039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5949897075146766,
+      "learning_rate": 2.992072420697164e-06,
+      "loss": 0.7347,
+      "step": 10040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5060359580118083,
+      "learning_rate": 2.991123226568954e-06,
+      "loss": 0.7422,
+      "step": 10041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.915718632043237,
+      "learning_rate": 2.990174118766651e-06,
+      "loss": 0.8044,
+      "step": 10042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.11014611611491,
+      "learning_rate": 2.9892250973310433e-06,
+      "loss": 0.771,
+      "step": 10043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8110293342908839,
+      "learning_rate": 2.98827616230291e-06,
+      "loss": 0.7967,
+      "step": 10044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5289746653539231,
+      "learning_rate": 2.987327313723033e-06,
+      "loss": 0.7392,
+      "step": 10045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2844317595387953,
+      "learning_rate": 2.9863785516321806e-06,
+      "loss": 0.6728,
+      "step": 10046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6730461215374364,
+      "learning_rate": 2.9854298760711285e-06,
+      "loss": 0.6314,
+      "step": 10047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6374794830840955,
+      "learning_rate": 2.9844812870806417e-06,
+      "loss": 0.7893,
+      "step": 10048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5415448229214448,
+      "learning_rate": 2.9835327847014816e-06,
+      "loss": 0.6274,
+      "step": 10049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4240247482342716,
+      "learning_rate": 2.982584368974408e-06,
+      "loss": 0.6036,
+      "step": 10050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9939837655322912,
+      "learning_rate": 2.9816360399401777e-06,
+      "loss": 0.7433,
+      "step": 10051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6146241476798393,
+      "learning_rate": 2.980687797639543e-06,
+      "loss": 0.6431,
+      "step": 10052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7519762062445103,
+      "learning_rate": 2.979739642113251e-06,
+      "loss": 0.7259,
+      "step": 10053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7439624232067292,
+      "learning_rate": 2.9787915734020446e-06,
+      "loss": 0.6768,
+      "step": 10054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6382058764292462,
+      "learning_rate": 2.9778435915466674e-06,
+      "loss": 0.8378,
+      "step": 10055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.772873627866152,
+      "learning_rate": 2.9768956965878526e-06,
+      "loss": 0.6561,
+      "step": 10056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7478318466129528,
+      "learning_rate": 2.9759478885663362e-06,
+      "loss": 0.6556,
+      "step": 10057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.647281879143175,
+      "learning_rate": 2.975000167522846e-06,
+      "loss": 0.7758,
+      "step": 10058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9596063779768897,
+      "learning_rate": 2.9740525334981105e-06,
+      "loss": 0.6283,
+      "step": 10059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7979395059614134,
+      "learning_rate": 2.9731049865328486e-06,
+      "loss": 0.6819,
+      "step": 10060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7149155302274366,
+      "learning_rate": 2.9721575266677783e-06,
+      "loss": 0.6316,
+      "step": 10061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4563254284075924,
+      "learning_rate": 2.971210153943617e-06,
+      "loss": 0.7056,
+      "step": 10062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.726155704095348,
+      "learning_rate": 2.970262868401075e-06,
+      "loss": 0.591,
+      "step": 10063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7201047966659218,
+      "learning_rate": 2.9693156700808556e-06,
+      "loss": 0.6223,
+      "step": 10064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3047303813094933,
+      "learning_rate": 2.968368559023664e-06,
+      "loss": 0.6702,
+      "step": 10065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1132735408246055,
+      "learning_rate": 2.967421535270203e-06,
+      "loss": 0.6469,
+      "step": 10066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8923589140351174,
+      "learning_rate": 2.966474598861164e-06,
+      "loss": 0.6513,
+      "step": 10067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4340495456921627,
+      "learning_rate": 2.965527749837239e-06,
+      "loss": 0.6269,
+      "step": 10068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.225035427156675,
+      "learning_rate": 2.9645809882391187e-06,
+      "loss": 0.7477,
+      "step": 10069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.422450812452344,
+      "learning_rate": 2.9636343141074873e-06,
+      "loss": 0.7435,
+      "step": 10070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4867089501712942,
+      "learning_rate": 2.962687727483024e-06,
+      "loss": 0.6922,
+      "step": 10071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6553649789725802,
+      "learning_rate": 2.961741228406405e-06,
+      "loss": 0.7962,
+      "step": 10072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2997545688602434,
+      "learning_rate": 2.9607948169183077e-06,
+      "loss": 0.8049,
+      "step": 10073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.327050934870249,
+      "learning_rate": 2.959848493059396e-06,
+      "loss": 0.6909,
+      "step": 10074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.551101445427766,
+      "learning_rate": 2.958902256870339e-06,
+      "loss": 0.6151,
+      "step": 10075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.430154758117175,
+      "learning_rate": 2.9579561083917984e-06,
+      "loss": 0.717,
+      "step": 10076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6328830808268824,
+      "learning_rate": 2.9570100476644325e-06,
+      "loss": 0.6276,
+      "step": 10077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8015950080793304,
+      "learning_rate": 2.9560640747288925e-06,
+      "loss": 0.6711,
+      "step": 10078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6996299391763514,
+      "learning_rate": 2.9551181896258317e-06,
+      "loss": 0.7346,
+      "step": 10079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8249397921271617,
+      "learning_rate": 2.9541723923958975e-06,
+      "loss": 0.7638,
+      "step": 10080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1263869946526321,
+      "learning_rate": 2.9532266830797336e-06,
+      "loss": 0.7932,
+      "step": 10081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8137702561206555,
+      "learning_rate": 2.9522810617179742e-06,
+      "loss": 0.6802,
+      "step": 10082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7108447952319041,
+      "learning_rate": 2.951335528351259e-06,
+      "loss": 0.7102,
+      "step": 10083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9897310174686702,
+      "learning_rate": 2.9503900830202202e-06,
+      "loss": 0.687,
+      "step": 10084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7748731895485956,
+      "learning_rate": 2.9494447257654836e-06,
+      "loss": 0.7559,
+      "step": 10085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4629397466826655,
+      "learning_rate": 2.948499456627674e-06,
+      "loss": 0.6636,
+      "step": 10086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8184082934365673,
+      "learning_rate": 2.94755427564741e-06,
+      "loss": 0.7423,
+      "step": 10087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8804261115921768,
+      "learning_rate": 2.9466091828653127e-06,
+      "loss": 0.8095,
+      "step": 10088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1008459445352141,
+      "learning_rate": 2.9456641783219897e-06,
+      "loss": 0.6106,
+      "step": 10089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8794082880886556,
+      "learning_rate": 2.9447192620580532e-06,
+      "loss": 0.7405,
+      "step": 10090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7785938190576243,
+      "learning_rate": 2.9437744341141084e-06,
+      "loss": 0.7609,
+      "step": 10091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.70304675100566,
+      "learning_rate": 2.9428296945307533e-06,
+      "loss": 0.6667,
+      "step": 10092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6623374453216526,
+      "learning_rate": 2.9418850433485877e-06,
+      "loss": 0.7701,
+      "step": 10093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5135166738907408,
+      "learning_rate": 2.9409404806082077e-06,
+      "loss": 0.7254,
+      "step": 10094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9310284403140623,
+      "learning_rate": 2.939996006350201e-06,
+      "loss": 0.7351,
+      "step": 10095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9916571242728193,
+      "learning_rate": 2.939051620615152e-06,
+      "loss": 0.5785,
+      "step": 10096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.743455930943593,
+      "learning_rate": 2.938107323443645e-06,
+      "loss": 0.7339,
+      "step": 10097
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1488393082920316,
+      "learning_rate": 2.9371631148762592e-06,
+      "loss": 0.6547,
+      "step": 10098
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.493546700139166,
+      "learning_rate": 2.936218994953568e-06,
+      "loss": 0.68,
+      "step": 10099
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2014550276359584,
+      "learning_rate": 2.935274963716144e-06,
+      "loss": 0.7332,
+      "step": 10100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.615042335564596,
+      "learning_rate": 2.934331021204551e-06,
+      "loss": 0.7526,
+      "step": 10101
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.667686624609267,
+      "learning_rate": 2.9333871674593573e-06,
+      "loss": 0.7057,
+      "step": 10102
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8732573294958206,
+      "learning_rate": 2.9324434025211173e-06,
+      "loss": 0.7199,
+      "step": 10103
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3096301976594844,
+      "learning_rate": 2.93149972643039e-06,
+      "loss": 0.6675,
+      "step": 10104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.566702184022064,
+      "learning_rate": 2.930556139227726e-06,
+      "loss": 0.6999,
+      "step": 10105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5121120452910932,
+      "learning_rate": 2.929612640953675e-06,
+      "loss": 0.6835,
+      "step": 10106
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.088964213589581,
+      "learning_rate": 2.9286692316487776e-06,
+      "loss": 0.6987,
+      "step": 10107
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6145357065338586,
+      "learning_rate": 2.9277259113535774e-06,
+      "loss": 0.7764,
+      "step": 10108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.178392779559164,
+      "learning_rate": 2.9267826801086103e-06,
+      "loss": 0.5359,
+      "step": 10109
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.159271493611286,
+      "learning_rate": 2.9258395379544085e-06,
+      "loss": 0.7126,
+      "step": 10110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5962469280407336,
+      "learning_rate": 2.9248964849315004e-06,
+      "loss": 0.6669,
+      "step": 10111
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6847264056306734,
+      "learning_rate": 2.9239535210804115e-06,
+      "loss": 0.8306,
+      "step": 10112
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4686329702823164,
+      "learning_rate": 2.923010646441666e-06,
+      "loss": 0.7163,
+      "step": 10113
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6459806301698148,
+      "learning_rate": 2.9220678610557773e-06,
+      "loss": 0.7311,
+      "step": 10114
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6049926399579977,
+      "learning_rate": 2.9211251649632587e-06,
+      "loss": 0.7714,
+      "step": 10115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5619653461389944,
+      "learning_rate": 2.920182558204623e-06,
+      "loss": 0.755,
+      "step": 10116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7119968491522046,
+      "learning_rate": 2.919240040820376e-06,
+      "loss": 0.7246,
+      "step": 10117
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6511456483641436,
+      "learning_rate": 2.9182976128510177e-06,
+      "loss": 0.6891,
+      "step": 10118
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0583679734007412,
+      "learning_rate": 2.9173552743370454e-06,
+      "loss": 0.6382,
+      "step": 10119
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2921408729784885,
+      "learning_rate": 2.9164130253189594e-06,
+      "loss": 0.6075,
+      "step": 10120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.934496817502542,
+      "learning_rate": 2.9154708658372433e-06,
+      "loss": 0.6723,
+      "step": 10121
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6162268049000366,
+      "learning_rate": 2.9145287959323852e-06,
+      "loss": 0.691,
+      "step": 10122
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5959898618358663,
+      "learning_rate": 2.913586815644871e-06,
+      "loss": 0.6918,
+      "step": 10123
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5940751710583667,
+      "learning_rate": 2.912644925015179e-06,
+      "loss": 0.6396,
+      "step": 10124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1895371661452008,
+      "learning_rate": 2.9117031240837825e-06,
+      "loss": 0.6817,
+      "step": 10125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.164126984333239,
+      "learning_rate": 2.910761412891153e-06,
+      "loss": 0.7438,
+      "step": 10126
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7219590371542022,
+      "learning_rate": 2.9098197914777575e-06,
+      "loss": 0.7198,
+      "step": 10127
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9088994999917737,
+      "learning_rate": 2.9088782598840646e-06,
+      "loss": 0.7249,
+      "step": 10128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.823501939431761,
+      "learning_rate": 2.9079368181505263e-06,
+      "loss": 0.6726,
+      "step": 10129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7763919533734764,
+      "learning_rate": 2.9069954663176036e-06,
+      "loss": 0.7734,
+      "step": 10130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.180523404296923,
+      "learning_rate": 2.9060542044257487e-06,
+      "loss": 0.8099,
+      "step": 10131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0415898183906975,
+      "learning_rate": 2.9051130325154053e-06,
+      "loss": 0.5515,
+      "step": 10132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.818207492668701,
+      "learning_rate": 2.904171950627022e-06,
+      "loss": 0.7946,
+      "step": 10133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7547007376216324,
+      "learning_rate": 2.9032309588010372e-06,
+      "loss": 0.7921,
+      "step": 10134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.016880212055877,
+      "learning_rate": 2.9022900570778876e-06,
+      "loss": 0.7117,
+      "step": 10135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6456700915791491,
+      "learning_rate": 2.9013492454980074e-06,
+      "loss": 0.6686,
+      "step": 10136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9545401489807175,
+      "learning_rate": 2.9004085241018208e-06,
+      "loss": 0.7077,
+      "step": 10137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9807140322611938,
+      "learning_rate": 2.899467892929761e-06,
+      "loss": 0.7317,
+      "step": 10138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5681955984166602,
+      "learning_rate": 2.8985273520222414e-06,
+      "loss": 0.6594,
+      "step": 10139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7800741585389497,
+      "learning_rate": 2.8975869014196797e-06,
+      "loss": 0.8369,
+      "step": 10140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4772746386766853,
+      "learning_rate": 2.896646541162493e-06,
+      "loss": 0.7922,
+      "step": 10141
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9911878198649324,
+      "learning_rate": 2.895706271291089e-06,
+      "loss": 0.6758,
+      "step": 10142
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1591433204678916,
+      "learning_rate": 2.894766091845873e-06,
+      "loss": 0.7449,
+      "step": 10143
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.547990528119237,
+      "learning_rate": 2.893826002867247e-06,
+      "loss": 0.6937,
+      "step": 10144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8433494711737244,
+      "learning_rate": 2.8928860043956062e-06,
+      "loss": 0.7134,
+      "step": 10145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6907292373857539,
+      "learning_rate": 2.8919460964713507e-06,
+      "loss": 0.7197,
+      "step": 10146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6459677061263227,
+      "learning_rate": 2.891006279134862e-06,
+      "loss": 0.6574,
+      "step": 10147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5967264034889648,
+      "learning_rate": 2.8900665524265326e-06,
+      "loss": 0.8777,
+      "step": 10148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.67230018830918,
+      "learning_rate": 2.889126916386744e-06,
+      "loss": 0.7535,
+      "step": 10149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.090749232467143,
+      "learning_rate": 2.88818737105587e-06,
+      "loss": 0.6699,
+      "step": 10150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3874370966324652,
+      "learning_rate": 2.8872479164742896e-06,
+      "loss": 0.6552,
+      "step": 10151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7571365828717944,
+      "learning_rate": 2.886308552682371e-06,
+      "loss": 0.7772,
+      "step": 10152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8455339723646302,
+      "learning_rate": 2.885369279720482e-06,
+      "loss": 0.6817,
+      "step": 10153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7922225364477482,
+      "learning_rate": 2.884430097628984e-06,
+      "loss": 0.7934,
+      "step": 10154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7091760943511094,
+      "learning_rate": 2.883491006448235e-06,
+      "loss": 0.7369,
+      "step": 10155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7023821663927976,
+      "learning_rate": 2.8825520062185945e-06,
+      "loss": 0.7795,
+      "step": 10156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8849676911221867,
+      "learning_rate": 2.881613096980407e-06,
+      "loss": 0.7713,
+      "step": 10157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8255337547046786,
+      "learning_rate": 2.8806742787740237e-06,
+      "loss": 0.795,
+      "step": 10158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5728967043846005,
+      "learning_rate": 2.879735551639787e-06,
+      "loss": 0.6154,
+      "step": 10159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5446845983354105,
+      "learning_rate": 2.8787969156180353e-06,
+      "loss": 0.6421,
+      "step": 10160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1118910531265362,
+      "learning_rate": 2.877858370749105e-06,
+      "loss": 0.5908,
+      "step": 10161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0050792828111907,
+      "learning_rate": 2.8769199170733265e-06,
+      "loss": 0.6755,
+      "step": 10162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5361740443635676,
+      "learning_rate": 2.8759815546310254e-06,
+      "loss": 0.6563,
+      "step": 10163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5459708254657392,
+      "learning_rate": 2.8750432834625312e-06,
+      "loss": 0.6718,
+      "step": 10164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6133681688823018,
+      "learning_rate": 2.874105103608156e-06,
+      "loss": 0.7899,
+      "step": 10165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7477569057081244,
+      "learning_rate": 2.8731670151082213e-06,
+      "loss": 0.7865,
+      "step": 10166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.016257135630012,
+      "learning_rate": 2.872229018003036e-06,
+      "loss": 0.6958,
+      "step": 10167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1317622162055605,
+      "learning_rate": 2.871291112332909e-06,
+      "loss": 0.6753,
+      "step": 10168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.684570430636243,
+      "learning_rate": 2.8703532981381437e-06,
+      "loss": 0.6095,
+      "step": 10169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8520346084515282,
+      "learning_rate": 2.869415575459038e-06,
+      "loss": 0.7002,
+      "step": 10170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.664017812848687,
+      "learning_rate": 2.8684779443358945e-06,
+      "loss": 0.6322,
+      "step": 10171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6011454100285405,
+      "learning_rate": 2.8675404048089987e-06,
+      "loss": 0.7272,
+      "step": 10172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7773641877598871,
+      "learning_rate": 2.8666029569186393e-06,
+      "loss": 0.6713,
+      "step": 10173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.013333104523175,
+      "learning_rate": 2.8656656007051055e-06,
+      "loss": 0.8389,
+      "step": 10174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2618597258555133,
+      "learning_rate": 2.864728336208671e-06,
+      "loss": 0.7083,
+      "step": 10175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7682900936467247,
+      "learning_rate": 2.8637911634696165e-06,
+      "loss": 0.5985,
+      "step": 10176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9777716509586112,
+      "learning_rate": 2.862854082528214e-06,
+      "loss": 0.7288,
+      "step": 10177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7667456653610905,
+      "learning_rate": 2.861917093424731e-06,
+      "loss": 0.6856,
+      "step": 10178
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6983198810961302,
+      "learning_rate": 2.860980196199432e-06,
+      "loss": 0.7247,
+      "step": 10179
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.663671054467873,
+      "learning_rate": 2.8600433908925764e-06,
+      "loss": 0.7498,
+      "step": 10180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.847480702595042,
+      "learning_rate": 2.859106677544424e-06,
+      "loss": 0.7167,
+      "step": 10181
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6161333420029433,
+      "learning_rate": 2.8581700561952274e-06,
+      "loss": 0.7701,
+      "step": 10182
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5077392324646648,
+      "learning_rate": 2.8572335268852297e-06,
+      "loss": 0.7106,
+      "step": 10183
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7301871529983603,
+      "learning_rate": 2.8562970896546815e-06,
+      "loss": 0.6173,
+      "step": 10184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5736294595616331,
+      "learning_rate": 2.855360744543822e-06,
+      "loss": 0.7377,
+      "step": 10185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7596973031606198,
+      "learning_rate": 2.8544244915928876e-06,
+      "loss": 0.8081,
+      "step": 10186
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5000739967024768,
+      "learning_rate": 2.8534883308421114e-06,
+      "loss": 0.7149,
+      "step": 10187
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0474334965884915,
+      "learning_rate": 2.852552262331719e-06,
+      "loss": 0.6203,
+      "step": 10188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6537192815931425,
+      "learning_rate": 2.8516162861019437e-06,
+      "loss": 0.7087,
+      "step": 10189
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0419401269376856,
+      "learning_rate": 2.850680402192999e-06,
+      "loss": 0.8054,
+      "step": 10190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7602106257217547,
+      "learning_rate": 2.8497446106451022e-06,
+      "loss": 0.7168,
+      "step": 10191
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.020504964553268,
+      "learning_rate": 2.8488089114984725e-06,
+      "loss": 0.6907,
+      "step": 10192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.92971009235384,
+      "learning_rate": 2.8478733047933103e-06,
+      "loss": 0.7522,
+      "step": 10193
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9706651063142409,
+      "learning_rate": 2.846937790569828e-06,
+      "loss": 0.7469,
+      "step": 10194
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4588167691974865,
+      "learning_rate": 2.8460023688682237e-06,
+      "loss": 0.7043,
+      "step": 10195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4464866618434729,
+      "learning_rate": 2.8450670397286945e-06,
+      "loss": 0.7695,
+      "step": 10196
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6462938451730316,
+      "learning_rate": 2.8441318031914344e-06,
+      "loss": 0.6644,
+      "step": 10197
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5436368517955819,
+      "learning_rate": 2.8431966592966297e-06,
+      "loss": 0.6908,
+      "step": 10198
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6748011963201694,
+      "learning_rate": 2.84226160808447e-06,
+      "loss": 0.6821,
+      "step": 10199
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7067410427055576,
+      "learning_rate": 2.8413266495951366e-06,
+      "loss": 0.7026,
+      "step": 10200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8503081115592956,
+      "learning_rate": 2.8403917838688007e-06,
+      "loss": 0.7092,
+      "step": 10201
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0341858839825837,
+      "learning_rate": 2.8394570109456413e-06,
+      "loss": 0.6204,
+      "step": 10202
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6373206586041922,
+      "learning_rate": 2.838522330865826e-06,
+      "loss": 0.6421,
+      "step": 10203
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.059380348079675,
+      "learning_rate": 2.837587743669521e-06,
+      "loss": 0.5827,
+      "step": 10204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.830126438630299,
+      "learning_rate": 2.8366532493968858e-06,
+      "loss": 0.7452,
+      "step": 10205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.472770705092664,
+      "learning_rate": 2.835718848088076e-06,
+      "loss": 0.6797,
+      "step": 10206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.776906957040893,
+      "learning_rate": 2.8347845397832517e-06,
+      "loss": 0.7347,
+      "step": 10207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2941850825474823,
+      "learning_rate": 2.8338503245225544e-06,
+      "loss": 0.6646,
+      "step": 10208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7181986080458986,
+      "learning_rate": 2.8329162023461355e-06,
+      "loss": 0.6496,
+      "step": 10209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5599415205752571,
+      "learning_rate": 2.8319821732941345e-06,
+      "loss": 0.7473,
+      "step": 10210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8034548287513112,
+      "learning_rate": 2.8310482374066843e-06,
+      "loss": 0.7573,
+      "step": 10211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2221287438900659,
+      "learning_rate": 2.830114394723924e-06,
+      "loss": 0.6716,
+      "step": 10212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.665964134468667,
+      "learning_rate": 2.8291806452859803e-06,
+      "loss": 0.6822,
+      "step": 10213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2587034104689616,
+      "learning_rate": 2.82824698913298e-06,
+      "loss": 0.8463,
+      "step": 10214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0928148834086553,
+      "learning_rate": 2.827313426305042e-06,
+      "loss": 0.6863,
+      "step": 10215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4948670211220776,
+      "learning_rate": 2.826379956842284e-06,
+      "loss": 0.6485,
+      "step": 10216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1740963542560139,
+      "learning_rate": 2.825446580784822e-06,
+      "loss": 0.7433,
+      "step": 10217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5915484297783107,
+      "learning_rate": 2.824513298172763e-06,
+      "loss": 0.7472,
+      "step": 10218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1266766288010714,
+      "learning_rate": 2.823580109046212e-06,
+      "loss": 0.6126,
+      "step": 10219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3721292981266862,
+      "learning_rate": 2.822647013445272e-06,
+      "loss": 0.6583,
+      "step": 10220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5380622481122155,
+      "learning_rate": 2.821714011410039e-06,
+      "loss": 0.6076,
+      "step": 10221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.23816772235924,
+      "learning_rate": 2.820781102980606e-06,
+      "loss": 0.7625,
+      "step": 10222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.142011301780901,
+      "learning_rate": 2.8198482881970626e-06,
+      "loss": 0.6224,
+      "step": 10223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1781194136637538,
+      "learning_rate": 2.8189155670994913e-06,
+      "loss": 0.6513,
+      "step": 10224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.824225737860651,
+      "learning_rate": 2.8179829397279795e-06,
+      "loss": 0.7056,
+      "step": 10225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9255919115614897,
+      "learning_rate": 2.8170504061225966e-06,
+      "loss": 0.6389,
+      "step": 10226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.596740886753346,
+      "learning_rate": 2.8161179663234215e-06,
+      "loss": 0.7446,
+      "step": 10227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7283977461353515,
+      "learning_rate": 2.8151856203705215e-06,
+      "loss": 0.7527,
+      "step": 10228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6715504983486444,
+      "learning_rate": 2.814253368303961e-06,
+      "loss": 0.8768,
+      "step": 10229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5984868003097914,
+      "learning_rate": 2.813321210163801e-06,
+      "loss": 0.6784,
+      "step": 10230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7581669396891255,
+      "learning_rate": 2.8123891459900967e-06,
+      "loss": 0.6543,
+      "step": 10231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6808560798745886,
+      "learning_rate": 2.8114571758229064e-06,
+      "loss": 0.7195,
+      "step": 10232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6642781138944374,
+      "learning_rate": 2.8105252997022737e-06,
+      "loss": 0.7668,
+      "step": 10233
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.73366777749611,
+      "learning_rate": 2.809593517668243e-06,
+      "loss": 0.6783,
+      "step": 10234
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6674133851465205,
+      "learning_rate": 2.80866182976086e-06,
+      "loss": 0.6547,
+      "step": 10235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7365309476172017,
+      "learning_rate": 2.8077302360201575e-06,
+      "loss": 0.6983,
+      "step": 10236
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.745849331051467,
+      "learning_rate": 2.80679873648617e-06,
+      "loss": 0.7459,
+      "step": 10237
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0931409544770276,
+      "learning_rate": 2.8058673311989247e-06,
+      "loss": 0.6082,
+      "step": 10238
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2736020459108222,
+      "learning_rate": 2.804936020198447e-06,
+      "loss": 0.679,
+      "step": 10239
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5198148835775966,
+      "learning_rate": 2.8040048035247576e-06,
+      "loss": 0.6884,
+      "step": 10240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8943955169037698,
+      "learning_rate": 2.8030736812178717e-06,
+      "loss": 0.7904,
+      "step": 10241
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.826307623583513,
+      "learning_rate": 2.8021426533178007e-06,
+      "loss": 0.8341,
+      "step": 10242
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5836224379735044,
+      "learning_rate": 2.8012117198645596e-06,
+      "loss": 0.7253,
+      "step": 10243
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7099127772623313,
+      "learning_rate": 2.800280880898143e-06,
+      "loss": 0.6741,
+      "step": 10244
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.180426687703099,
+      "learning_rate": 2.7993501364585584e-06,
+      "loss": 0.7424,
+      "step": 10245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6485412888561775,
+      "learning_rate": 2.7984194865857994e-06,
+      "loss": 0.7539,
+      "step": 10246
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6337707383479725,
+      "learning_rate": 2.7974889313198584e-06,
+      "loss": 0.7116,
+      "step": 10247
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7849971824924815,
+      "learning_rate": 2.796558470700723e-06,
+      "loss": 0.7224,
+      "step": 10248
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7389733228019353,
+      "learning_rate": 2.795628104768376e-06,
+      "loss": 0.7293,
+      "step": 10249
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4560042987622233,
+      "learning_rate": 2.7946978335628023e-06,
+      "loss": 0.7055,
+      "step": 10250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0507223964449017,
+      "learning_rate": 2.793767657123972e-06,
+      "loss": 0.7109,
+      "step": 10251
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1686300134664227,
+      "learning_rate": 2.792837575491857e-06,
+      "loss": 0.6907,
+      "step": 10252
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.076952219197869,
+      "learning_rate": 2.7919075887064296e-06,
+      "loss": 0.7913,
+      "step": 10253
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5329657271817294,
+      "learning_rate": 2.79097769680765e-06,
+      "loss": 0.6377,
+      "step": 10254
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1145452014760964,
+      "learning_rate": 2.790047899835479e-06,
+      "loss": 0.7075,
+      "step": 10255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3211579940258007,
+      "learning_rate": 2.789118197829872e-06,
+      "loss": 0.6878,
+      "step": 10256
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7877295776844107,
+      "learning_rate": 2.7881885908307794e-06,
+      "loss": 0.7252,
+      "step": 10257
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8977728970876215,
+      "learning_rate": 2.7872590788781483e-06,
+      "loss": 0.7814,
+      "step": 10258
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.660347095911191,
+      "learning_rate": 2.7863296620119217e-06,
+      "loss": 0.6336,
+      "step": 10259
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7023482057186736,
+      "learning_rate": 2.7854003402720416e-06,
+      "loss": 0.6872,
+      "step": 10260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7367259410347673,
+      "learning_rate": 2.7844711136984425e-06,
+      "loss": 0.6621,
+      "step": 10261
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3567726240918443,
+      "learning_rate": 2.7835419823310507e-06,
+      "loss": 0.7029,
+      "step": 10262
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.785817253867552,
+      "learning_rate": 2.7826129462097974e-06,
+      "loss": 0.8133,
+      "step": 10263
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6864320873749277,
+      "learning_rate": 2.781684005374604e-06,
+      "loss": 0.7279,
+      "step": 10264
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.300104066236513,
+      "learning_rate": 2.7807551598653905e-06,
+      "loss": 0.6262,
+      "step": 10265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.595091719214806,
+      "learning_rate": 2.7798264097220694e-06,
+      "loss": 0.7884,
+      "step": 10266
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.77547907632745,
+      "learning_rate": 2.7788977549845503e-06,
+      "loss": 0.6565,
+      "step": 10267
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4944848416185403,
+      "learning_rate": 2.777969195692745e-06,
+      "loss": 0.6975,
+      "step": 10268
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7823529750461984,
+      "learning_rate": 2.777040731886549e-06,
+      "loss": 0.8039,
+      "step": 10269
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.934210712838425,
+      "learning_rate": 2.7761123636058645e-06,
+      "loss": 0.7649,
+      "step": 10270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1488497324403113,
+      "learning_rate": 2.775184090890585e-06,
+      "loss": 0.7663,
+      "step": 10271
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3945415870307782,
+      "learning_rate": 2.7742559137806e-06,
+      "loss": 0.6589,
+      "step": 10272
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8138277365573943,
+      "learning_rate": 2.773327832315796e-06,
+      "loss": 0.679,
+      "step": 10273
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.988298627892705,
+      "learning_rate": 2.7723998465360537e-06,
+      "loss": 0.7247,
+      "step": 10274
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4587762386075007,
+      "learning_rate": 2.771471956481251e-06,
+      "loss": 0.7201,
+      "step": 10275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8007972684910742,
+      "learning_rate": 2.770544162191261e-06,
+      "loss": 0.6729,
+      "step": 10276
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.618347905053086,
+      "learning_rate": 2.7696164637059527e-06,
+      "loss": 0.6799,
+      "step": 10277
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6673685513043428,
+      "learning_rate": 2.7686888610651936e-06,
+      "loss": 0.7068,
+      "step": 10278
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6805538501954356,
+      "learning_rate": 2.7677613543088432e-06,
+      "loss": 0.6681,
+      "step": 10279
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.875288171986117,
+      "learning_rate": 2.76683394347676e-06,
+      "loss": 0.7954,
+      "step": 10280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0607776561272977,
+      "learning_rate": 2.7659066286087947e-06,
+      "loss": 0.6382,
+      "step": 10281
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.645136997201299,
+      "learning_rate": 2.7649794097447956e-06,
+      "loss": 0.6834,
+      "step": 10282
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0258185041489951,
+      "learning_rate": 2.7640522869246134e-06,
+      "loss": 0.6385,
+      "step": 10283
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5397467898761095,
+      "learning_rate": 2.7631252601880816e-06,
+      "loss": 0.7059,
+      "step": 10284
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5631194604923808,
+      "learning_rate": 2.762198329575037e-06,
+      "loss": 0.8319,
+      "step": 10285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5777997467421654,
+      "learning_rate": 2.7612714951253183e-06,
+      "loss": 0.635,
+      "step": 10286
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.607849383336294,
+      "learning_rate": 2.7603447568787456e-06,
+      "loss": 0.644,
+      "step": 10287
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1016102524515905,
+      "learning_rate": 2.759418114875148e-06,
+      "loss": 0.5964,
+      "step": 10288
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7832090885153489,
+      "learning_rate": 2.7584915691543444e-06,
+      "loss": 0.6941,
+      "step": 10289
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7885753910829532,
+      "learning_rate": 2.7575651197561504e-06,
+      "loss": 1.0469,
+      "step": 10290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.924772952950346,
+      "learning_rate": 2.7566387667203776e-06,
+      "loss": 0.7789,
+      "step": 10291
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.048402376911717,
+      "learning_rate": 2.7557125100868327e-06,
+      "loss": 0.7515,
+      "step": 10292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.645683827325185,
+      "learning_rate": 2.75478634989532e-06,
+      "loss": 0.7876,
+      "step": 10293
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6939002393221017,
+      "learning_rate": 2.753860286185637e-06,
+      "loss": 0.6571,
+      "step": 10294
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7226175182943362,
+      "learning_rate": 2.7529343189975788e-06,
+      "loss": 0.7083,
+      "step": 10295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8268097177236229,
+      "learning_rate": 2.7520084483709397e-06,
+      "loss": 0.7498,
+      "step": 10296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8665860344230825,
+      "learning_rate": 2.7510826743455037e-06,
+      "loss": 0.6491,
+      "step": 10297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5226839932174903,
+      "learning_rate": 2.7501569969610544e-06,
+      "loss": 0.6461,
+      "step": 10298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4819120895792477,
+      "learning_rate": 2.7492314162573687e-06,
+      "loss": 0.6505,
+      "step": 10299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9054542385401723,
+      "learning_rate": 2.74830593227422e-06,
+      "loss": 0.7226,
+      "step": 10300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5614631667544066,
+      "learning_rate": 2.7473805450513846e-06,
+      "loss": 0.7416,
+      "step": 10301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6135413804089365,
+      "learning_rate": 2.7464552546286217e-06,
+      "loss": 0.6592,
+      "step": 10302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7742722599430225,
+      "learning_rate": 2.7455300610456935e-06,
+      "loss": 0.6566,
+      "step": 10303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8042520484154945,
+      "learning_rate": 2.744604964342364e-06,
+      "loss": 0.7759,
+      "step": 10304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5597558458486023,
+      "learning_rate": 2.743679964558378e-06,
+      "loss": 0.7331,
+      "step": 10305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9583163631467906,
+      "learning_rate": 2.7427550617334905e-06,
+      "loss": 0.6898,
+      "step": 10306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6847307109979333,
+      "learning_rate": 2.741830255907445e-06,
+      "loss": 0.6488,
+      "step": 10307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5336275789689948,
+      "learning_rate": 2.740905547119983e-06,
+      "loss": 0.8064,
+      "step": 10308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.110523453109782,
+      "learning_rate": 2.7399809354108415e-06,
+      "loss": 0.7296,
+      "step": 10309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.766007826850305,
+      "learning_rate": 2.73905642081975e-06,
+      "loss": 0.7141,
+      "step": 10310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5751335446462993,
+      "learning_rate": 2.7381320033864434e-06,
+      "loss": 0.7569,
+      "step": 10311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6715484196492325,
+      "learning_rate": 2.73720768315064e-06,
+      "loss": 0.5584,
+      "step": 10312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.552963291942047,
+      "learning_rate": 2.7362834601520603e-06,
+      "loss": 0.691,
+      "step": 10313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6705103422394538,
+      "learning_rate": 2.735359334430424e-06,
+      "loss": 0.7319,
+      "step": 10314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.979645219149573,
+      "learning_rate": 2.73443530602544e-06,
+      "loss": 0.7418,
+      "step": 10315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.668969336803651,
+      "learning_rate": 2.733511374976817e-06,
+      "loss": 0.6416,
+      "step": 10316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8362520285775532,
+      "learning_rate": 2.7325875413242585e-06,
+      "loss": 0.6612,
+      "step": 10317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6291373107995932,
+      "learning_rate": 2.7316638051074605e-06,
+      "loss": 0.65,
+      "step": 10318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7717279713795935,
+      "learning_rate": 2.7307401663661247e-06,
+      "loss": 0.7092,
+      "step": 10319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9736580480768946,
+      "learning_rate": 2.729816625139934e-06,
+      "loss": 0.7135,
+      "step": 10320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6561368796224785,
+      "learning_rate": 2.728893181468581e-06,
+      "loss": 0.6257,
+      "step": 10321
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4683139751873626,
+      "learning_rate": 2.7279698353917476e-06,
+      "loss": 0.6374,
+      "step": 10322
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6670858565471207,
+      "learning_rate": 2.727046586949107e-06,
+      "loss": 0.6565,
+      "step": 10323
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6083742450723413,
+      "learning_rate": 2.7261234361803383e-06,
+      "loss": 0.7244,
+      "step": 10324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5692252749957505,
+      "learning_rate": 2.72520038312511e-06,
+      "loss": 0.7113,
+      "step": 10325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6996839471844947,
+      "learning_rate": 2.7242774278230876e-06,
+      "loss": 0.6342,
+      "step": 10326
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6947544407930264,
+      "learning_rate": 2.723354570313932e-06,
+      "loss": 0.7515,
+      "step": 10327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.71157338120634,
+      "learning_rate": 2.7224318106372995e-06,
+      "loss": 0.7847,
+      "step": 10328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7306746094311332,
+      "learning_rate": 2.721509148832847e-06,
+      "loss": 0.6311,
+      "step": 10329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.450519017608128,
+      "learning_rate": 2.7205865849402187e-06,
+      "loss": 0.7229,
+      "step": 10330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.777536826159543,
+      "learning_rate": 2.7196641189990623e-06,
+      "loss": 0.8754,
+      "step": 10331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6017292596246189,
+      "learning_rate": 2.7187417510490176e-06,
+      "loss": 0.7704,
+      "step": 10332
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6386308700303711,
+      "learning_rate": 2.7178194811297204e-06,
+      "loss": 0.704,
+      "step": 10333
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.036792025992493,
+      "learning_rate": 2.7168973092808025e-06,
+      "loss": 0.6711,
+      "step": 10334
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8562452640461484,
+      "learning_rate": 2.7159752355418927e-06,
+      "loss": 0.7183,
+      "step": 10335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5635539631629214,
+      "learning_rate": 2.7150532599526114e-06,
+      "loss": 0.7886,
+      "step": 10336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6977898469160215,
+      "learning_rate": 2.7141313825525845e-06,
+      "loss": 0.6571,
+      "step": 10337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9858316730725927,
+      "learning_rate": 2.7132096033814193e-06,
+      "loss": 0.6806,
+      "step": 10338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8100284555224775,
+      "learning_rate": 2.7122879224787315e-06,
+      "loss": 0.7681,
+      "step": 10339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7783973388795395,
+      "learning_rate": 2.7113663398841296e-06,
+      "loss": 0.6307,
+      "step": 10340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6221606052247755,
+      "learning_rate": 2.7104448556372086e-06,
+      "loss": 0.7329,
+      "step": 10341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5754743983220223,
+      "learning_rate": 2.709523469777573e-06,
+      "loss": 0.8051,
+      "step": 10342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9132642062178415,
+      "learning_rate": 2.7086021823448155e-06,
+      "loss": 0.6027,
+      "step": 10343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7850269864997448,
+      "learning_rate": 2.7076809933785254e-06,
+      "loss": 0.7337,
+      "step": 10344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1693732786414053,
+      "learning_rate": 2.7067599029182884e-06,
+      "loss": 0.6982,
+      "step": 10345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7870594947129292,
+      "learning_rate": 2.7058389110036835e-06,
+      "loss": 0.7186,
+      "step": 10346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9123630835079686,
+      "learning_rate": 2.704918017674294e-06,
+      "loss": 0.8272,
+      "step": 10347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6141392043774727,
+      "learning_rate": 2.7039972229696847e-06,
+      "loss": 0.7942,
+      "step": 10348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7595275661417118,
+      "learning_rate": 2.70307652692943e-06,
+      "loss": 0.6636,
+      "step": 10349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.261828737139329,
+      "learning_rate": 2.702155929593093e-06,
+      "loss": 0.7077,
+      "step": 10350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4733586069374927,
+      "learning_rate": 2.7012354310002333e-06,
+      "loss": 0.707,
+      "step": 10351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.248245552161178,
+      "learning_rate": 2.700315031190406e-06,
+      "loss": 0.7475,
+      "step": 10352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6525653628039134,
+      "learning_rate": 2.6993947302031643e-06,
+      "loss": 0.701,
+      "step": 10353
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8563700881163556,
+      "learning_rate": 2.6984745280780524e-06,
+      "loss": 0.7803,
+      "step": 10354
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7556102273767749,
+      "learning_rate": 2.6975544248546195e-06,
+      "loss": 0.668,
+      "step": 10355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.696712215838192,
+      "learning_rate": 2.696634420572397e-06,
+      "loss": 0.703,
+      "step": 10356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9132305658733788,
+      "learning_rate": 2.6957145152709253e-06,
+      "loss": 0.7376,
+      "step": 10357
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6905654388645899,
+      "learning_rate": 2.694794708989733e-06,
+      "loss": 0.7473,
+      "step": 10358
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.014817944135091,
+      "learning_rate": 2.6938750017683457e-06,
+      "loss": 0.7408,
+      "step": 10359
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7919151934615754,
+      "learning_rate": 2.692955393646286e-06,
+      "loss": 0.7567,
+      "step": 10360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7762471820729637,
+      "learning_rate": 2.6920358846630688e-06,
+      "loss": 0.6746,
+      "step": 10361
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.026392613920886,
+      "learning_rate": 2.691116474858214e-06,
+      "loss": 0.625,
+      "step": 10362
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.834197709056897,
+      "learning_rate": 2.690197164271224e-06,
+      "loss": 0.7704,
+      "step": 10363
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6722286620907607,
+      "learning_rate": 2.6892779529416045e-06,
+      "loss": 0.5753,
+      "step": 10364
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6502635101290346,
+      "learning_rate": 2.6883588409088597e-06,
+      "loss": 0.6158,
+      "step": 10365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8798080866728877,
+      "learning_rate": 2.6874398282124837e-06,
+      "loss": 0.6929,
+      "step": 10366
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3292155899945874,
+      "learning_rate": 2.686520914891968e-06,
+      "loss": 0.5944,
+      "step": 10367
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1883413803363367,
+      "learning_rate": 2.6856021009868016e-06,
+      "loss": 0.67,
+      "step": 10368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8017255789968862,
+      "learning_rate": 2.6846833865364674e-06,
+      "loss": 0.6564,
+      "step": 10369
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0499894939626055,
+      "learning_rate": 2.6837647715804444e-06,
+      "loss": 0.6902,
+      "step": 10370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8355637630647939,
+      "learning_rate": 2.682846256158206e-06,
+      "loss": 0.7368,
+      "step": 10371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7942964992928703,
+      "learning_rate": 2.6819278403092258e-06,
+      "loss": 0.7198,
+      "step": 10372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4110467667026387,
+      "learning_rate": 2.6810095240729715e-06,
+      "loss": 0.7026,
+      "step": 10373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8021256187408863,
+      "learning_rate": 2.6800913074888984e-06,
+      "loss": 0.7511,
+      "step": 10374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7675836587690135,
+      "learning_rate": 2.679173190596471e-06,
+      "loss": 0.8028,
+      "step": 10375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5736167877666967,
+      "learning_rate": 2.6782551734351396e-06,
+      "loss": 0.6596,
+      "step": 10376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1612095958843662,
+      "learning_rate": 2.6773372560443543e-06,
+      "loss": 0.6636,
+      "step": 10377
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.108528417636058,
+      "learning_rate": 2.6764194384635607e-06,
+      "loss": 0.8294,
+      "step": 10378
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8414279700383511,
+      "learning_rate": 2.6755017207321964e-06,
+      "loss": 0.8175,
+      "step": 10379
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7903056034353126,
+      "learning_rate": 2.6745841028897034e-06,
+      "loss": 0.7337,
+      "step": 10380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.800747888976762,
+      "learning_rate": 2.6736665849755073e-06,
+      "loss": 0.8395,
+      "step": 10381
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8529589345020818,
+      "learning_rate": 2.6727491670290405e-06,
+      "loss": 0.8135,
+      "step": 10382
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5798625066566292,
+      "learning_rate": 2.6718318490897256e-06,
+      "loss": 0.7007,
+      "step": 10383
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7589296755542951,
+      "learning_rate": 2.6709146311969813e-06,
+      "loss": 0.8783,
+      "step": 10384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6963246787328483,
+      "learning_rate": 2.669997513390222e-06,
+      "loss": 0.638,
+      "step": 10385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7847728761655626,
+      "learning_rate": 2.669080495708859e-06,
+      "loss": 0.7187,
+      "step": 10386
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8360468629215458,
+      "learning_rate": 2.668163578192298e-06,
+      "loss": 0.673,
+      "step": 10387
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7497800490743904,
+      "learning_rate": 2.6672467608799413e-06,
+      "loss": 0.8088,
+      "step": 10388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7666687153781946,
+      "learning_rate": 2.666330043811185e-06,
+      "loss": 0.7426,
+      "step": 10389
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4010956287141885,
+      "learning_rate": 2.6654134270254263e-06,
+      "loss": 0.6555,
+      "step": 10390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7165628721402157,
+      "learning_rate": 2.6644969105620535e-06,
+      "loss": 0.7257,
+      "step": 10391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9080192580107953,
+      "learning_rate": 2.663580494460446e-06,
+      "loss": 0.6898,
+      "step": 10392
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.520482467345118,
+      "learning_rate": 2.6626641787599904e-06,
+      "loss": 0.6578,
+      "step": 10393
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7502953163601782,
+      "learning_rate": 2.66174796350006e-06,
+      "loss": 0.6693,
+      "step": 10394
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.750393687503268,
+      "learning_rate": 2.660831848720028e-06,
+      "loss": 0.7298,
+      "step": 10395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7804769598959083,
+      "learning_rate": 2.6599158344592606e-06,
+      "loss": 0.7958,
+      "step": 10396
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.690162307646254,
+      "learning_rate": 2.658999920757119e-06,
+      "loss": 0.589,
+      "step": 10397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6234415755903195,
+      "learning_rate": 2.6580841076529697e-06,
+      "loss": 0.6919,
+      "step": 10398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.88525317611182,
+      "learning_rate": 2.657168395186157e-06,
+      "loss": 0.7904,
+      "step": 10399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0652841493702967,
+      "learning_rate": 2.656252783396038e-06,
+      "loss": 0.7962,
+      "step": 10400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1079406190645313,
+      "learning_rate": 2.655337272321957e-06,
+      "loss": 0.754,
+      "step": 10401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0559071906785251,
+      "learning_rate": 2.654421862003256e-06,
+      "loss": 0.7112,
+      "step": 10402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6052936604511026,
+      "learning_rate": 2.653506552479271e-06,
+      "loss": 0.7684,
+      "step": 10403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7521941511929038,
+      "learning_rate": 2.6525913437893346e-06,
+      "loss": 0.7825,
+      "step": 10404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7950328053532796,
+      "learning_rate": 2.651676235972777e-06,
+      "loss": 0.6929,
+      "step": 10405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5595744525070954,
+      "learning_rate": 2.650761229068921e-06,
+      "loss": 0.6388,
+      "step": 10406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7233350393549582,
+      "learning_rate": 2.649846323117086e-06,
+      "loss": 0.7052,
+      "step": 10407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6096846336405501,
+      "learning_rate": 2.6489315181565896e-06,
+      "loss": 0.5905,
+      "step": 10408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8530972148770426,
+      "learning_rate": 2.648016814226742e-06,
+      "loss": 0.6934,
+      "step": 10409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.927311252339605,
+      "learning_rate": 2.6471022113668505e-06,
+      "loss": 0.7359,
+      "step": 10410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7918767576934873,
+      "learning_rate": 2.646187709616217e-06,
+      "loss": 0.683,
+      "step": 10411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7618548724906788,
+      "learning_rate": 2.645273309014139e-06,
+      "loss": 0.6786,
+      "step": 10412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7344395202946257,
+      "learning_rate": 2.6443590095999127e-06,
+      "loss": 0.6394,
+      "step": 10413
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9592479851637847,
+      "learning_rate": 2.6434448114128252e-06,
+      "loss": 0.7567,
+      "step": 10414
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4428998378305729,
+      "learning_rate": 2.642530714492161e-06,
+      "loss": 0.6686,
+      "step": 10415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8480111660463685,
+      "learning_rate": 2.6416167188772052e-06,
+      "loss": 0.7799,
+      "step": 10416
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7803799572932957,
+      "learning_rate": 2.6407028246072287e-06,
+      "loss": 0.7947,
+      "step": 10417
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8978090827466867,
+      "learning_rate": 2.6397890317215067e-06,
+      "loss": 0.7451,
+      "step": 10418
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.551392772292236,
+      "learning_rate": 2.6388753402593083e-06,
+      "loss": 0.851,
+      "step": 10419
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9753223685707314,
+      "learning_rate": 2.637961750259894e-06,
+      "loss": 0.6045,
+      "step": 10420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.716204156967755,
+      "learning_rate": 2.6370482617625236e-06,
+      "loss": 0.703,
+      "step": 10421
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.908289824264548,
+      "learning_rate": 2.636134874806451e-06,
+      "loss": 0.8564,
+      "step": 10422
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8370875774649313,
+      "learning_rate": 2.6352215894309306e-06,
+      "loss": 0.845,
+      "step": 10423
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.722718642348664,
+      "learning_rate": 2.6343084056752032e-06,
+      "loss": 0.6316,
+      "step": 10424
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8224775579907164,
+      "learning_rate": 2.6333953235785105e-06,
+      "loss": 0.6441,
+      "step": 10425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6452531028877158,
+      "learning_rate": 2.632482343180094e-06,
+      "loss": 0.8294,
+      "step": 10426
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6595099329616647,
+      "learning_rate": 2.631569464519183e-06,
+      "loss": 0.6944,
+      "step": 10427
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7248355321126951,
+      "learning_rate": 2.6306566876350072e-06,
+      "loss": 0.6948,
+      "step": 10428
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6790843194968228,
+      "learning_rate": 2.6297440125667904e-06,
+      "loss": 0.7725,
+      "step": 10429
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4721040027741186,
+      "learning_rate": 2.6288314393537522e-06,
+      "loss": 0.6397,
+      "step": 10430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.516007291494785,
+      "learning_rate": 2.627918968035108e-06,
+      "loss": 0.7205,
+      "step": 10431
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6068312855094504,
+      "learning_rate": 2.627006598650067e-06,
+      "loss": 0.6861,
+      "step": 10432
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6416513675766167,
+      "learning_rate": 2.6260943312378384e-06,
+      "loss": 0.7098,
+      "step": 10433
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5864880802929027,
+      "learning_rate": 2.6251821658376265e-06,
+      "loss": 0.7594,
+      "step": 10434
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.796270753084642,
+      "learning_rate": 2.624270102488621e-06,
+      "loss": 0.7348,
+      "step": 10435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8066193147666638,
+      "learning_rate": 2.6233581412300222e-06,
+      "loss": 0.6752,
+      "step": 10436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7888733144812272,
+      "learning_rate": 2.6224462821010185e-06,
+      "loss": 0.8245,
+      "step": 10437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.851882177980771,
+      "learning_rate": 2.621534525140792e-06,
+      "loss": 0.7533,
+      "step": 10438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6775335978280725,
+      "learning_rate": 2.620622870388524e-06,
+      "loss": 0.7472,
+      "step": 10439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.607016335319473,
+      "learning_rate": 2.619711317883389e-06,
+      "loss": 0.632,
+      "step": 10440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6827872716445604,
+      "learning_rate": 2.6187998676645633e-06,
+      "loss": 0.756,
+      "step": 10441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7193239466084536,
+      "learning_rate": 2.617888519771207e-06,
+      "loss": 0.6643,
+      "step": 10442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.008069833221627,
+      "learning_rate": 2.616977274242488e-06,
+      "loss": 0.6562,
+      "step": 10443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7061227410745219,
+      "learning_rate": 2.616066131117563e-06,
+      "loss": 0.6905,
+      "step": 10444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.205811432289692,
+      "learning_rate": 2.6151550904355853e-06,
+      "loss": 0.7959,
+      "step": 10445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7167766179491926,
+      "learning_rate": 2.6142441522357055e-06,
+      "loss": 0.6764,
+      "step": 10446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.851203445789448,
+      "learning_rate": 2.6133333165570667e-06,
+      "loss": 0.693,
+      "step": 10447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0501696986021605,
+      "learning_rate": 2.612422583438811e-06,
+      "loss": 0.7048,
+      "step": 10448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.1821944926963095,
+      "learning_rate": 2.6115119529200748e-06,
+      "loss": 0.7899,
+      "step": 10449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2089522529998478,
+      "learning_rate": 2.6106014250399868e-06,
+      "loss": 0.6829,
+      "step": 10450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4122596977405646,
+      "learning_rate": 2.6096909998376794e-06,
+      "loss": 0.8131,
+      "step": 10451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6060847368681588,
+      "learning_rate": 2.6087806773522755e-06,
+      "loss": 0.657,
+      "step": 10452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8574539685029208,
+      "learning_rate": 2.6078704576228876e-06,
+      "loss": 0.6981,
+      "step": 10453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.158574680762581,
+      "learning_rate": 2.6069603406886347e-06,
+      "loss": 0.7,
+      "step": 10454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1737676824011345,
+      "learning_rate": 2.606050326588626e-06,
+      "loss": 0.7082,
+      "step": 10455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3715731755996488,
+      "learning_rate": 2.605140415361967e-06,
+      "loss": 0.8099,
+      "step": 10456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5750149987308701,
+      "learning_rate": 2.6042306070477576e-06,
+      "loss": 0.6299,
+      "step": 10457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9266178327480117,
+      "learning_rate": 2.6033209016850926e-06,
+      "loss": 0.7936,
+      "step": 10458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6466835922271112,
+      "learning_rate": 2.60241129931307e-06,
+      "loss": 0.7956,
+      "step": 10459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9385480049101178,
+      "learning_rate": 2.601501799970769e-06,
+      "loss": 0.7276,
+      "step": 10460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.608856082878401,
+      "learning_rate": 2.6005924036972807e-06,
+      "loss": 0.753,
+      "step": 10461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9522997107205229,
+      "learning_rate": 2.59968311053168e-06,
+      "loss": 0.6087,
+      "step": 10462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1378098689416714,
+      "learning_rate": 2.5987739205130413e-06,
+      "loss": 0.6114,
+      "step": 10463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.752483983383137,
+      "learning_rate": 2.597864833680436e-06,
+      "loss": 0.7042,
+      "step": 10464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1122509312982087,
+      "learning_rate": 2.596955850072928e-06,
+      "loss": 0.7101,
+      "step": 10465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2055331789068473,
+      "learning_rate": 2.5960469697295794e-06,
+      "loss": 0.6613,
+      "step": 10466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4555837291371785,
+      "learning_rate": 2.595138192689446e-06,
+      "loss": 0.6583,
+      "step": 10467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7957357116630368,
+      "learning_rate": 2.594229518991579e-06,
+      "loss": 0.7228,
+      "step": 10468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5929960202802076,
+      "learning_rate": 2.593320948675029e-06,
+      "loss": 0.5781,
+      "step": 10469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.083734756641774,
+      "learning_rate": 2.5924124817788377e-06,
+      "loss": 0.7833,
+      "step": 10470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7968667058611616,
+      "learning_rate": 2.591504118342044e-06,
+      "loss": 0.8427,
+      "step": 10471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6591464502460875,
+      "learning_rate": 2.5905958584036826e-06,
+      "loss": 0.6969,
+      "step": 10472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8279685069941567,
+      "learning_rate": 2.5896877020027813e-06,
+      "loss": 0.7443,
+      "step": 10473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7311214325796915,
+      "learning_rate": 2.588779649178371e-06,
+      "loss": 0.718,
+      "step": 10474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.819709876344288,
+      "learning_rate": 2.5878716999694675e-06,
+      "loss": 0.8469,
+      "step": 10475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8688470803922386,
+      "learning_rate": 2.5869638544150876e-06,
+      "loss": 0.7652,
+      "step": 10476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.647290946213791,
+      "learning_rate": 2.586056112554248e-06,
+      "loss": 0.6296,
+      "step": 10477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5785081310272526,
+      "learning_rate": 2.5851484744259503e-06,
+      "loss": 0.6179,
+      "step": 10478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6031453719538988,
+      "learning_rate": 2.5842409400692026e-06,
+      "loss": 0.6943,
+      "step": 10479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1359777214013682,
+      "learning_rate": 2.5833335095230016e-06,
+      "loss": 0.6358,
+      "step": 10480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7531407547297522,
+      "learning_rate": 2.5824261828263424e-06,
+      "loss": 0.7663,
+      "step": 10481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.229782639567281,
+      "learning_rate": 2.5815189600182145e-06,
+      "loss": 0.6232,
+      "step": 10482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8089787815655904,
+      "learning_rate": 2.5806118411376e-06,
+      "loss": 0.7837,
+      "step": 10483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9697742123580277,
+      "learning_rate": 2.579704826223488e-06,
+      "loss": 0.6008,
+      "step": 10484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.540917836061923,
+      "learning_rate": 2.578797915314848e-06,
+      "loss": 0.6722,
+      "step": 10485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8942532309699966,
+      "learning_rate": 2.577891108450651e-06,
+      "loss": 0.7439,
+      "step": 10486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.741115018224377,
+      "learning_rate": 2.5769844056698702e-06,
+      "loss": 0.7083,
+      "step": 10487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1541905599557811,
+      "learning_rate": 2.576077807011465e-06,
+      "loss": 0.6648,
+      "step": 10488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.402332218000173,
+      "learning_rate": 2.575171312514395e-06,
+      "loss": 0.6922,
+      "step": 10489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6433962113935232,
+      "learning_rate": 2.5742649222176147e-06,
+      "loss": 0.6514,
+      "step": 10490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8105246384310498,
+      "learning_rate": 2.573358636160071e-06,
+      "loss": 0.7541,
+      "step": 10491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6555814965192102,
+      "learning_rate": 2.572452454380715e-06,
+      "loss": 0.6875,
+      "step": 10492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6116531682708628,
+      "learning_rate": 2.571546376918479e-06,
+      "loss": 0.7329,
+      "step": 10493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.640912579249429,
+      "learning_rate": 2.570640403812306e-06,
+      "loss": 0.7702,
+      "step": 10494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6025482498191597,
+      "learning_rate": 2.569734535101127e-06,
+      "loss": 0.7039,
+      "step": 10495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6633616322049054,
+      "learning_rate": 2.5688287708238647e-06,
+      "loss": 0.748,
+      "step": 10496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.406053126535999,
+      "learning_rate": 2.5679231110194454e-06,
+      "loss": 0.693,
+      "step": 10497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6205778789878251,
+      "learning_rate": 2.567017555726787e-06,
+      "loss": 0.7006,
+      "step": 10498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.250686232127316,
+      "learning_rate": 2.5661121049848026e-06,
+      "loss": 0.7329,
+      "step": 10499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5656326594461123,
+      "learning_rate": 2.5652067588324015e-06,
+      "loss": 0.5921,
+      "step": 10500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8213397060619398,
+      "learning_rate": 2.564301517308487e-06,
+      "loss": 0.6896,
+      "step": 10501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9510996055131338,
+      "learning_rate": 2.5633963804519623e-06,
+      "loss": 0.812,
+      "step": 10502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3198092027785786,
+      "learning_rate": 2.5624913483017246e-06,
+      "loss": 0.618,
+      "step": 10503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7108704039032405,
+      "learning_rate": 2.5615864208966573e-06,
+      "loss": 0.6986,
+      "step": 10504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.691175362216201,
+      "learning_rate": 2.560681598275655e-06,
+      "loss": 0.6876,
+      "step": 10505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7844443924610105,
+      "learning_rate": 2.559776880477596e-06,
+      "loss": 0.7384,
+      "step": 10506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9310484976419107,
+      "learning_rate": 2.55887226754136e-06,
+      "loss": 0.7645,
+      "step": 10507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6672167446108173,
+      "learning_rate": 2.5579677595058196e-06,
+      "loss": 0.6354,
+      "step": 10508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0548314562161867,
+      "learning_rate": 2.55706335640984e-06,
+      "loss": 0.8021,
+      "step": 10509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5042419814098358,
+      "learning_rate": 2.5561590582922934e-06,
+      "loss": 0.6052,
+      "step": 10510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6818314594304793,
+      "learning_rate": 2.5552548651920317e-06,
+      "loss": 0.7064,
+      "step": 10511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8073722620630568,
+      "learning_rate": 2.5543507771479137e-06,
+      "loss": 0.7005,
+      "step": 10512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.049582298842044,
+      "learning_rate": 2.553446794198793e-06,
+      "loss": 0.9208,
+      "step": 10513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7418826065824329,
+      "learning_rate": 2.552542916383507e-06,
+      "loss": 0.6815,
+      "step": 10514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.761180467988135,
+      "learning_rate": 2.551639143740905e-06,
+      "loss": 0.681,
+      "step": 10515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.629209770936907,
+      "learning_rate": 2.5507354763098226e-06,
+      "loss": 0.7159,
+      "step": 10516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5145606937111586,
+      "learning_rate": 2.5498319141290904e-06,
+      "loss": 0.6186,
+      "step": 10517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21015722824601,
+      "learning_rate": 2.548928457237538e-06,
+      "loss": 0.6414,
+      "step": 10518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8088713774633514,
+      "learning_rate": 2.5480251056739874e-06,
+      "loss": 0.7261,
+      "step": 10519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.708442428336887,
+      "learning_rate": 2.5471218594772602e-06,
+      "loss": 0.7644,
+      "step": 10520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1766833697950756,
+      "learning_rate": 2.5462187186861697e-06,
+      "loss": 0.7838,
+      "step": 10521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7186407301242206,
+      "learning_rate": 2.5453156833395254e-06,
+      "loss": 0.7229,
+      "step": 10522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9384132702642054,
+      "learning_rate": 2.544412753476134e-06,
+      "loss": 0.7875,
+      "step": 10523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7509691916894927,
+      "learning_rate": 2.543509929134794e-06,
+      "loss": 0.6859,
+      "step": 10524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5397069694655818,
+      "learning_rate": 2.5426072103543043e-06,
+      "loss": 0.5566,
+      "step": 10525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0872598704143694,
+      "learning_rate": 2.5417045971734556e-06,
+      "loss": 0.7972,
+      "step": 10526
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.703735738213567,
+      "learning_rate": 2.5408020896310333e-06,
+      "loss": 0.7775,
+      "step": 10527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5869172430661111,
+      "learning_rate": 2.5398996877658256e-06,
+      "loss": 0.7128,
+      "step": 10528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.084600236443474,
+      "learning_rate": 2.5389973916166037e-06,
+      "loss": 0.8665,
+      "step": 10529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6098482937997791,
+      "learning_rate": 2.5380952012221473e-06,
+      "loss": 0.6896,
+      "step": 10530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5773585363322016,
+      "learning_rate": 2.5371931166212216e-06,
+      "loss": 0.6354,
+      "step": 10531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9070404815012973,
+      "learning_rate": 2.536291137852594e-06,
+      "loss": 0.7419,
+      "step": 10532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.757528383812611,
+      "learning_rate": 2.535389264955023e-06,
+      "loss": 0.7585,
+      "step": 10533
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5038020371807834,
+      "learning_rate": 2.534487497967262e-06,
+      "loss": 0.6338,
+      "step": 10534
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6300457753136284,
+      "learning_rate": 2.5335858369280674e-06,
+      "loss": 0.6969,
+      "step": 10535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9336369901187365,
+      "learning_rate": 2.5326842818761814e-06,
+      "loss": 0.677,
+      "step": 10536
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.902125872256862,
+      "learning_rate": 2.5317828328503445e-06,
+      "loss": 0.7434,
+      "step": 10537
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7787775840495346,
+      "learning_rate": 2.5308814898892985e-06,
+      "loss": 0.7176,
+      "step": 10538
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7463553598872457,
+      "learning_rate": 2.529980253031774e-06,
+      "loss": 0.639,
+      "step": 10539
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.820054890367816,
+      "learning_rate": 2.529079122316499e-06,
+      "loss": 0.7291,
+      "step": 10540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7362714310742928,
+      "learning_rate": 2.5281780977821967e-06,
+      "loss": 0.6073,
+      "step": 10541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.152771980640751,
+      "learning_rate": 2.5272771794675866e-06,
+      "loss": 0.7302,
+      "step": 10542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5436647357017448,
+      "learning_rate": 2.526376367411384e-06,
+      "loss": 0.8766,
+      "step": 10543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6681064926813716,
+      "learning_rate": 2.5254756616522953e-06,
+      "loss": 0.6637,
+      "step": 10544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7700042495717256,
+      "learning_rate": 2.52457506222903e-06,
+      "loss": 0.7758,
+      "step": 10545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.694031585368257,
+      "learning_rate": 2.52367456918029e-06,
+      "loss": 0.7353,
+      "step": 10546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6762144213983587,
+      "learning_rate": 2.522774182544765e-06,
+      "loss": 0.7833,
+      "step": 10547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0113644504219044,
+      "learning_rate": 2.5218739023611523e-06,
+      "loss": 0.5908,
+      "step": 10548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5008515029406695,
+      "learning_rate": 2.5209737286681367e-06,
+      "loss": 0.7072,
+      "step": 10549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1049765879470523,
+      "learning_rate": 2.5200736615044012e-06,
+      "loss": 0.6885,
+      "step": 10550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8723263234676062,
+      "learning_rate": 2.5191737009086235e-06,
+      "loss": 0.7178,
+      "step": 10551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.347455274998888,
+      "learning_rate": 2.5182738469194757e-06,
+      "loss": 0.7519,
+      "step": 10552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6676340527487266,
+      "learning_rate": 2.5173740995756313e-06,
+      "loss": 0.6544,
+      "step": 10553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2779805709855199,
+      "learning_rate": 2.5164744589157488e-06,
+      "loss": 0.61,
+      "step": 10554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.61174920105065,
+      "learning_rate": 2.5155749249784876e-06,
+      "loss": 0.7278,
+      "step": 10555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6737955860251992,
+      "learning_rate": 2.514675497802508e-06,
+      "loss": 0.8479,
+      "step": 10556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.921924987644472,
+      "learning_rate": 2.513776177426457e-06,
+      "loss": 0.6979,
+      "step": 10557
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.629925609026353,
+      "learning_rate": 2.51287696388898e-06,
+      "loss": 0.7021,
+      "step": 10558
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8001406646782832,
+      "learning_rate": 2.5119778572287195e-06,
+      "loss": 0.8028,
+      "step": 10559
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7928948262419429,
+      "learning_rate": 2.5110788574843113e-06,
+      "loss": 0.7344,
+      "step": 10560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.477538695286704,
+      "learning_rate": 2.510179964694388e-06,
+      "loss": 0.6024,
+      "step": 10561
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7239365190547553,
+      "learning_rate": 2.5092811788975754e-06,
+      "loss": 0.6709,
+      "step": 10562
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7244964055226673,
+      "learning_rate": 2.508382500132499e-06,
+      "loss": 0.6314,
+      "step": 10563
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5816340883185187,
+      "learning_rate": 2.5074839284377774e-06,
+      "loss": 0.8013,
+      "step": 10564
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6421124191514476,
+      "learning_rate": 2.506585463852019e-06,
+      "loss": 0.9379,
+      "step": 10565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6102725760124343,
+      "learning_rate": 2.5056871064138383e-06,
+      "loss": 0.7377,
+      "step": 10566
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3904147095149932,
+      "learning_rate": 2.504788856161838e-06,
+      "loss": 0.8134,
+      "step": 10567
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.198173620667642,
+      "learning_rate": 2.5038907131346168e-06,
+      "loss": 0.7275,
+      "step": 10568
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.218727502905834,
+      "learning_rate": 2.5029926773707713e-06,
+      "loss": 0.5849,
+      "step": 10569
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8778106325221307,
+      "learning_rate": 2.50209474890889e-06,
+      "loss": 0.7266,
+      "step": 10570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6333668186516024,
+      "learning_rate": 2.5011969277875643e-06,
+      "loss": 0.6771,
+      "step": 10571
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.455726557808762,
+      "learning_rate": 2.5002992140453674e-06,
+      "loss": 0.73,
+      "step": 10572
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6816599440207043,
+      "learning_rate": 2.499401607720883e-06,
+      "loss": 0.7155,
+      "step": 10573
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8920297699225725,
+      "learning_rate": 2.49850410885268e-06,
+      "loss": 0.7093,
+      "step": 10574
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5940065646555972,
+      "learning_rate": 2.4976067174793268e-06,
+      "loss": 0.6525,
+      "step": 10575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6403968382883,
+      "learning_rate": 2.496709433639386e-06,
+      "loss": 0.749,
+      "step": 10576
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6944093572444265,
+      "learning_rate": 2.495812257371416e-06,
+      "loss": 0.6166,
+      "step": 10577
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2199868951120774,
+      "learning_rate": 2.4949151887139705e-06,
+      "loss": 0.7563,
+      "step": 10578
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9266035229817122,
+      "learning_rate": 2.4940182277055987e-06,
+      "loss": 0.7155,
+      "step": 10579
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5870030110312,
+      "learning_rate": 2.4931213743848428e-06,
+      "loss": 0.6689,
+      "step": 10580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5334628005053585,
+      "learning_rate": 2.4922246287902458e-06,
+      "loss": 0.7193,
+      "step": 10581
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7112037381952068,
+      "learning_rate": 2.491327990960342e-06,
+      "loss": 0.6416,
+      "step": 10582
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2718239016339272,
+      "learning_rate": 2.4904314609336616e-06,
+      "loss": 0.6466,
+      "step": 10583
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6960580603036075,
+      "learning_rate": 2.4895350387487304e-06,
+      "loss": 0.74,
+      "step": 10584
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.700271544487396,
+      "learning_rate": 2.48863872444407e-06,
+      "loss": 0.727,
+      "step": 10585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.983631413071237,
+      "learning_rate": 2.4877425180581964e-06,
+      "loss": 0.7854,
+      "step": 10586
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.838643164180294,
+      "learning_rate": 2.486846419629622e-06,
+      "loss": 0.6784,
+      "step": 10587
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7204804784928363,
+      "learning_rate": 2.4859504291968522e-06,
+      "loss": 0.6823,
+      "step": 10588
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0033323164357153,
+      "learning_rate": 2.485054546798395e-06,
+      "loss": 0.6103,
+      "step": 10589
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0601152639868934,
+      "learning_rate": 2.4841587724727416e-06,
+      "loss": 0.6887,
+      "step": 10590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6567914792045537,
+      "learning_rate": 2.4832631062583906e-06,
+      "loss": 0.7597,
+      "step": 10591
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3748457185655303,
+      "learning_rate": 2.4823675481938293e-06,
+      "loss": 0.737,
+      "step": 10592
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5641165387250426,
+      "learning_rate": 2.4814720983175413e-06,
+      "loss": 0.7621,
+      "step": 10593
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.218780306928261,
+      "learning_rate": 2.4805767566680057e-06,
+      "loss": 0.7658,
+      "step": 10594
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6299619262571958,
+      "learning_rate": 2.4796815232836965e-06,
+      "loss": 0.7127,
+      "step": 10595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8889531549936374,
+      "learning_rate": 2.478786398203089e-06,
+      "loss": 0.7772,
+      "step": 10596
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.554839622634349,
+      "learning_rate": 2.4778913814646433e-06,
+      "loss": 0.7052,
+      "step": 10597
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.729528744608259,
+      "learning_rate": 2.47699647310682e-06,
+      "loss": 0.8415,
+      "step": 10598
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6533043920557324,
+      "learning_rate": 2.4761016731680792e-06,
+      "loss": 0.7374,
+      "step": 10599
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.852139597788533,
+      "learning_rate": 2.4752069816868707e-06,
+      "loss": 0.598,
+      "step": 10600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9336088074080644,
+      "learning_rate": 2.474312398701641e-06,
+      "loss": 0.7513,
+      "step": 10601
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1154716766665216,
+      "learning_rate": 2.473417924250833e-06,
+      "loss": 0.6251,
+      "step": 10602
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8433546862193424,
+      "learning_rate": 2.472523558372883e-06,
+      "loss": 0.7042,
+      "step": 10603
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.500656569035716,
+      "learning_rate": 2.4716293011062248e-06,
+      "loss": 0.7255,
+      "step": 10604
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0782039104418197,
+      "learning_rate": 2.470735152489287e-06,
+      "loss": 0.6701,
+      "step": 10605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6045341060594935,
+      "learning_rate": 2.4698411125604903e-06,
+      "loss": 0.7218,
+      "step": 10606
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8935689397122388,
+      "learning_rate": 2.46894718135826e-06,
+      "loss": 0.7564,
+      "step": 10607
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8095594584962333,
+      "learning_rate": 2.4680533589210028e-06,
+      "loss": 0.8235,
+      "step": 10608
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0293679977302457,
+      "learning_rate": 2.467159645287133e-06,
+      "loss": 0.6848,
+      "step": 10609
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7024149954838674,
+      "learning_rate": 2.4662660404950544e-06,
+      "loss": 0.7953,
+      "step": 10610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9785485745843823,
+      "learning_rate": 2.465372544583167e-06,
+      "loss": 0.7803,
+      "step": 10611
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0218593689812157,
+      "learning_rate": 2.4644791575898665e-06,
+      "loss": 0.6374,
+      "step": 10612
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0840340091294445,
+      "learning_rate": 2.4635858795535413e-06,
+      "loss": 0.6255,
+      "step": 10613
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6450411984569986,
+      "learning_rate": 2.4626927105125834e-06,
+      "loss": 0.6163,
+      "step": 10614
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6980965252602491,
+      "learning_rate": 2.461799650505369e-06,
+      "loss": 0.766,
+      "step": 10615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6960708715576722,
+      "learning_rate": 2.4609066995702745e-06,
+      "loss": 0.6612,
+      "step": 10616
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7795068981539444,
+      "learning_rate": 2.460013857745676e-06,
+      "loss": 0.7837,
+      "step": 10617
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4475561888398922,
+      "learning_rate": 2.4591211250699393e-06,
+      "loss": 0.7027,
+      "step": 10618
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8561129598609247,
+      "learning_rate": 2.4582285015814263e-06,
+      "loss": 0.7513,
+      "step": 10619
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2827205141165798,
+      "learning_rate": 2.457335987318495e-06,
+      "loss": 0.5528,
+      "step": 10620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6577288996688193,
+      "learning_rate": 2.4564435823194977e-06,
+      "loss": 0.7085,
+      "step": 10621
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5465365405140383,
+      "learning_rate": 2.455551286622788e-06,
+      "loss": 0.548,
+      "step": 10622
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5595684548752489,
+      "learning_rate": 2.454659100266703e-06,
+      "loss": 0.6834,
+      "step": 10623
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0921204092734238,
+      "learning_rate": 2.4537670232895866e-06,
+      "loss": 0.6406,
+      "step": 10624
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6565002905428763,
+      "learning_rate": 2.4528750557297745e-06,
+      "loss": 0.6843,
+      "step": 10625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2379417144042524,
+      "learning_rate": 2.4519831976255892e-06,
+      "loss": 0.7072,
+      "step": 10626
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2731371250942067,
+      "learning_rate": 2.451091449015364e-06,
+      "loss": 0.6446,
+      "step": 10627
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7956659119815377,
+      "learning_rate": 2.4501998099374147e-06,
+      "loss": 0.7462,
+      "step": 10628
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9535516588924444,
+      "learning_rate": 2.4493082804300585e-06,
+      "loss": 0.7938,
+      "step": 10629
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1283595632748067,
+      "learning_rate": 2.4484168605316062e-06,
+      "loss": 0.7026,
+      "step": 10630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6223242230826975,
+      "learning_rate": 2.4475255502803624e-06,
+      "loss": 0.6987,
+      "step": 10631
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8944889191566425,
+      "learning_rate": 2.4466343497146345e-06,
+      "loss": 0.778,
+      "step": 10632
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8037880764169036,
+      "learning_rate": 2.445743258872711e-06,
+      "loss": 0.6594,
+      "step": 10633
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6666083715190771,
+      "learning_rate": 2.4448522777928903e-06,
+      "loss": 0.5805,
+      "step": 10634
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.61852169465093,
+      "learning_rate": 2.4439614065134576e-06,
+      "loss": 0.6511,
+      "step": 10635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.978945830528889,
+      "learning_rate": 2.443070645072696e-06,
+      "loss": 0.7292,
+      "step": 10636
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7546815247775593,
+      "learning_rate": 2.4421799935088837e-06,
+      "loss": 0.8114,
+      "step": 10637
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9079193530730336,
+      "learning_rate": 2.441289451860293e-06,
+      "loss": 0.6524,
+      "step": 10638
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5745835839804052,
+      "learning_rate": 2.4403990201651915e-06,
+      "loss": 0.735,
+      "step": 10639
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6900907708469355,
+      "learning_rate": 2.4395086984618486e-06,
+      "loss": 0.7416,
+      "step": 10640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.734743782955788,
+      "learning_rate": 2.4386184867885154e-06,
+      "loss": 0.7121,
+      "step": 10641
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.210821098089137,
+      "learning_rate": 2.437728385183453e-06,
+      "loss": 0.6085,
+      "step": 10642
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9118941359079593,
+      "learning_rate": 2.436838393684907e-06,
+      "loss": 0.8164,
+      "step": 10643
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.06640149603053,
+      "learning_rate": 2.435948512331125e-06,
+      "loss": 0.8404,
+      "step": 10644
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2244547477811227,
+      "learning_rate": 2.4350587411603454e-06,
+      "loss": 0.6473,
+      "step": 10645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.123567995744661,
+      "learning_rate": 2.434169080210802e-06,
+      "loss": 0.5771,
+      "step": 10646
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6271380017774673,
+      "learning_rate": 2.433279529520732e-06,
+      "loss": 0.7193,
+      "step": 10647
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5795063892246821,
+      "learning_rate": 2.432390089128354e-06,
+      "loss": 0.7118,
+      "step": 10648
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.043020064963528,
+      "learning_rate": 2.4315007590718913e-06,
+      "loss": 0.6058,
+      "step": 10649
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7361983255332676,
+      "learning_rate": 2.4306115393895645e-06,
+      "loss": 0.7015,
+      "step": 10650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7527560597489038,
+      "learning_rate": 2.4297224301195795e-06,
+      "loss": 0.8455,
+      "step": 10651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6937826585993505,
+      "learning_rate": 2.4288334313001467e-06,
+      "loss": 0.7287,
+      "step": 10652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8419275464146092,
+      "learning_rate": 2.4279445429694683e-06,
+      "loss": 0.7917,
+      "step": 10653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.137812439786945,
+      "learning_rate": 2.427055765165741e-06,
+      "loss": 0.6359,
+      "step": 10654
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5699109800603892,
+      "learning_rate": 2.4261670979271577e-06,
+      "loss": 0.7189,
+      "step": 10655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1004844589592573,
+      "learning_rate": 2.4252785412919065e-06,
+      "loss": 0.6005,
+      "step": 10656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.072827018810702,
+      "learning_rate": 2.4243900952981685e-06,
+      "loss": 0.7494,
+      "step": 10657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.700810972880891,
+      "learning_rate": 2.423501759984128e-06,
+      "loss": 0.7725,
+      "step": 10658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.686717168440388,
+      "learning_rate": 2.4226135353879516e-06,
+      "loss": 0.7894,
+      "step": 10659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6393909557137045,
+      "learning_rate": 2.4217254215478137e-06,
+      "loss": 0.6708,
+      "step": 10660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0970011226294671,
+      "learning_rate": 2.420837418501876e-06,
+      "loss": 0.6475,
+      "step": 10661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6943415008315292,
+      "learning_rate": 2.4199495262882993e-06,
+      "loss": 0.6947,
+      "step": 10662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6874430540418923,
+      "learning_rate": 2.419061744945237e-06,
+      "loss": 0.6547,
+      "step": 10663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6225462778342652,
+      "learning_rate": 2.4181740745108377e-06,
+      "loss": 0.6368,
+      "step": 10664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9783305222817045,
+      "learning_rate": 2.4172865150232528e-06,
+      "loss": 0.6365,
+      "step": 10665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1099981190472286,
+      "learning_rate": 2.4163990665206153e-06,
+      "loss": 0.6683,
+      "step": 10666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0331372612570413,
+      "learning_rate": 2.415511729041063e-06,
+      "loss": 0.6195,
+      "step": 10667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6210027740907167,
+      "learning_rate": 2.414624502622731e-06,
+      "loss": 0.714,
+      "step": 10668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5641882398344453,
+      "learning_rate": 2.413737387303739e-06,
+      "loss": 0.657,
+      "step": 10669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4971274710938005,
+      "learning_rate": 2.412850383122213e-06,
+      "loss": 0.7055,
+      "step": 10670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2515291262770805,
+      "learning_rate": 2.411963490116267e-06,
+      "loss": 0.6898,
+      "step": 10671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8392004296032591,
+      "learning_rate": 2.4110767083240145e-06,
+      "loss": 0.7046,
+      "step": 10672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6421864430074096,
+      "learning_rate": 2.410190037783562e-06,
+      "loss": 0.6797,
+      "step": 10673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8585409278287903,
+      "learning_rate": 2.4093034785330087e-06,
+      "loss": 0.6383,
+      "step": 10674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.806372685346883,
+      "learning_rate": 2.408417030610457e-06,
+      "loss": 0.575,
+      "step": 10675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.039400751083578,
+      "learning_rate": 2.4075306940539988e-06,
+      "loss": 0.7471,
+      "step": 10676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.010018813566582,
+      "learning_rate": 2.4066444689017173e-06,
+      "loss": 0.5679,
+      "step": 10677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.774454161239747,
+      "learning_rate": 2.4057583551917e-06,
+      "loss": 0.7764,
+      "step": 10678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8074676677816706,
+      "learning_rate": 2.4048723529620246e-06,
+      "loss": 0.7597,
+      "step": 10679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7158931876370862,
+      "learning_rate": 2.403986462250763e-06,
+      "loss": 0.6172,
+      "step": 10680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6265111147443365,
+      "learning_rate": 2.4031006830959853e-06,
+      "loss": 0.7798,
+      "step": 10681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7953418247104216,
+      "learning_rate": 2.4022150155357526e-06,
+      "loss": 0.6799,
+      "step": 10682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4921104083551096,
+      "learning_rate": 2.40132945960813e-06,
+      "loss": 0.6171,
+      "step": 10683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5889258051474955,
+      "learning_rate": 2.4004440153511642e-06,
+      "loss": 0.7127,
+      "step": 10684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7639832920498395,
+      "learning_rate": 2.399558682802911e-06,
+      "loss": 0.7144,
+      "step": 10685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6454687282855232,
+      "learning_rate": 2.3986734620014146e-06,
+      "loss": 0.7149,
+      "step": 10686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6901160070615373,
+      "learning_rate": 2.3977883529847087e-06,
+      "loss": 0.7652,
+      "step": 10687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6062493484950198,
+      "learning_rate": 2.396903355790835e-06,
+      "loss": 0.7223,
+      "step": 10688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0978200400693368,
+      "learning_rate": 2.396018470457821e-06,
+      "loss": 0.5863,
+      "step": 10689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5193121645562813,
+      "learning_rate": 2.395133697023693e-06,
+      "loss": 0.5962,
+      "step": 10690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7280156715414121,
+      "learning_rate": 2.394249035526472e-06,
+      "loss": 0.8052,
+      "step": 10691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6422483499058715,
+      "learning_rate": 2.3933644860041703e-06,
+      "loss": 0.7525,
+      "step": 10692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.188274727721925,
+      "learning_rate": 2.3924800484948045e-06,
+      "loss": 0.648,
+      "step": 10693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7715454938430748,
+      "learning_rate": 2.3915957230363783e-06,
+      "loss": 0.6432,
+      "step": 10694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.583764611219115,
+      "learning_rate": 2.390711509666893e-06,
+      "loss": 0.7449,
+      "step": 10695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1088439685797828,
+      "learning_rate": 2.389827408424345e-06,
+      "loss": 0.7029,
+      "step": 10696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0507909654670726,
+      "learning_rate": 2.388943419346727e-06,
+      "loss": 0.7304,
+      "step": 10697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7557474050361235,
+      "learning_rate": 2.388059542472025e-06,
+      "loss": 0.7412,
+      "step": 10698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5738379042042008,
+      "learning_rate": 2.3871757778382216e-06,
+      "loss": 0.6807,
+      "step": 10699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9819789237144885,
+      "learning_rate": 2.386292125483292e-06,
+      "loss": 0.6715,
+      "step": 10700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.111401454546964,
+      "learning_rate": 2.385408585445215e-06,
+      "loss": 0.6752,
+      "step": 10701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8092032436943637,
+      "learning_rate": 2.38452515776195e-06,
+      "loss": 0.6657,
+      "step": 10702
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.665096397623257,
+      "learning_rate": 2.3836418424714665e-06,
+      "loss": 0.7929,
+      "step": 10703
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8124376702941174,
+      "learning_rate": 2.3827586396117207e-06,
+      "loss": 0.6866,
+      "step": 10704
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7452369028674888,
+      "learning_rate": 2.3818755492206624e-06,
+      "loss": 0.7943,
+      "step": 10705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5330841018586547,
+      "learning_rate": 2.3809925713362437e-06,
+      "loss": 0.6691,
+      "step": 10706
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.784817003410383,
+      "learning_rate": 2.3801097059964075e-06,
+      "loss": 0.6389,
+      "step": 10707
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7997585001710033,
+      "learning_rate": 2.3792269532390923e-06,
+      "loss": 0.7352,
+      "step": 10708
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.66013629300847,
+      "learning_rate": 2.378344313102231e-06,
+      "loss": 0.5853,
+      "step": 10709
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2763327373614386,
+      "learning_rate": 2.377461785623752e-06,
+      "loss": 0.6887,
+      "step": 10710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9731377910709078,
+      "learning_rate": 2.3765793708415825e-06,
+      "loss": 0.6729,
+      "step": 10711
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.17629765436105,
+      "learning_rate": 2.3756970687936397e-06,
+      "loss": 0.6882,
+      "step": 10712
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4190954564929414,
+      "learning_rate": 2.3748148795178387e-06,
+      "loss": 0.6895,
+      "step": 10713
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7356184381876214,
+      "learning_rate": 2.373932803052089e-06,
+      "loss": 0.792,
+      "step": 10714
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.630977426810605,
+      "learning_rate": 2.3730508394342957e-06,
+      "loss": 0.6725,
+      "step": 10715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9085017355394898,
+      "learning_rate": 2.372168988702358e-06,
+      "loss": 0.6357,
+      "step": 10716
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7319208033722784,
+      "learning_rate": 2.3712872508941714e-06,
+      "loss": 0.7647,
+      "step": 10717
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.613035566548936,
+      "learning_rate": 2.3704056260476237e-06,
+      "loss": 0.7511,
+      "step": 10718
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0221659932973828,
+      "learning_rate": 2.369524114200607e-06,
+      "loss": 0.5753,
+      "step": 10719
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5696670263070787,
+      "learning_rate": 2.368642715390994e-06,
+      "loss": 0.6543,
+      "step": 10720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6051283991463394,
+      "learning_rate": 2.3677614296566654e-06,
+      "loss": 0.7179,
+      "step": 10721
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.232550333643952,
+      "learning_rate": 2.3668802570354903e-06,
+      "loss": 0.8189,
+      "step": 10722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5710322955357805,
+      "learning_rate": 2.365999197565335e-06,
+      "loss": 0.6351,
+      "step": 10723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5965852661332487,
+      "learning_rate": 2.3651182512840604e-06,
+      "loss": 0.6876,
+      "step": 10724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.53479235237334,
+      "learning_rate": 2.3642374182295207e-06,
+      "loss": 0.6834,
+      "step": 10725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4294072298298992,
+      "learning_rate": 2.363356698439573e-06,
+      "loss": 0.7085,
+      "step": 10726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.62655664025636,
+      "learning_rate": 2.362476091952059e-06,
+      "loss": 0.7819,
+      "step": 10727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6735706207020715,
+      "learning_rate": 2.361595598804819e-06,
+      "loss": 0.7432,
+      "step": 10728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8437453137411053,
+      "learning_rate": 2.360715219035694e-06,
+      "loss": 0.8337,
+      "step": 10729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.80330613022894,
+      "learning_rate": 2.3598349526825137e-06,
+      "loss": 0.8547,
+      "step": 10730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1435349500936416,
+      "learning_rate": 2.358954799783106e-06,
+      "loss": 0.6094,
+      "step": 10731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7203293157023745,
+      "learning_rate": 2.358074760375292e-06,
+      "loss": 0.8066,
+      "step": 10732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1032645243304782,
+      "learning_rate": 2.3571948344968896e-06,
+      "loss": 0.6503,
+      "step": 10733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6731596439058765,
+      "learning_rate": 2.356315022185712e-06,
+      "loss": 0.6937,
+      "step": 10734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0150236619802742,
+      "learning_rate": 2.3554353234795624e-06,
+      "loss": 0.7853,
+      "step": 10735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.754878012188957,
+      "learning_rate": 2.3545557384162494e-06,
+      "loss": 0.7317,
+      "step": 10736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9695581331993304,
+      "learning_rate": 2.3536762670335696e-06,
+      "loss": 0.7514,
+      "step": 10737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6739786025485257,
+      "learning_rate": 2.3527969093693105e-06,
+      "loss": 0.6197,
+      "step": 10738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7660566355812128,
+      "learning_rate": 2.3519176654612657e-06,
+      "loss": 0.714,
+      "step": 10739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7232359730059035,
+      "learning_rate": 2.3510385353472163e-06,
+      "loss": 0.6519,
+      "step": 10740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5956014597596082,
+      "learning_rate": 2.3501595190649405e-06,
+      "loss": 0.7241,
+      "step": 10741
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2030018514721792,
+      "learning_rate": 2.3492806166522125e-06,
+      "loss": 0.6195,
+      "step": 10742
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5531121049930008,
+      "learning_rate": 2.3484018281467973e-06,
+      "loss": 0.6755,
+      "step": 10743
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3053193069337277,
+      "learning_rate": 2.3475231535864653e-06,
+      "loss": 0.7036,
+      "step": 10744
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0874999785364636,
+      "learning_rate": 2.346644593008966e-06,
+      "loss": 0.5083,
+      "step": 10745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6848512838030414,
+      "learning_rate": 2.3457661464520605e-06,
+      "loss": 0.8349,
+      "step": 10746
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5532564235701547,
+      "learning_rate": 2.3448878139534948e-06,
+      "loss": 0.7078,
+      "step": 10747
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.152815002009677,
+      "learning_rate": 2.344009595551013e-06,
+      "loss": 0.6276,
+      "step": 10748
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9105902499861813,
+      "learning_rate": 2.3431314912823543e-06,
+      "loss": 0.785,
+      "step": 10749
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7210805265825289,
+      "learning_rate": 2.3422535011852522e-06,
+      "loss": 0.6687,
+      "step": 10750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7733258036213746,
+      "learning_rate": 2.3413756252974366e-06,
+      "loss": 0.706,
+      "step": 10751
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2516409122793022,
+      "learning_rate": 2.3404978636566312e-06,
+      "loss": 0.7187,
+      "step": 10752
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0311984171086093,
+      "learning_rate": 2.3396202163005536e-06,
+      "loss": 0.6528,
+      "step": 10753
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9063244338311807,
+      "learning_rate": 2.338742683266923e-06,
+      "loss": 0.788,
+      "step": 10754
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7915892481928049,
+      "learning_rate": 2.3378652645934473e-06,
+      "loss": 0.7202,
+      "step": 10755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8266279697208871,
+      "learning_rate": 2.3369879603178264e-06,
+      "loss": 0.5982,
+      "step": 10756
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8743480386960922,
+      "learning_rate": 2.336110770477766e-06,
+      "loss": 0.7754,
+      "step": 10757
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5024670045566326,
+      "learning_rate": 2.3352336951109584e-06,
+      "loss": 0.8069,
+      "step": 10758
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8510370063200547,
+      "learning_rate": 2.3343567342550933e-06,
+      "loss": 0.7665,
+      "step": 10759
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7857493334964303,
+      "learning_rate": 2.3334798879478563e-06,
+      "loss": 0.6715,
+      "step": 10760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6328160352605068,
+      "learning_rate": 2.332603156226926e-06,
+      "loss": 0.6507,
+      "step": 10761
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.804126954005363,
+      "learning_rate": 2.3317265391299813e-06,
+      "loss": 0.7299,
+      "step": 10762
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5570352422744702,
+      "learning_rate": 2.330850036694687e-06,
+      "loss": 0.6744,
+      "step": 10763
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7468168559969073,
+      "learning_rate": 2.329973648958712e-06,
+      "loss": 0.7641,
+      "step": 10764
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7920149993016432,
+      "learning_rate": 2.3290973759597165e-06,
+      "loss": 0.7449,
+      "step": 10765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.836392293617268,
+      "learning_rate": 2.328221217735355e-06,
+      "loss": 0.6865,
+      "step": 10766
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9691779387469392,
+      "learning_rate": 2.3273451743232777e-06,
+      "loss": 0.7434,
+      "step": 10767
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6613221407262435,
+      "learning_rate": 2.326469245761131e-06,
+      "loss": 0.6655,
+      "step": 10768
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3729057289722,
+      "learning_rate": 2.3255934320865555e-06,
+      "loss": 0.5828,
+      "step": 10769
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6890272436900213,
+      "learning_rate": 2.324717733337186e-06,
+      "loss": 0.7031,
+      "step": 10770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7860913155304814,
+      "learning_rate": 2.3238421495506515e-06,
+      "loss": 0.6847,
+      "step": 10771
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6851891657504199,
+      "learning_rate": 2.322966680764582e-06,
+      "loss": 0.6565,
+      "step": 10772
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.87902805617575,
+      "learning_rate": 2.322091327016597e-06,
+      "loss": 0.6283,
+      "step": 10773
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6534331416393322,
+      "learning_rate": 2.3212160883443107e-06,
+      "loss": 0.7286,
+      "step": 10774
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0097014626057152,
+      "learning_rate": 2.320340964785336e-06,
+      "loss": 0.6401,
+      "step": 10775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7613601783298174,
+      "learning_rate": 2.3194659563772754e-06,
+      "loss": 0.7091,
+      "step": 10776
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.727765825749901,
+      "learning_rate": 2.318591063157736e-06,
+      "loss": 0.7444,
+      "step": 10777
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6510510444731277,
+      "learning_rate": 2.317716285164309e-06,
+      "loss": 0.7583,
+      "step": 10778
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8252536675351836,
+      "learning_rate": 2.316841622434586e-06,
+      "loss": 0.7503,
+      "step": 10779
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4070712392326299,
+      "learning_rate": 2.3159670750061563e-06,
+      "loss": 0.5941,
+      "step": 10780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.741747352005612,
+      "learning_rate": 2.3150926429165967e-06,
+      "loss": 0.674,
+      "step": 10781
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.220570609047074,
+      "learning_rate": 2.3142183262034865e-06,
+      "loss": 0.538,
+      "step": 10782
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2390980174478317,
+      "learning_rate": 2.3133441249043973e-06,
+      "loss": 0.6868,
+      "step": 10783
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.65350495125348,
+      "learning_rate": 2.3124700390568945e-06,
+      "loss": 0.7583,
+      "step": 10784
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5622719055475534,
+      "learning_rate": 2.311596068698539e-06,
+      "loss": 0.6835,
+      "step": 10785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4529100221323858,
+      "learning_rate": 2.3107222138668863e-06,
+      "loss": 0.7787,
+      "step": 10786
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0042963863123568,
+      "learning_rate": 2.3098484745994933e-06,
+      "loss": 0.6918,
+      "step": 10787
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.559998411900406,
+      "learning_rate": 2.3089748509339007e-06,
+      "loss": 0.8365,
+      "step": 10788
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.526295822279976,
+      "learning_rate": 2.30810134290765e-06,
+      "loss": 0.7851,
+      "step": 10789
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5437169324925868,
+      "learning_rate": 2.307227950558281e-06,
+      "loss": 0.6944,
+      "step": 10790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9801035669759766,
+      "learning_rate": 2.306354673923325e-06,
+      "loss": 0.565,
+      "step": 10791
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0393520774579,
+      "learning_rate": 2.3054815130403073e-06,
+      "loss": 0.7455,
+      "step": 10792
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.728256899910923,
+      "learning_rate": 2.3046084679467506e-06,
+      "loss": 0.7504,
+      "step": 10793
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7415149326689554,
+      "learning_rate": 2.3037355386801683e-06,
+      "loss": 0.6766,
+      "step": 10794
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.100539711889053,
+      "learning_rate": 2.3028627252780794e-06,
+      "loss": 0.8094,
+      "step": 10795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5774928281306821,
+      "learning_rate": 2.3019900277779816e-06,
+      "loss": 0.6556,
+      "step": 10796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8757002061702206,
+      "learning_rate": 2.3011174462173836e-06,
+      "loss": 0.7399,
+      "step": 10797
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.288272951575751,
+      "learning_rate": 2.3002449806337807e-06,
+      "loss": 0.646,
+      "step": 10798
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.802879089839074,
+      "learning_rate": 2.2993726310646603e-06,
+      "loss": 0.5553,
+      "step": 10799
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6198801741071946,
+      "learning_rate": 2.2985003975475146e-06,
+      "loss": 0.6585,
+      "step": 10800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6358748095655227,
+      "learning_rate": 2.2976282801198237e-06,
+      "loss": 0.6827,
+      "step": 10801
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7494034129948848,
+      "learning_rate": 2.2967562788190633e-06,
+      "loss": 0.8395,
+      "step": 10802
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6456164872794434,
+      "learning_rate": 2.2958843936827064e-06,
+      "loss": 0.6834,
+      "step": 10803
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8844434099779603,
+      "learning_rate": 2.2950126247482178e-06,
+      "loss": 0.7468,
+      "step": 10804
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.603514599595401,
+      "learning_rate": 2.2941409720530642e-06,
+      "loss": 0.7402,
+      "step": 10805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1499152308500156,
+      "learning_rate": 2.2932694356346986e-06,
+      "loss": 0.6311,
+      "step": 10806
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7321887235869011,
+      "learning_rate": 2.292398015530571e-06,
+      "loss": 0.6566,
+      "step": 10807
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4756829187618514,
+      "learning_rate": 2.2915267117781328e-06,
+      "loss": 0.7123,
+      "step": 10808
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8642357869219564,
+      "learning_rate": 2.2906555244148233e-06,
+      "loss": 0.8017,
+      "step": 10809
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8822613950396763,
+      "learning_rate": 2.2897844534780805e-06,
+      "loss": 0.8079,
+      "step": 10810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7363635856019497,
+      "learning_rate": 2.2889134990053356e-06,
+      "loss": 0.6791,
+      "step": 10811
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1685411710793792,
+      "learning_rate": 2.2880426610340135e-06,
+      "loss": 0.6972,
+      "step": 10812
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8636262703049782,
+      "learning_rate": 2.2871719396015423e-06,
+      "loss": 0.7505,
+      "step": 10813
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4630000481597512,
+      "learning_rate": 2.2863013347453305e-06,
+      "loss": 0.6922,
+      "step": 10814
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.246980673019529,
+      "learning_rate": 2.2854308465027963e-06,
+      "loss": 0.5657,
+      "step": 10815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5620296944096947,
+      "learning_rate": 2.2845604749113457e-06,
+      "loss": 0.6671,
+      "step": 10816
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5672813505606782,
+      "learning_rate": 2.2836902200083762e-06,
+      "loss": 0.5558,
+      "step": 10817
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6584837070914114,
+      "learning_rate": 2.2828200818312886e-06,
+      "loss": 0.7807,
+      "step": 10818
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4022389782339526,
+      "learning_rate": 2.2819500604174733e-06,
+      "loss": 0.6534,
+      "step": 10819
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9285182983449882,
+      "learning_rate": 2.2810801558043177e-06,
+      "loss": 0.6888,
+      "step": 10820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6849100747122707,
+      "learning_rate": 2.2802103680292037e-06,
+      "loss": 0.7063,
+      "step": 10821
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8004334870109053,
+      "learning_rate": 2.279340697129505e-06,
+      "loss": 0.7011,
+      "step": 10822
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6401399072100662,
+      "learning_rate": 2.2784711431426e-06,
+      "loss": 0.7145,
+      "step": 10823
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7991386066424044,
+      "learning_rate": 2.277601706105847e-06,
+      "loss": 0.6797,
+      "step": 10824
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6218288713248405,
+      "learning_rate": 2.2767323860566137e-06,
+      "loss": 0.7207,
+      "step": 10825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5434988002577197,
+      "learning_rate": 2.2758631830322546e-06,
+      "loss": 0.6249,
+      "step": 10826
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6653324396796858,
+      "learning_rate": 2.274994097070122e-06,
+      "loss": 0.6377,
+      "step": 10827
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.686627276131989,
+      "learning_rate": 2.2741251282075617e-06,
+      "loss": 0.6863,
+      "step": 10828
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6788159617521423,
+      "learning_rate": 2.2732562764819157e-06,
+      "loss": 0.7209,
+      "step": 10829
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5979206062911346,
+      "learning_rate": 2.272387541930518e-06,
+      "loss": 0.7589,
+      "step": 10830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6830078335225507,
+      "learning_rate": 2.2715189245907066e-06,
+      "loss": 0.597,
+      "step": 10831
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5916916990177676,
+      "learning_rate": 2.2706504244997997e-06,
+      "loss": 0.6611,
+      "step": 10832
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7569327512305084,
+      "learning_rate": 2.269782041695124e-06,
+      "loss": 0.7056,
+      "step": 10833
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7422975938578633,
+      "learning_rate": 2.2689137762139952e-06,
+      "loss": 0.6919,
+      "step": 10834
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6570358295219136,
+      "learning_rate": 2.2680456280937234e-06,
+      "loss": 0.7807,
+      "step": 10835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6896080260099917,
+      "learning_rate": 2.267177597371616e-06,
+      "loss": 0.7592,
+      "step": 10836
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6300449311400005,
+      "learning_rate": 2.266309684084971e-06,
+      "loss": 0.7674,
+      "step": 10837
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.438979893953582,
+      "learning_rate": 2.265441888271091e-06,
+      "loss": 0.7382,
+      "step": 10838
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0333494589405428,
+      "learning_rate": 2.264574209967262e-06,
+      "loss": 0.7668,
+      "step": 10839
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7825156624566307,
+      "learning_rate": 2.263706649210769e-06,
+      "loss": 0.6629,
+      "step": 10840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8629312951497374,
+      "learning_rate": 2.2628392060388986e-06,
+      "loss": 0.8159,
+      "step": 10841
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7708346407219888,
+      "learning_rate": 2.2619718804889203e-06,
+      "loss": 0.6445,
+      "step": 10842
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6549659897530637,
+      "learning_rate": 2.26110467259811e-06,
+      "loss": 0.6321,
+      "step": 10843
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7250807253982194,
+      "learning_rate": 2.260237582403732e-06,
+      "loss": 0.7128,
+      "step": 10844
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.72955097813344,
+      "learning_rate": 2.259370609943046e-06,
+      "loss": 0.6929,
+      "step": 10845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5945183317141356,
+      "learning_rate": 2.2585037552533097e-06,
+      "loss": 0.5409,
+      "step": 10846
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5132341909846403,
+      "learning_rate": 2.2576370183717704e-06,
+      "loss": 0.672,
+      "step": 10847
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1400869550085753,
+      "learning_rate": 2.2567703993356775e-06,
+      "loss": 0.579,
+      "step": 10848
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4659794614162525,
+      "learning_rate": 2.2559038981822724e-06,
+      "loss": 0.6688,
+      "step": 10849
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9985440981305742,
+      "learning_rate": 2.255037514948785e-06,
+      "loss": 0.7462,
+      "step": 10850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8488698070046936,
+      "learning_rate": 2.2541712496724495e-06,
+      "loss": 0.7854,
+      "step": 10851
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7241919436092699,
+      "learning_rate": 2.2533051023904913e-06,
+      "loss": 0.6382,
+      "step": 10852
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.616116137302536,
+      "learning_rate": 2.25243907314013e-06,
+      "loss": 0.7246,
+      "step": 10853
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7017204410787108,
+      "learning_rate": 2.2515731619585814e-06,
+      "loss": 0.6781,
+      "step": 10854
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9411795250865245,
+      "learning_rate": 2.250707368883052e-06,
+      "loss": 0.787,
+      "step": 10855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4567998035778889,
+      "learning_rate": 2.249841693950755e-06,
+      "loss": 0.7808,
+      "step": 10856
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.390620795945843,
+      "learning_rate": 2.2489761371988826e-06,
+      "loss": 0.671,
+      "step": 10857
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5642667857897166,
+      "learning_rate": 2.2481106986646314e-06,
+      "loss": 0.6471,
+      "step": 10858
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7534131209449295,
+      "learning_rate": 2.247245378385195e-06,
+      "loss": 0.6479,
+      "step": 10859
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6756995433949684,
+      "learning_rate": 2.246380176397752e-06,
+      "loss": 0.7305,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0053528858727394,
+      "learning_rate": 2.245515092739488e-06,
+      "loss": 0.7854,
+      "step": 10861
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0196665222267847,
+      "learning_rate": 2.2446501274475747e-06,
+      "loss": 0.6135,
+      "step": 10862
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5991633655920516,
+      "learning_rate": 2.2437852805591815e-06,
+      "loss": 0.7049,
+      "step": 10863
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.849001588679011,
+      "learning_rate": 2.242920552111473e-06,
+      "loss": 0.7975,
+      "step": 10864
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9610046450592853,
+      "learning_rate": 2.2420559421416075e-06,
+      "loss": 0.6535,
+      "step": 10865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8008598441969794,
+      "learning_rate": 2.241191450686743e-06,
+      "loss": 0.7744,
+      "step": 10866
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8447274956985995,
+      "learning_rate": 2.2403270777840262e-06,
+      "loss": 0.6826,
+      "step": 10867
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.622134292433357,
+      "learning_rate": 2.239462823470598e-06,
+      "loss": 0.7312,
+      "step": 10868
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6051752282321383,
+      "learning_rate": 2.238598687783603e-06,
+      "loss": 0.6799,
+      "step": 10869
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.650472743308951,
+      "learning_rate": 2.237734670760172e-06,
+      "loss": 0.7295,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.659667254369723,
+      "learning_rate": 2.236870772437433e-06,
+      "loss": 0.7445,
+      "step": 10871
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7579116594644997,
+      "learning_rate": 2.2360069928525118e-06,
+      "loss": 0.8021,
+      "step": 10872
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7183990679831456,
+      "learning_rate": 2.2351433320425237e-06,
+      "loss": 0.7741,
+      "step": 10873
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7303690883995355,
+      "learning_rate": 2.234279790044588e-06,
+      "loss": 0.7681,
+      "step": 10874
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8623689282417317,
+      "learning_rate": 2.2334163668958055e-06,
+      "loss": 0.709,
+      "step": 10875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.053895438639393,
+      "learning_rate": 2.2325530626332852e-06,
+      "loss": 0.7205,
+      "step": 10876
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8370594412179477,
+      "learning_rate": 2.231689877294123e-06,
+      "loss": 0.6361,
+      "step": 10877
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7623981129040969,
+      "learning_rate": 2.2308268109154126e-06,
+      "loss": 0.8131,
+      "step": 10878
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.79650249028902,
+      "learning_rate": 2.229963863534241e-06,
+      "loss": 0.7672,
+      "step": 10879
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4334106116439695,
+      "learning_rate": 2.229101035187692e-06,
+      "loss": 0.6543,
+      "step": 10880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.85230574458866,
+      "learning_rate": 2.2282383259128416e-06,
+      "loss": 0.6969,
+      "step": 10881
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6166436412146659,
+      "learning_rate": 2.227375735746764e-06,
+      "loss": 0.7958,
+      "step": 10882
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6944080395738306,
+      "learning_rate": 2.2265132647265243e-06,
+      "loss": 0.7444,
+      "step": 10883
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3349381179288986,
+      "learning_rate": 2.225650912889188e-06,
+      "loss": 0.6575,
+      "step": 10884
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5575577575293058,
+      "learning_rate": 2.224788680271811e-06,
+      "loss": 0.744,
+      "step": 10885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.641401099806086,
+      "learning_rate": 2.223926566911445e-06,
+      "loss": 0.6948,
+      "step": 10886
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9453153259768134,
+      "learning_rate": 2.223064572845137e-06,
+      "loss": 0.8595,
+      "step": 10887
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1387011736757797,
+      "learning_rate": 2.222202698109929e-06,
+      "loss": 0.7956,
+      "step": 10888
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.445861948963439,
+      "learning_rate": 2.221340942742858e-06,
+      "loss": 0.5981,
+      "step": 10889
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5632003737772122,
+      "learning_rate": 2.2204793067809545e-06,
+      "loss": 0.7661,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0836010520743784,
+      "learning_rate": 2.219617790261244e-06,
+      "loss": 0.6061,
+      "step": 10891
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7337408341071945,
+      "learning_rate": 2.218756393220753e-06,
+      "loss": 0.7362,
+      "step": 10892
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.719205165501837,
+      "learning_rate": 2.2178951156964895e-06,
+      "loss": 0.7097,
+      "step": 10893
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.799621061261963,
+      "learning_rate": 2.2170339577254714e-06,
+      "loss": 0.5969,
+      "step": 10894
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.546789178618193,
+      "learning_rate": 2.2161729193447024e-06,
+      "loss": 0.6694,
+      "step": 10895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.069401115816898,
+      "learning_rate": 2.2153120005911825e-06,
+      "loss": 0.8285,
+      "step": 10896
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.443866145969787,
+      "learning_rate": 2.214451201501908e-06,
+      "loss": 0.7222,
+      "step": 10897
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5590071102945744,
+      "learning_rate": 2.213590522113867e-06,
+      "loss": 0.6664,
+      "step": 10898
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5792993893384817,
+      "learning_rate": 2.212729962464051e-06,
+      "loss": 0.5835,
+      "step": 10899
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3568027485057104,
+      "learning_rate": 2.2118695225894343e-06,
+      "loss": 0.5989,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.631303324226593,
+      "learning_rate": 2.2110092025269915e-06,
+      "loss": 0.6672,
+      "step": 10901
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0663468552506359,
+      "learning_rate": 2.210149002313697e-06,
+      "loss": 0.6366,
+      "step": 10902
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0741767095531096,
+      "learning_rate": 2.2092889219865127e-06,
+      "loss": 0.7904,
+      "step": 10903
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.809614576868289,
+      "learning_rate": 2.208428961582399e-06,
+      "loss": 0.8374,
+      "step": 10904
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.531154913281785,
+      "learning_rate": 2.2075691211383094e-06,
+      "loss": 0.7117,
+      "step": 10905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6459168909984379,
+      "learning_rate": 2.2067094006911943e-06,
+      "loss": 0.7588,
+      "step": 10906
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6875428363012672,
+      "learning_rate": 2.205849800277997e-06,
+      "loss": 0.8664,
+      "step": 10907
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4398916232227212,
+      "learning_rate": 2.204990319935657e-06,
+      "loss": 0.6911,
+      "step": 10908
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6621546214043217,
+      "learning_rate": 2.2041309597011057e-06,
+      "loss": 0.6734,
+      "step": 10909
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6557300568169904,
+      "learning_rate": 2.2032717196112776e-06,
+      "loss": 0.7208,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3596009979632733,
+      "learning_rate": 2.2024125997030885e-06,
+      "loss": 0.726,
+      "step": 10911
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7281463663471774,
+      "learning_rate": 2.2015536000134624e-06,
+      "loss": 0.7238,
+      "step": 10912
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5258694365759162,
+      "learning_rate": 2.2006947205793107e-06,
+      "loss": 0.7595,
+      "step": 10913
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5917074375036067,
+      "learning_rate": 2.1998359614375412e-06,
+      "loss": 0.6952,
+      "step": 10914
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.575070791581636,
+      "learning_rate": 2.198977322625057e-06,
+      "loss": 0.6279,
+      "step": 10915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5413480929341974,
+      "learning_rate": 2.1981188041787534e-06,
+      "loss": 0.6615,
+      "step": 10916
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5531660837500603,
+      "learning_rate": 2.1972604061355285e-06,
+      "loss": 0.6019,
+      "step": 10917
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7899324558592955,
+      "learning_rate": 2.196402128532264e-06,
+      "loss": 0.6291,
+      "step": 10918
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0185552115249776,
+      "learning_rate": 2.1955439714058422e-06,
+      "loss": 0.9019,
+      "step": 10919
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.299986849031111,
+      "learning_rate": 2.1946859347931442e-06,
+      "loss": 0.7152,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9200178232070786,
+      "learning_rate": 2.193828018731039e-06,
+      "loss": 0.7028,
+      "step": 10921
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7068634092631465,
+      "learning_rate": 2.192970223256393e-06,
+      "loss": 0.7585,
+      "step": 10922
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8617245091208035,
+      "learning_rate": 2.1921125484060692e-06,
+      "loss": 0.8036,
+      "step": 10923
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8929519129783416,
+      "learning_rate": 2.191254994216922e-06,
+      "loss": 0.6801,
+      "step": 10924
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.12433929228347,
+      "learning_rate": 2.1903975607258025e-06,
+      "loss": 0.7875,
+      "step": 10925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6727812547883771,
+      "learning_rate": 2.1895402479695566e-06,
+      "loss": 0.7302,
+      "step": 10926
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0218862050008068,
+      "learning_rate": 2.1886830559850264e-06,
+      "loss": 0.6765,
+      "step": 10927
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.329326880883483,
+      "learning_rate": 2.187825984809049e-06,
+      "loss": 0.6387,
+      "step": 10928
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3993522704459391,
+      "learning_rate": 2.186969034478448e-06,
+      "loss": 0.6626,
+      "step": 10929
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6562084066768215,
+      "learning_rate": 2.186112205030054e-06,
+      "loss": 0.6758,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7617086037119556,
+      "learning_rate": 2.1852554965006854e-06,
+      "loss": 0.6816,
+      "step": 10931
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9781588802831636,
+      "learning_rate": 2.1843989089271572e-06,
+      "loss": 0.6055,
+      "step": 10932
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7644987107065064,
+      "learning_rate": 2.183542442346279e-06,
+      "loss": 0.7131,
+      "step": 10933
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.454530310719767,
+      "learning_rate": 2.182686096794852e-06,
+      "loss": 0.6002,
+      "step": 10934
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6408718155133322,
+      "learning_rate": 2.181829872309682e-06,
+      "loss": 0.6914,
+      "step": 10935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6957203312789986,
+      "learning_rate": 2.180973768927555e-06,
+      "loss": 0.7271,
+      "step": 10936
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7465005445064787,
+      "learning_rate": 2.1801177866852658e-06,
+      "loss": 0.7326,
+      "step": 10937
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8312254571434192,
+      "learning_rate": 2.179261925619595e-06,
+      "loss": 0.6893,
+      "step": 10938
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1975189208316481,
+      "learning_rate": 2.1784061857673217e-06,
+      "loss": 0.6424,
+      "step": 10939
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1567564282806404,
+      "learning_rate": 2.177550567165218e-06,
+      "loss": 0.7142,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6624541289532557,
+      "learning_rate": 2.176695069850053e-06,
+      "loss": 0.6594,
+      "step": 10941
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5693330388180948,
+      "learning_rate": 2.1758396938585885e-06,
+      "loss": 0.6486,
+      "step": 10942
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7745652270227936,
+      "learning_rate": 2.174984439227582e-06,
+      "loss": 0.7449,
+      "step": 10943
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6953616886773626,
+      "learning_rate": 2.174129305993784e-06,
+      "loss": 0.8274,
+      "step": 10944
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7129823223503433,
+      "learning_rate": 2.173274294193945e-06,
+      "loss": 0.8337,
+      "step": 10945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5185827254127737,
+      "learning_rate": 2.1724194038648057e-06,
+      "loss": 0.6464,
+      "step": 10946
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7732686291552804,
+      "learning_rate": 2.1715646350431015e-06,
+      "loss": 0.7528,
+      "step": 10947
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3650389708509505,
+      "learning_rate": 2.1707099877655634e-06,
+      "loss": 0.7456,
+      "step": 10948
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1271996084120057,
+      "learning_rate": 2.1698554620689178e-06,
+      "loss": 0.6205,
+      "step": 10949
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7633031729135424,
+      "learning_rate": 2.169001057989889e-06,
+      "loss": 0.8002,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7789303842662407,
+      "learning_rate": 2.1681467755651882e-06,
+      "loss": 0.6705,
+      "step": 10951
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.579052126002995,
+      "learning_rate": 2.1672926148315247e-06,
+      "loss": 0.7173,
+      "step": 10952
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3900719799619903,
+      "learning_rate": 2.1664385758256106e-06,
+      "loss": 0.6312,
+      "step": 10953
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5769426311898078,
+      "learning_rate": 2.165584658584138e-06,
+      "loss": 0.6468,
+      "step": 10954
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8027918333196824,
+      "learning_rate": 2.1647308631438068e-06,
+      "loss": 0.7766,
+      "step": 10955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6811616147144774,
+      "learning_rate": 2.163877189541305e-06,
+      "loss": 0.7029,
+      "step": 10956
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0333232672517862,
+      "learning_rate": 2.1630236378133168e-06,
+      "loss": 0.75,
+      "step": 10957
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.613602787511423,
+      "learning_rate": 2.1621702079965214e-06,
+      "loss": 0.7484,
+      "step": 10958
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4552719991683194,
+      "learning_rate": 2.16131690012759e-06,
+      "loss": 0.6643,
+      "step": 10959
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7423120066600886,
+      "learning_rate": 2.1604637142431978e-06,
+      "loss": 0.6708,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9936632773420215,
+      "learning_rate": 2.1596106503800013e-06,
+      "loss": 0.7123,
+      "step": 10961
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.599686959480425,
+      "learning_rate": 2.1587577085746596e-06,
+      "loss": 0.7361,
+      "step": 10962
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.608407952885165,
+      "learning_rate": 2.157904888863829e-06,
+      "loss": 0.7344,
+      "step": 10963
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8296909540482187,
+      "learning_rate": 2.157052191284154e-06,
+      "loss": 0.7814,
+      "step": 10964
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5745101650809028,
+      "learning_rate": 2.1561996158722785e-06,
+      "loss": 0.6762,
+      "step": 10965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.069619154685728,
+      "learning_rate": 2.1553471626648392e-06,
+      "loss": 0.6823,
+      "step": 10966
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6772091700254264,
+      "learning_rate": 2.1544948316984658e-06,
+      "loss": 0.7992,
+      "step": 10967
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5244415868950925,
+      "learning_rate": 2.1536426230097906e-06,
+      "loss": 0.7183,
+      "step": 10968
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.583250773378166,
+      "learning_rate": 2.1527905366354292e-06,
+      "loss": 0.8703,
+      "step": 10969
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5786928740201431,
+      "learning_rate": 2.1519385726119972e-06,
+      "loss": 0.7059,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0564557759089024,
+      "learning_rate": 2.151086730976112e-06,
+      "loss": 0.6946,
+      "step": 10971
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6485380107663594,
+      "learning_rate": 2.1502350117643704e-06,
+      "loss": 0.6121,
+      "step": 10972
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.239561994908667,
+      "learning_rate": 2.1493834150133795e-06,
+      "loss": 0.64,
+      "step": 10973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6746388438363164,
+      "learning_rate": 2.1485319407597315e-06,
+      "loss": 0.7024,
+      "step": 10974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.744785130762634,
+      "learning_rate": 2.1476805890400172e-06,
+      "loss": 0.7438,
+      "step": 10975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359472211838223,
+      "learning_rate": 2.14682935989082e-06,
+      "loss": 0.7368,
+      "step": 10976
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.647039179495285,
+      "learning_rate": 2.1459782533487176e-06,
+      "loss": 0.6973,
+      "step": 10977
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1614614598599817,
+      "learning_rate": 2.14512726945029e-06,
+      "loss": 0.7285,
+      "step": 10978
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.95303926379565,
+      "learning_rate": 2.1442764082321e-06,
+      "loss": 0.7478,
+      "step": 10979
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5286513797913601,
+      "learning_rate": 2.1434256697307103e-06,
+      "loss": 0.659,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8544994475823524,
+      "learning_rate": 2.142575053982684e-06,
+      "loss": 0.7521,
+      "step": 10981
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.856572035276391,
+      "learning_rate": 2.1417245610245714e-06,
+      "loss": 0.7393,
+      "step": 10982
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.884303382420816,
+      "learning_rate": 2.14087419089292e-06,
+      "loss": 0.7625,
+      "step": 10983
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.428531142738337,
+      "learning_rate": 2.140023943624272e-06,
+      "loss": 0.6446,
+      "step": 10984
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7161973694108348,
+      "learning_rate": 2.139173819255163e-06,
+      "loss": 0.7321,
+      "step": 10985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.57849514729423,
+      "learning_rate": 2.13832381782213e-06,
+      "loss": 0.8222,
+      "step": 10986
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.639941148056352,
+      "learning_rate": 2.1374739393616923e-06,
+      "loss": 0.6866,
+      "step": 10987
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6101417047546824,
+      "learning_rate": 2.1366241839103764e-06,
+      "loss": 0.659,
+      "step": 10988
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.127370880169811,
+      "learning_rate": 2.135774551504698e-06,
+      "loss": 0.7286,
+      "step": 10989
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.012017161226355,
+      "learning_rate": 2.1349250421811622e-06,
+      "loss": 0.7311,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7116344412619777,
+      "learning_rate": 2.1340756559762808e-06,
+      "loss": 0.7999,
+      "step": 10991
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5823545836727309,
+      "learning_rate": 2.1332263929265505e-06,
+      "loss": 0.7125,
+      "step": 10992
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0213737190240948,
+      "learning_rate": 2.132377253068467e-06,
+      "loss": 0.5448,
+      "step": 10993
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.081495623001012,
+      "learning_rate": 2.1315282364385197e-06,
+      "loss": 0.6624,
+      "step": 10994
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6749131697485558,
+      "learning_rate": 2.1306793430731908e-06,
+      "loss": 0.7893,
+      "step": 10995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2870554600253596,
+      "learning_rate": 2.1298305730089626e-06,
+      "loss": 0.6954,
+      "step": 10996
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9155549684247937,
+      "learning_rate": 2.1289819262823065e-06,
+      "loss": 0.6979,
+      "step": 10997
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7709373111600182,
+      "learning_rate": 2.1281334029296913e-06,
+      "loss": 0.8131,
+      "step": 10998
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.692823014263838,
+      "learning_rate": 2.1272850029875802e-06,
+      "loss": 0.7258,
+      "step": 10999
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6043701133549275,
+      "learning_rate": 2.1264367264924295e-06,
+      "loss": 0.7319,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.92717182762909,
+      "learning_rate": 2.125588573480693e-06,
+      "loss": 0.6922,
+      "step": 11001
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7314118277816795,
+      "learning_rate": 2.1247405439888173e-06,
+      "loss": 0.6923,
+      "step": 11002
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0133525220926756,
+      "learning_rate": 2.1238926380532423e-06,
+      "loss": 0.8004,
+      "step": 11003
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4873899425933892,
+      "learning_rate": 2.1230448557104087e-06,
+      "loss": 0.6548,
+      "step": 11004
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6301688594397084,
+      "learning_rate": 2.1221971969967424e-06,
+      "loss": 0.7402,
+      "step": 11005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.624688537789284,
+      "learning_rate": 2.121349661948673e-06,
+      "loss": 0.6804,
+      "step": 11006
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.998539303709603,
+      "learning_rate": 2.12050225060262e-06,
+      "loss": 0.7928,
+      "step": 11007
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.713831054696385,
+      "learning_rate": 2.119654962994998e-06,
+      "loss": 0.7762,
+      "step": 11008
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6160821517264579,
+      "learning_rate": 2.1188077991622174e-06,
+      "loss": 0.6945,
+      "step": 11009
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6006627647066072,
+      "learning_rate": 2.1179607591406807e-06,
+      "loss": 0.7681,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5603806221391656,
+      "learning_rate": 2.117113842966792e-06,
+      "loss": 0.6352,
+      "step": 11011
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4736912637134774,
+      "learning_rate": 2.1162670506769408e-06,
+      "loss": 0.5747,
+      "step": 11012
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8140748245877956,
+      "learning_rate": 2.1154203823075146e-06,
+      "loss": 0.7249,
+      "step": 11013
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1913627636649184,
+      "learning_rate": 2.1145738378949004e-06,
+      "loss": 0.6821,
+      "step": 11014
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6696116986530294,
+      "learning_rate": 2.113727417475475e-06,
+      "loss": 0.8658,
+      "step": 11015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0598534808822286,
+      "learning_rate": 2.112881121085611e-06,
+      "loss": 0.634,
+      "step": 11016
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.436142195660563,
+      "learning_rate": 2.112034948761674e-06,
+      "loss": 0.7384,
+      "step": 11017
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5484243550046066,
+      "learning_rate": 2.111188900540028e-06,
+      "loss": 0.6915,
+      "step": 11018
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.719377672206668,
+      "learning_rate": 2.110342976457029e-06,
+      "loss": 0.7708,
+      "step": 11019
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7314789570007987,
+      "learning_rate": 2.1094971765490266e-06,
+      "loss": 0.7656,
+      "step": 11020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8668535486707087,
+      "learning_rate": 2.108651500852367e-06,
+      "loss": 0.8131,
+      "step": 11021
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1437292784168482,
+      "learning_rate": 2.107805949403396e-06,
+      "loss": 0.7945,
+      "step": 11022
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6844737652152248,
+      "learning_rate": 2.10696052223844e-06,
+      "loss": 0.7802,
+      "step": 11023
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6582233660424102,
+      "learning_rate": 2.1061152193938355e-06,
+      "loss": 0.7316,
+      "step": 11024
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6425174138382532,
+      "learning_rate": 2.1052700409059057e-06,
+      "loss": 0.784,
+      "step": 11025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.882868264092631,
+      "learning_rate": 2.1044249868109683e-06,
+      "loss": 0.6855,
+      "step": 11026
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5715189003943277,
+      "learning_rate": 2.1035800571453386e-06,
+      "loss": 0.693,
+      "step": 11027
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5507219043098708,
+      "learning_rate": 2.1027352519453225e-06,
+      "loss": 0.7909,
+      "step": 11028
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.703149209640331,
+      "learning_rate": 2.1018905712472285e-06,
+      "loss": 0.6426,
+      "step": 11029
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1746274713009868,
+      "learning_rate": 2.1010460150873495e-06,
+      "loss": 0.6743,
+      "step": 11030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6325427598648892,
+      "learning_rate": 2.1002015835019777e-06,
+      "loss": 0.7034,
+      "step": 11031
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.990186115079494,
+      "learning_rate": 2.0993572765274044e-06,
+      "loss": 0.6925,
+      "step": 11032
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.205977002965741,
+      "learning_rate": 2.098513094199909e-06,
+      "loss": 0.6779,
+      "step": 11033
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.137554556058046,
+      "learning_rate": 2.0976690365557673e-06,
+      "loss": 0.7019,
+      "step": 11034
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9557033117340403,
+      "learning_rate": 2.096825103631251e-06,
+      "loss": 0.6726,
+      "step": 11035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9270208223803673,
+      "learning_rate": 2.0959812954626256e-06,
+      "loss": 0.7713,
+      "step": 11036
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.13191680256557,
+      "learning_rate": 2.0951376120861523e-06,
+      "loss": 0.6264,
+      "step": 11037
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8310721383803183,
+      "learning_rate": 2.0942940535380823e-06,
+      "loss": 0.7143,
+      "step": 11038
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4004979388934022,
+      "learning_rate": 2.093450619854671e-06,
+      "loss": 0.6074,
+      "step": 11039
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5143243108565974,
+      "learning_rate": 2.0926073110721613e-06,
+      "loss": 0.6048,
+      "step": 11040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.946173293699006,
+      "learning_rate": 2.0917641272267867e-06,
+      "loss": 0.6995,
+      "step": 11041
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6977132766604788,
+      "learning_rate": 2.0909210683547865e-06,
+      "loss": 0.7909,
+      "step": 11042
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.671472831175908,
+      "learning_rate": 2.0900781344923872e-06,
+      "loss": 0.7066,
+      "step": 11043
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7558782840635307,
+      "learning_rate": 2.0892353256758107e-06,
+      "loss": 0.6849,
+      "step": 11044
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4912291064987993,
+      "learning_rate": 2.0883926419412748e-06,
+      "loss": 0.7424,
+      "step": 11045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9341821226539104,
+      "learning_rate": 2.08755008332499e-06,
+      "loss": 0.7106,
+      "step": 11046
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.357337905246405,
+      "learning_rate": 2.0867076498631685e-06,
+      "loss": 0.8017,
+      "step": 11047
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5905152171382004,
+      "learning_rate": 2.0858653415920034e-06,
+      "loss": 0.6991,
+      "step": 11048
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8335542571608934,
+      "learning_rate": 2.0850231585476965e-06,
+      "loss": 0.7553,
+      "step": 11049
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8866332465718245,
+      "learning_rate": 2.084181100766437e-06,
+      "loss": 0.7753,
+      "step": 11050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7076270022235738,
+      "learning_rate": 2.083339168284409e-06,
+      "loss": 0.6206,
+      "step": 11051
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9153472157415492,
+      "learning_rate": 2.0824973611377924e-06,
+      "loss": 0.638,
+      "step": 11052
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6002976910560107,
+      "learning_rate": 2.0816556793627624e-06,
+      "loss": 0.7499,
+      "step": 11053
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6888631296949415,
+      "learning_rate": 2.0808141229954876e-06,
+      "loss": 0.7074,
+      "step": 11054
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0154980404928007,
+      "learning_rate": 2.079972692072131e-06,
+      "loss": 0.6304,
+      "step": 11055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6756753751416718,
+      "learning_rate": 2.0791313866288485e-06,
+      "loss": 0.6618,
+      "step": 11056
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.907016964335436,
+      "learning_rate": 2.078290206701798e-06,
+      "loss": 0.7056,
+      "step": 11057
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7713845462482318,
+      "learning_rate": 2.0774491523271243e-06,
+      "loss": 0.6897,
+      "step": 11058
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6810382497599277,
+      "learning_rate": 2.0766082235409695e-06,
+      "loss": 0.6978,
+      "step": 11059
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0780567029050303,
+      "learning_rate": 2.0757674203794696e-06,
+      "loss": 0.4827,
+      "step": 11060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6782786943711154,
+      "learning_rate": 2.074926742878757e-06,
+      "loss": 0.6905,
+      "step": 11061
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.980796041372373,
+      "learning_rate": 2.074086191074956e-06,
+      "loss": 0.6921,
+      "step": 11062
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7984660946034807,
+      "learning_rate": 2.0732457650041884e-06,
+      "loss": 0.7383,
+      "step": 11063
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6022073661755483,
+      "learning_rate": 2.072405464702566e-06,
+      "loss": 0.7107,
+      "step": 11064
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6674449917615468,
+      "learning_rate": 2.071565290206205e-06,
+      "loss": 0.7883,
+      "step": 11065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2696140956944477,
+      "learning_rate": 2.0707252415512015e-06,
+      "loss": 0.7442,
+      "step": 11066
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5667674519478851,
+      "learning_rate": 2.06988531877366e-06,
+      "loss": 0.6531,
+      "step": 11067
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8287848429195728,
+      "learning_rate": 2.069045521909672e-06,
+      "loss": 0.7373,
+      "step": 11068
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.73155359751554,
+      "learning_rate": 2.068205850995326e-06,
+      "loss": 0.7108,
+      "step": 11069
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4791980143669636,
+      "learning_rate": 2.067366306066704e-06,
+      "loss": 0.7238,
+      "step": 11070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7586799328637759,
+      "learning_rate": 2.0665268871598832e-06,
+      "loss": 0.7314,
+      "step": 11071
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4852605337373948,
+      "learning_rate": 2.065687594310935e-06,
+      "loss": 0.7801,
+      "step": 11072
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7241161161925782,
+      "learning_rate": 2.064848427555926e-06,
+      "loss": 0.6962,
+      "step": 11073
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5682425502195716,
+      "learning_rate": 2.064009386930915e-06,
+      "loss": 0.599,
+      "step": 11074
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7849786203515992,
+      "learning_rate": 2.0631704724719613e-06,
+      "loss": 0.6313,
+      "step": 11075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.848283859668674,
+      "learning_rate": 2.0623316842151132e-06,
+      "loss": 0.7599,
+      "step": 11076
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7439888021164505,
+      "learning_rate": 2.061493022196415e-06,
+      "loss": 0.664,
+      "step": 11077
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.685457676932178,
+      "learning_rate": 2.0606544864519064e-06,
+      "loss": 0.7568,
+      "step": 11078
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5651374493629915,
+      "learning_rate": 2.0598160770176208e-06,
+      "loss": 0.6658,
+      "step": 11079
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5798876825974355,
+      "learning_rate": 2.0589777939295855e-06,
+      "loss": 0.6506,
+      "step": 11080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.121182235378748,
+      "learning_rate": 2.0581396372238254e-06,
+      "loss": 0.6216,
+      "step": 11081
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5451538712878414,
+      "learning_rate": 2.0573016069363554e-06,
+      "loss": 0.6068,
+      "step": 11082
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4125409761671734,
+      "learning_rate": 2.056463703103192e-06,
+      "loss": 0.8228,
+      "step": 11083
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6224701024838515,
+      "learning_rate": 2.0556259257603355e-06,
+      "loss": 0.6126,
+      "step": 11084
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2214343732265722,
+      "learning_rate": 2.0547882749437923e-06,
+      "loss": 0.6557,
+      "step": 11085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6839010518380504,
+      "learning_rate": 2.053950750689556e-06,
+      "loss": 0.8936,
+      "step": 11086
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.714995193768694,
+      "learning_rate": 2.053113353033618e-06,
+      "loss": 0.6781,
+      "step": 11087
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.565774198306855,
+      "learning_rate": 2.0522760820119615e-06,
+      "loss": 0.7615,
+      "step": 11088
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7185164846005547,
+      "learning_rate": 2.0514389376605646e-06,
+      "loss": 0.718,
+      "step": 11089
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0765048016798215,
+      "learning_rate": 2.0506019200154076e-06,
+      "loss": 0.6972,
+      "step": 11090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0443858915702937,
+      "learning_rate": 2.049765029112453e-06,
+      "loss": 0.7297,
+      "step": 11091
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1078573062752004,
+      "learning_rate": 2.0489282649876633e-06,
+      "loss": 0.7019,
+      "step": 11092
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9085758414283676,
+      "learning_rate": 2.0480916276770005e-06,
+      "loss": 0.7394,
+      "step": 11093
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3283458176920373,
+      "learning_rate": 2.0472551172164152e-06,
+      "loss": 0.6268,
+      "step": 11094
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7508786321673286,
+      "learning_rate": 2.046418733641853e-06,
+      "loss": 0.6965,
+      "step": 11095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.699887617810572,
+      "learning_rate": 2.0455824769892556e-06,
+      "loss": 0.7762,
+      "step": 11096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8150825344950552,
+      "learning_rate": 2.0447463472945593e-06,
+      "loss": 0.6636,
+      "step": 11097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8326913463930834,
+      "learning_rate": 2.0439103445936946e-06,
+      "loss": 0.7117,
+      "step": 11098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0579216601755546,
+      "learning_rate": 2.0430744689225833e-06,
+      "loss": 0.6265,
+      "step": 11099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5110739893123748,
+      "learning_rate": 2.0422387203171503e-06,
+      "loss": 0.639,
+      "step": 11100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7127189899738327,
+      "learning_rate": 2.0414030988133076e-06,
+      "loss": 0.6394,
+      "step": 11101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5972117742304472,
+      "learning_rate": 2.04056760444696e-06,
+      "loss": 0.5052,
+      "step": 11102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.910245351160794,
+      "learning_rate": 2.039732237254015e-06,
+      "loss": 0.799,
+      "step": 11103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9490512473106743,
+      "learning_rate": 2.0388969972703688e-06,
+      "loss": 0.6998,
+      "step": 11104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7524668975791122,
+      "learning_rate": 2.0380618845319132e-06,
+      "loss": 0.6499,
+      "step": 11105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7529140395183378,
+      "learning_rate": 2.0372268990745347e-06,
+      "loss": 0.7914,
+      "step": 11106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7156698396461945,
+      "learning_rate": 2.036392040934114e-06,
+      "loss": 0.6206,
+      "step": 11107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.153560320466811,
+      "learning_rate": 2.03555731014653e-06,
+      "loss": 0.7676,
+      "step": 11108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6016436828344407,
+      "learning_rate": 2.0347227067476478e-06,
+      "loss": 0.6496,
+      "step": 11109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5359253410426308,
+      "learning_rate": 2.033888230773336e-06,
+      "loss": 0.7254,
+      "step": 11110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0876756217882824,
+      "learning_rate": 2.0330538822594532e-06,
+      "loss": 0.6674,
+      "step": 11111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8608150005286082,
+      "learning_rate": 2.0322196612418527e-06,
+      "loss": 0.7122,
+      "step": 11112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7024364271750811,
+      "learning_rate": 2.031385567756383e-06,
+      "loss": 0.6953,
+      "step": 11113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5682722684721544,
+      "learning_rate": 2.030551601838887e-06,
+      "loss": 0.7671,
+      "step": 11114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6963831739134858,
+      "learning_rate": 2.029717763525202e-06,
+      "loss": 0.6616,
+      "step": 11115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.891357589683478,
+      "learning_rate": 2.02888405285116e-06,
+      "loss": 0.847,
+      "step": 11116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9673828551945425,
+      "learning_rate": 2.0280504698525856e-06,
+      "loss": 0.858,
+      "step": 11117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8774471429006245,
+      "learning_rate": 2.0272170145653025e-06,
+      "loss": 0.8354,
+      "step": 11118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8004268777260004,
+      "learning_rate": 2.0263836870251277e-06,
+      "loss": 0.8026,
+      "step": 11119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0168118109355997,
+      "learning_rate": 2.025550487267864e-06,
+      "loss": 0.7265,
+      "step": 11120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5549591814132666,
+      "learning_rate": 2.0247174153293226e-06,
+      "loss": 0.663,
+      "step": 11121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6323236663268446,
+      "learning_rate": 2.0238844712453e-06,
+      "loss": 0.6765,
+      "step": 11122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8370770896274586,
+      "learning_rate": 2.02305165505159e-06,
+      "loss": 0.6414,
+      "step": 11123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5746198636849638,
+      "learning_rate": 2.0222189667839805e-06,
+      "loss": 0.6665,
+      "step": 11124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.82645831633097,
+      "learning_rate": 2.0213864064782516e-06,
+      "loss": 0.7215,
+      "step": 11125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4945633043880855,
+      "learning_rate": 2.020553974170186e-06,
+      "loss": 0.6586,
+      "step": 11126
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.115730676875051,
+      "learning_rate": 2.019721669895549e-06,
+      "loss": 0.7844,
+      "step": 11127
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4956351493697435,
+      "learning_rate": 2.018889493690111e-06,
+      "loss": 0.6963,
+      "step": 11128
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7756155397830478,
+      "learning_rate": 2.01805744558963e-06,
+      "loss": 0.686,
+      "step": 11129
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1389208555430335,
+      "learning_rate": 2.0172255256298623e-06,
+      "loss": 0.6298,
+      "step": 11130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1956337172443314,
+      "learning_rate": 2.016393733846557e-06,
+      "loss": 0.6386,
+      "step": 11131
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3904058379825444,
+      "learning_rate": 2.015562070275457e-06,
+      "loss": 0.6587,
+      "step": 11132
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.134295350709825,
+      "learning_rate": 2.0147305349523e-06,
+      "loss": 0.9235,
+      "step": 11133
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7455648964508887,
+      "learning_rate": 2.013899127912824e-06,
+      "loss": 0.6623,
+      "step": 11134
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2813072629792135,
+      "learning_rate": 2.0130678491927495e-06,
+      "loss": 0.7601,
+      "step": 11135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7596671173115697,
+      "learning_rate": 2.0122366988278035e-06,
+      "loss": 0.7384,
+      "step": 11136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7469630506361982,
+      "learning_rate": 2.0114056768537005e-06,
+      "loss": 0.6765,
+      "step": 11137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4964295079667271,
+      "learning_rate": 2.0105747833061516e-06,
+      "loss": 0.7121,
+      "step": 11138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5856062879588368,
+      "learning_rate": 2.009744018220863e-06,
+      "loss": 0.6441,
+      "step": 11139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3447823154286511,
+      "learning_rate": 2.0089133816335303e-06,
+      "loss": 0.6777,
+      "step": 11140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4494658594910557,
+      "learning_rate": 2.0080828735798554e-06,
+      "loss": 0.5845,
+      "step": 11141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.756112228860545,
+      "learning_rate": 2.0072524940955213e-06,
+      "loss": 0.7357,
+      "step": 11142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.854451234872745,
+      "learning_rate": 2.00642224321621e-06,
+      "loss": 0.6802,
+      "step": 11143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0167349594184167,
+      "learning_rate": 2.005592120977606e-06,
+      "loss": 0.7392,
+      "step": 11144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6024253543178162,
+      "learning_rate": 2.0047621274153746e-06,
+      "loss": 0.7157,
+      "step": 11145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.13144902223188,
+      "learning_rate": 2.0039322625651865e-06,
+      "loss": 0.7686,
+      "step": 11146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5567230273790287,
+      "learning_rate": 2.0031025264627015e-06,
+      "loss": 0.7794,
+      "step": 11147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.99633806606309,
+      "learning_rate": 2.002272919143576e-06,
+      "loss": 0.6491,
+      "step": 11148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.995504153985394,
+      "learning_rate": 2.0014434406434584e-06,
+      "loss": 0.8137,
+      "step": 11149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.071521923320979,
+      "learning_rate": 2.0006140909979937e-06,
+      "loss": 0.6363,
+      "step": 11150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6736072619660318,
+      "learning_rate": 1.9997848702428226e-06,
+      "loss": 0.7549,
+      "step": 11151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7096010154662302,
+      "learning_rate": 1.9989557784135792e-06,
+      "loss": 0.7698,
+      "step": 11152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6857641997618227,
+      "learning_rate": 1.9981268155458864e-06,
+      "loss": 0.6653,
+      "step": 11153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9889407384678621,
+      "learning_rate": 1.9972979816753717e-06,
+      "loss": 0.6382,
+      "step": 11154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6583131643620035,
+      "learning_rate": 1.9964692768376497e-06,
+      "loss": 0.7026,
+      "step": 11155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6543154647559388,
+      "learning_rate": 1.9956407010683324e-06,
+      "loss": 0.7387,
+      "step": 11156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.058390235014486,
+      "learning_rate": 1.994812254403025e-06,
+      "loss": 0.6219,
+      "step": 11157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7254319023844116,
+      "learning_rate": 1.9939839368773267e-06,
+      "loss": 0.6396,
+      "step": 11158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8281006452217943,
+      "learning_rate": 1.9931557485268365e-06,
+      "loss": 0.6832,
+      "step": 11159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0852721632029683,
+      "learning_rate": 1.992327689387137e-06,
+      "loss": 0.6114,
+      "step": 11160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.073912717320372,
+      "learning_rate": 1.991499759493817e-06,
+      "loss": 0.7179,
+      "step": 11161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7852779746791958,
+      "learning_rate": 1.990671958882454e-06,
+      "loss": 0.717,
+      "step": 11162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3035082359932204,
+      "learning_rate": 1.9898442875886157e-06,
+      "loss": 0.6792,
+      "step": 11163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0126792767707045,
+      "learning_rate": 1.9890167456478748e-06,
+      "loss": 0.5962,
+      "step": 11164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2131876116598215,
+      "learning_rate": 1.9881893330957893e-06,
+      "loss": 0.7275,
+      "step": 11165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.512512109876676,
+      "learning_rate": 1.9873620499679157e-06,
+      "loss": 0.7062,
+      "step": 11166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3322268102987527,
+      "learning_rate": 1.986534896299805e-06,
+      "loss": 0.6348,
+      "step": 11167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5489910642494742,
+      "learning_rate": 1.985707872126999e-06,
+      "loss": 0.7381,
+      "step": 11168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5597453611681007,
+      "learning_rate": 1.984880977485041e-06,
+      "loss": 0.6505,
+      "step": 11169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0497809118617367,
+      "learning_rate": 1.9840542124094635e-06,
+      "loss": 0.6346,
+      "step": 11170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.549151800395625,
+      "learning_rate": 1.9832275769357906e-06,
+      "loss": 0.6203,
+      "step": 11171
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7096525122509014,
+      "learning_rate": 1.982401071099549e-06,
+      "loss": 0.7828,
+      "step": 11172
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5067439896465595,
+      "learning_rate": 1.9815746949362536e-06,
+      "loss": 0.7474,
+      "step": 11173
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7163278022166175,
+      "learning_rate": 1.980748448481416e-06,
+      "loss": 0.7363,
+      "step": 11174
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7518923856464974,
+      "learning_rate": 1.9799223317705426e-06,
+      "loss": 0.7947,
+      "step": 11175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8866956826513013,
+      "learning_rate": 1.97909634483913e-06,
+      "loss": 0.9296,
+      "step": 11176
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7595134303373612,
+      "learning_rate": 1.9782704877226796e-06,
+      "loss": 0.6562,
+      "step": 11177
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6814881462608269,
+      "learning_rate": 1.9774447604566715e-06,
+      "loss": 0.7972,
+      "step": 11178
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6363324952790264,
+      "learning_rate": 1.9766191630765964e-06,
+      "loss": 0.765,
+      "step": 11179
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7324081730727934,
+      "learning_rate": 1.9757936956179308e-06,
+      "loss": 0.7404,
+      "step": 11180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0235886951289714,
+      "learning_rate": 1.9749683581161417e-06,
+      "loss": 0.512,
+      "step": 11181
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9855435909223613,
+      "learning_rate": 1.9741431506067012e-06,
+      "loss": 0.7309,
+      "step": 11182
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6890570511877943,
+      "learning_rate": 1.9733180731250683e-06,
+      "loss": 0.6557,
+      "step": 11183
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6514441870876422,
+      "learning_rate": 1.9724931257066988e-06,
+      "loss": 0.7386,
+      "step": 11184
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2052141748996976,
+      "learning_rate": 1.971668308387042e-06,
+      "loss": 0.5689,
+      "step": 11185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6794004084444827,
+      "learning_rate": 1.970843621201541e-06,
+      "loss": 0.7461,
+      "step": 11186
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5296540031963715,
+      "learning_rate": 1.9700190641856378e-06,
+      "loss": 0.706,
+      "step": 11187
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9614196320357888,
+      "learning_rate": 1.9691946373747626e-06,
+      "loss": 0.8237,
+      "step": 11188
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0826789568951618,
+      "learning_rate": 1.9683703408043447e-06,
+      "loss": 0.6494,
+      "step": 11189
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9858646196634715,
+      "learning_rate": 1.967546174509805e-06,
+      "loss": 0.8372,
+      "step": 11190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8599148223999464,
+      "learning_rate": 1.9667221385265595e-06,
+      "loss": 0.7809,
+      "step": 11191
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4429920236434288,
+      "learning_rate": 1.9658982328900195e-06,
+      "loss": 0.7568,
+      "step": 11192
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.951270279034149,
+      "learning_rate": 1.9650744576355894e-06,
+      "loss": 0.6171,
+      "step": 11193
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2650547747071497,
+      "learning_rate": 1.9642508127986676e-06,
+      "loss": 0.6322,
+      "step": 11194
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.768453207782512,
+      "learning_rate": 1.963427298414653e-06,
+      "loss": 0.644,
+      "step": 11195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6024680455031137,
+      "learning_rate": 1.9626039145189263e-06,
+      "loss": 0.7436,
+      "step": 11196
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.687551674654731,
+      "learning_rate": 1.9617806611468756e-06,
+      "loss": 0.6791,
+      "step": 11197
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5600419540435995,
+      "learning_rate": 1.960957538333877e-06,
+      "loss": 0.6746,
+      "step": 11198
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.728276660201186,
+      "learning_rate": 1.9601345461153005e-06,
+      "loss": 0.6746,
+      "step": 11199
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0328223835153898,
+      "learning_rate": 1.959311684526513e-06,
+      "loss": 0.6059,
+      "step": 11200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0561576586909087,
+      "learning_rate": 1.9584889536028728e-06,
+      "loss": 0.7561,
+      "step": 11201
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7485519015640951,
+      "learning_rate": 1.95766635337974e-06,
+      "loss": 0.7275,
+      "step": 11202
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7269328160186908,
+      "learning_rate": 1.956843883892457e-06,
+      "loss": 0.7589,
+      "step": 11203
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0039886088138517,
+      "learning_rate": 1.9560215451763685e-06,
+      "loss": 0.632,
+      "step": 11204
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8201734297093362,
+      "learning_rate": 1.9551993372668153e-06,
+      "loss": 0.7265,
+      "step": 11205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.795310800854075,
+      "learning_rate": 1.9543772601991277e-06,
+      "loss": 0.7504,
+      "step": 11206
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6610332867431974,
+      "learning_rate": 1.9535553140086322e-06,
+      "loss": 0.7601,
+      "step": 11207
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1289900784624463,
+      "learning_rate": 1.9527334987306495e-06,
+      "loss": 0.7305,
+      "step": 11208
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6858815052406533,
+      "learning_rate": 1.951911814400495e-06,
+      "loss": 0.6484,
+      "step": 11209
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.780854635201375,
+      "learning_rate": 1.9510902610534782e-06,
+      "loss": 0.8385,
+      "step": 11210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6476835083128867,
+      "learning_rate": 1.950268838724902e-06,
+      "loss": 0.7433,
+      "step": 11211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7861835726014768,
+      "learning_rate": 1.949447547450067e-06,
+      "loss": 0.7891,
+      "step": 11212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7360405602355409,
+      "learning_rate": 1.948626387264267e-06,
+      "loss": 0.7263,
+      "step": 11213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1699549666471276,
+      "learning_rate": 1.9478053582027826e-06,
+      "loss": 0.621,
+      "step": 11214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9684725803134926,
+      "learning_rate": 1.9469844603009026e-06,
+      "loss": 0.7137,
+      "step": 11215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7316069542730719,
+      "learning_rate": 1.946163693593899e-06,
+      "loss": 0.6681,
+      "step": 11216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7026241152592296,
+      "learning_rate": 1.945343058117043e-06,
+      "loss": 0.675,
+      "step": 11217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8350367433138433,
+      "learning_rate": 1.9445225539055982e-06,
+      "loss": 0.6408,
+      "step": 11218
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.96160089071406,
+      "learning_rate": 1.9437021809948232e-06,
+      "loss": 0.7295,
+      "step": 11219
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.311184018160164,
+      "learning_rate": 1.9428819394199753e-06,
+      "loss": 0.6395,
+      "step": 11220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9893948059333308,
+      "learning_rate": 1.9420618292162974e-06,
+      "loss": 0.7063,
+      "step": 11221
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8079735440640816,
+      "learning_rate": 1.9412418504190316e-06,
+      "loss": 0.7458,
+      "step": 11222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2783953233192973,
+      "learning_rate": 1.9404220030634173e-06,
+      "loss": 0.6441,
+      "step": 11223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5682637233639554,
+      "learning_rate": 1.9396022871846836e-06,
+      "loss": 0.687,
+      "step": 11224
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.844204566547277,
+      "learning_rate": 1.9387827028180556e-06,
+      "loss": 0.6428,
+      "step": 11225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6859865413143797,
+      "learning_rate": 1.937963249998752e-06,
+      "loss": 0.7691,
+      "step": 11226
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.319752884659196,
+      "learning_rate": 1.9371439287619876e-06,
+      "loss": 0.7743,
+      "step": 11227
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.188159162156767,
+      "learning_rate": 1.9363247391429695e-06,
+      "loss": 0.696,
+      "step": 11228
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8272046510944646,
+      "learning_rate": 1.935505681176899e-06,
+      "loss": 0.7728,
+      "step": 11229
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.568978710038078,
+      "learning_rate": 1.9346867548989767e-06,
+      "loss": 0.6239,
+      "step": 11230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6155368041098899,
+      "learning_rate": 1.9338679603443922e-06,
+      "loss": 0.6568,
+      "step": 11231
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6760549156864903,
+      "learning_rate": 1.933049297548327e-06,
+      "loss": 0.66,
+      "step": 11232
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0244173840350634,
+      "learning_rate": 1.932230766545966e-06,
+      "loss": 0.581,
+      "step": 11233
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0632967369550665,
+      "learning_rate": 1.9314123673724805e-06,
+      "loss": 0.7014,
+      "step": 11234
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7260315926949796,
+      "learning_rate": 1.93059410006304e-06,
+      "loss": 0.7185,
+      "step": 11235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6484516737116428,
+      "learning_rate": 1.9297759646528076e-06,
+      "loss": 0.6888,
+      "step": 11236
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0882348397296486,
+      "learning_rate": 1.928957961176937e-06,
+      "loss": 0.5921,
+      "step": 11237
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3360913578046707,
+      "learning_rate": 1.9281400896705865e-06,
+      "loss": 0.7229,
+      "step": 11238
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5539439597293006,
+      "learning_rate": 1.9273223501688943e-06,
+      "loss": 0.7747,
+      "step": 11239
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4287159422512758,
+      "learning_rate": 1.9265047427070054e-06,
+      "loss": 0.629,
+      "step": 11240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1110050289015603,
+      "learning_rate": 1.925687267320053e-06,
+      "loss": 0.654,
+      "step": 11241
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9565079260378098,
+      "learning_rate": 1.924869924043165e-06,
+      "loss": 0.8203,
+      "step": 11242
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.980966861682095,
+      "learning_rate": 1.9240527129114654e-06,
+      "loss": 0.6542,
+      "step": 11243
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6788772650304014,
+      "learning_rate": 1.9232356339600717e-06,
+      "loss": 0.666,
+      "step": 11244
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6056820640357508,
+      "learning_rate": 1.9224186872240946e-06,
+      "loss": 0.6204,
+      "step": 11245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1123407603765183,
+      "learning_rate": 1.921601872738641e-06,
+      "loss": 0.7641,
+      "step": 11246
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0841001007772397,
+      "learning_rate": 1.920785190538809e-06,
+      "loss": 0.6793,
+      "step": 11247
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6370762436248245,
+      "learning_rate": 1.9199686406596968e-06,
+      "loss": 0.6868,
+      "step": 11248
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7488581997802146,
+      "learning_rate": 1.919152223136391e-06,
+      "loss": 0.7758,
+      "step": 11249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9365798490282247,
+      "learning_rate": 1.918335938003977e-06,
+      "loss": 0.7324,
+      "step": 11250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7678393727012478,
+      "learning_rate": 1.9175197852975304e-06,
+      "loss": 0.7994,
+      "step": 11251
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6046244049445426,
+      "learning_rate": 1.9167037650521215e-06,
+      "loss": 0.6746,
+      "step": 11252
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5541338032996708,
+      "learning_rate": 1.915887877302823e-06,
+      "loss": 0.7213,
+      "step": 11253
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0564266975339534,
+      "learning_rate": 1.9150721220846884e-06,
+      "loss": 0.7061,
+      "step": 11254
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5727758015813753,
+      "learning_rate": 1.914256499432774e-06,
+      "loss": 0.7022,
+      "step": 11255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8064213992042475,
+      "learning_rate": 1.913441009382133e-06,
+      "loss": 0.769,
+      "step": 11256
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4790569875574016,
+      "learning_rate": 1.9126256519678034e-06,
+      "loss": 0.7548,
+      "step": 11257
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6533753132490294,
+      "learning_rate": 1.9118104272248266e-06,
+      "loss": 0.7481,
+      "step": 11258
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1422768401021015,
+      "learning_rate": 1.910995335188234e-06,
+      "loss": 0.5928,
+      "step": 11259
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6593387599274065,
+      "learning_rate": 1.9101803758930514e-06,
+      "loss": 0.6548,
+      "step": 11260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4756243115094236,
+      "learning_rate": 1.909365549374299e-06,
+      "loss": 0.6947,
+      "step": 11261
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9176248516161567,
+      "learning_rate": 1.9085508556669905e-06,
+      "loss": 0.6734,
+      "step": 11262
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.623561447944894,
+      "learning_rate": 1.9077362948061404e-06,
+      "loss": 0.7658,
+      "step": 11263
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5718236211693941,
+      "learning_rate": 1.906921866826747e-06,
+      "loss": 0.5702,
+      "step": 11264
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5227704899493089,
+      "learning_rate": 1.9061075717638078e-06,
+      "loss": 0.668,
+      "step": 11265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5596241577615695,
+      "learning_rate": 1.9052934096523184e-06,
+      "loss": 0.6619,
+      "step": 11266
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9968906237683073,
+      "learning_rate": 1.9044793805272633e-06,
+      "loss": 0.7372,
+      "step": 11267
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9213528829116409,
+      "learning_rate": 1.9036654844236234e-06,
+      "loss": 0.6568,
+      "step": 11268
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4717663553061988,
+      "learning_rate": 1.9028517213763737e-06,
+      "loss": 0.7279,
+      "step": 11269
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.958344717344376,
+      "learning_rate": 1.902038091420481e-06,
+      "loss": 0.6937,
+      "step": 11270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5858747620450568,
+      "learning_rate": 1.9012245945909153e-06,
+      "loss": 0.6738,
+      "step": 11271
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8546038475847297,
+      "learning_rate": 1.900411230922628e-06,
+      "loss": 0.6906,
+      "step": 11272
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7619908105335664,
+      "learning_rate": 1.8995980004505715e-06,
+      "loss": 0.7137,
+      "step": 11273
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.679382691614524,
+      "learning_rate": 1.8987849032096973e-06,
+      "loss": 0.7445,
+      "step": 11274
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6216820283444935,
+      "learning_rate": 1.8979719392349388e-06,
+      "loss": 0.657,
+      "step": 11275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5292043737417067,
+      "learning_rate": 1.8971591085612367e-06,
+      "loss": 0.6344,
+      "step": 11276
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7614549642768773,
+      "learning_rate": 1.8963464112235185e-06,
+      "loss": 0.6736,
+      "step": 11277
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6566303733521306,
+      "learning_rate": 1.895533847256707e-06,
+      "loss": 0.7428,
+      "step": 11278
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7484589024057473,
+      "learning_rate": 1.89472141669572e-06,
+      "loss": 0.78,
+      "step": 11279
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1886038747774894,
+      "learning_rate": 1.8939091195754678e-06,
+      "loss": 0.64,
+      "step": 11280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3593294430287777,
+      "learning_rate": 1.8930969559308615e-06,
+      "loss": 0.688,
+      "step": 11281
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6676909723938487,
+      "learning_rate": 1.8922849257967972e-06,
+      "loss": 0.7573,
+      "step": 11282
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6769545786229438,
+      "learning_rate": 1.891473029208169e-06,
+      "loss": 0.6993,
+      "step": 11283
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6196035142446519,
+      "learning_rate": 1.8906612661998698e-06,
+      "loss": 0.7654,
+      "step": 11284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1986898426715713,
+      "learning_rate": 1.8898496368067804e-06,
+      "loss": 0.6998,
+      "step": 11285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9847020407788571,
+      "learning_rate": 1.8890381410637786e-06,
+      "loss": 0.7336,
+      "step": 11286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6301589490499515,
+      "learning_rate": 1.8882267790057374e-06,
+      "loss": 0.7487,
+      "step": 11287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234826072240057,
+      "learning_rate": 1.887415550667519e-06,
+      "loss": 0.6658,
+      "step": 11288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8358987764152532,
+      "learning_rate": 1.8866044560839902e-06,
+      "loss": 0.7823,
+      "step": 11289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6272634045170793,
+      "learning_rate": 1.8857934952899987e-06,
+      "loss": 0.6585,
+      "step": 11290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5442623990935953,
+      "learning_rate": 1.884982668320398e-06,
+      "loss": 0.7152,
+      "step": 11291
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8695311471665033,
+      "learning_rate": 1.8841719752100313e-06,
+      "loss": 0.8405,
+      "step": 11292
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.998236880147267,
+      "learning_rate": 1.8833614159937307e-06,
+      "loss": 0.685,
+      "step": 11293
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5811562101356618,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.7108,
+      "step": 11294
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7123456089226474,
+      "learning_rate": 1.881740699382662e-06,
+      "loss": 0.7344,
+      "step": 11295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7835483637693241,
+      "learning_rate": 1.880930542057538e-06,
+      "loss": 0.6928,
+      "step": 11296
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.74146589762445,
+      "learning_rate": 1.880120518765775e-06,
+      "loss": 0.6709,
+      "step": 11297
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6718748715951353,
+      "learning_rate": 1.8793106295421797e-06,
+      "loss": 0.7158,
+      "step": 11298
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.107887624714788,
+      "learning_rate": 1.8785008744215606e-06,
+      "loss": 0.6749,
+      "step": 11299
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.823393156751884,
+      "learning_rate": 1.8776912534387076e-06,
+      "loss": 0.7258,
+      "step": 11300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1302975190292908,
+      "learning_rate": 1.8768817666284166e-06,
+      "loss": 0.725,
+      "step": 11301
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7161587408649253,
+      "learning_rate": 1.8760724140254716e-06,
+      "loss": 0.7075,
+      "step": 11302
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6127660271651745,
+      "learning_rate": 1.8752631956646522e-06,
+      "loss": 0.643,
+      "step": 11303
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.654281836023095,
+      "learning_rate": 1.874454111580733e-06,
+      "loss": 0.7956,
+      "step": 11304
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7644532402973307,
+      "learning_rate": 1.873645161808481e-06,
+      "loss": 0.6205,
+      "step": 11305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6868653230747148,
+      "learning_rate": 1.8728363463826577e-06,
+      "loss": 0.7181,
+      "step": 11306
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2866374053449903,
+      "learning_rate": 1.872027665338025e-06,
+      "loss": 0.6491,
+      "step": 11307
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5903696393377584,
+      "learning_rate": 1.871219118709326e-06,
+      "loss": 0.7474,
+      "step": 11308
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.619689412954021,
+      "learning_rate": 1.8704107065313116e-06,
+      "loss": 0.7536,
+      "step": 11309
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.060535985054271,
+      "learning_rate": 1.8696024288387189e-06,
+      "loss": 0.6197,
+      "step": 11310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0299874642036335,
+      "learning_rate": 1.8687942856662815e-06,
+      "loss": 0.7286,
+      "step": 11311
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7174291193801974,
+      "learning_rate": 1.8679862770487273e-06,
+      "loss": 0.6738,
+      "step": 11312
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9025572172386298,
+      "learning_rate": 1.8671784030207762e-06,
+      "loss": 0.7142,
+      "step": 11313
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2397266997901588,
+      "learning_rate": 1.8663706636171503e-06,
+      "loss": 0.5959,
+      "step": 11314
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6357709848799717,
+      "learning_rate": 1.8655630588725536e-06,
+      "loss": 0.6612,
+      "step": 11315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6476729425253296,
+      "learning_rate": 1.8647555888216918e-06,
+      "loss": 0.773,
+      "step": 11316
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6823249093816264,
+      "learning_rate": 1.8639482534992682e-06,
+      "loss": 0.8271,
+      "step": 11317
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5952934289909046,
+      "learning_rate": 1.863141052939969e-06,
+      "loss": 0.7061,
+      "step": 11318
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1741452677268653,
+      "learning_rate": 1.8623339871784869e-06,
+      "loss": 0.7389,
+      "step": 11319
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.571386640494329,
+      "learning_rate": 1.8615270562495014e-06,
+      "loss": 0.7236,
+      "step": 11320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4593165462909592,
+      "learning_rate": 1.8607202601876873e-06,
+      "loss": 0.6635,
+      "step": 11321
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1936013699855317,
+      "learning_rate": 1.8599135990277156e-06,
+      "loss": 0.6128,
+      "step": 11322
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.764148042774969,
+      "learning_rate": 1.8591070728042503e-06,
+      "loss": 0.7713,
+      "step": 11323
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.497699935710057,
+      "learning_rate": 1.8583006815519473e-06,
+      "loss": 0.6375,
+      "step": 11324
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.720314730226272,
+      "learning_rate": 1.8574944253054638e-06,
+      "loss": 0.6661,
+      "step": 11325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4557538320148664,
+      "learning_rate": 1.8566883040994411e-06,
+      "loss": 0.6915,
+      "step": 11326
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7822703761984686,
+      "learning_rate": 1.855882317968523e-06,
+      "loss": 0.771,
+      "step": 11327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7604363998820856,
+      "learning_rate": 1.8550764669473454e-06,
+      "loss": 0.8311,
+      "step": 11328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7117045469066483,
+      "learning_rate": 1.8542707510705355e-06,
+      "loss": 0.6777,
+      "step": 11329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6678611748073953,
+      "learning_rate": 1.853465170372717e-06,
+      "loss": 0.7054,
+      "step": 11330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9245031687706855,
+      "learning_rate": 1.8526597248885065e-06,
+      "loss": 0.7868,
+      "step": 11331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4707019826414176,
+      "learning_rate": 1.8518544146525203e-06,
+      "loss": 0.7008,
+      "step": 11332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4758169842025228,
+      "learning_rate": 1.8510492396993595e-06,
+      "loss": 0.6841,
+      "step": 11333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2339016668728955,
+      "learning_rate": 1.8502442000636246e-06,
+      "loss": 0.6253,
+      "step": 11334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.776843972952931,
+      "learning_rate": 1.8494392957799146e-06,
+      "loss": 0.6067,
+      "step": 11335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7706270746933355,
+      "learning_rate": 1.8486345268828115e-06,
+      "loss": 0.678,
+      "step": 11336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.851179202906125,
+      "learning_rate": 1.8478298934069028e-06,
+      "loss": 0.7883,
+      "step": 11337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5909160484220632,
+      "learning_rate": 1.8470253953867635e-06,
+      "loss": 0.6949,
+      "step": 11338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5817683485385556,
+      "learning_rate": 1.846221032856965e-06,
+      "loss": 0.6683,
+      "step": 11339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4330461494020597,
+      "learning_rate": 1.8454168058520732e-06,
+      "loss": 0.7085,
+      "step": 11340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7518121533140303,
+      "learning_rate": 1.8446127144066445e-06,
+      "loss": 0.7065,
+      "step": 11341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0636646431314096,
+      "learning_rate": 1.8438087585552371e-06,
+      "loss": 0.7573,
+      "step": 11342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6842100319705715,
+      "learning_rate": 1.8430049383323983e-06,
+      "loss": 0.6618,
+      "step": 11343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.155017877922409,
+      "learning_rate": 1.8422012537726646e-06,
+      "loss": 0.6021,
+      "step": 11344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4317707055843822,
+      "learning_rate": 1.8413977049105774e-06,
+      "loss": 0.715,
+      "step": 11345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8102161630024405,
+      "learning_rate": 1.8405942917806657e-06,
+      "loss": 0.7117,
+      "step": 11346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6826660318675943,
+      "learning_rate": 1.8397910144174536e-06,
+      "loss": 0.7023,
+      "step": 11347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0279087086968683,
+      "learning_rate": 1.8389878728554594e-06,
+      "loss": 0.6421,
+      "step": 11348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.673648471414835,
+      "learning_rate": 1.8381848671291953e-06,
+      "loss": 0.7802,
+      "step": 11349
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.045081275179511,
+      "learning_rate": 1.8373819972731721e-06,
+      "loss": 0.6177,
+      "step": 11350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.770783866902555,
+      "learning_rate": 1.8365792633218844e-06,
+      "loss": 0.6637,
+      "step": 11351
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6513236299732423,
+      "learning_rate": 1.835776665309833e-06,
+      "loss": 0.8371,
+      "step": 11352
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1647846856836126,
+      "learning_rate": 1.8349742032715072e-06,
+      "loss": 0.73,
+      "step": 11353
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6505590936668402,
+      "learning_rate": 1.8341718772413852e-06,
+      "loss": 0.7781,
+      "step": 11354
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7321930066566322,
+      "learning_rate": 1.8333696872539502e-06,
+      "loss": 0.7217,
+      "step": 11355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5575451100335678,
+      "learning_rate": 1.8325676333436727e-06,
+      "loss": 0.6749,
+      "step": 11356
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0698087283276942,
+      "learning_rate": 1.8317657155450174e-06,
+      "loss": 0.7067,
+      "step": 11357
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7848269401742407,
+      "learning_rate": 1.8309639338924456e-06,
+      "loss": 0.6245,
+      "step": 11358
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7089043800958224,
+      "learning_rate": 1.8301622884204096e-06,
+      "loss": 0.8024,
+      "step": 11359
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5845164923892907,
+      "learning_rate": 1.8293607791633617e-06,
+      "loss": 0.6697,
+      "step": 11360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2000085368189572,
+      "learning_rate": 1.8285594061557421e-06,
+      "loss": 0.5838,
+      "step": 11361
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.754594599647029,
+      "learning_rate": 1.8277581694319884e-06,
+      "loss": 0.7309,
+      "step": 11362
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8681483603770423,
+      "learning_rate": 1.8269570690265308e-06,
+      "loss": 0.7198,
+      "step": 11363
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6423222508730158,
+      "learning_rate": 1.8261561049737946e-06,
+      "loss": 0.7059,
+      "step": 11364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0632518368144106,
+      "learning_rate": 1.8253552773081995e-06,
+      "loss": 0.6423,
+      "step": 11365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.647741289926236,
+      "learning_rate": 1.8245545860641584e-06,
+      "loss": 0.6748,
+      "step": 11366
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9923223972858114,
+      "learning_rate": 1.8237540312760764e-06,
+      "loss": 0.613,
+      "step": 11367
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.672737549560842,
+      "learning_rate": 1.822953612978362e-06,
+      "loss": 0.6592,
+      "step": 11368
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5596615274616124,
+      "learning_rate": 1.8221533312054024e-06,
+      "loss": 0.6487,
+      "step": 11369
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0687687653928948,
+      "learning_rate": 1.8213531859915939e-06,
+      "loss": 0.7044,
+      "step": 11370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.812824209894235,
+      "learning_rate": 1.8205531773713175e-06,
+      "loss": 0.6659,
+      "step": 11371
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.636350757057022,
+      "learning_rate": 1.8197533053789524e-06,
+      "loss": 0.6663,
+      "step": 11372
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.780727757441638,
+      "learning_rate": 1.8189535700488708e-06,
+      "loss": 0.7095,
+      "step": 11373
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1115516596404524,
+      "learning_rate": 1.818153971415439e-06,
+      "loss": 0.5921,
+      "step": 11374
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5744998896690918,
+      "learning_rate": 1.817354509513017e-06,
+      "loss": 0.5983,
+      "step": 11375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6080489269185172,
+      "learning_rate": 1.81655518437596e-06,
+      "loss": 0.7363,
+      "step": 11376
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8583903391496852,
+      "learning_rate": 1.8157559960386146e-06,
+      "loss": 0.7265,
+      "step": 11377
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6467749400738492,
+      "learning_rate": 1.8149569445353277e-06,
+      "loss": 0.6145,
+      "step": 11378
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5684614880272187,
+      "learning_rate": 1.8141580299004342e-06,
+      "loss": 0.6678,
+      "step": 11379
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5518680139663312,
+      "learning_rate": 1.8133592521682657e-06,
+      "loss": 0.8319,
+      "step": 11380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7541795536371747,
+      "learning_rate": 1.812560611373147e-06,
+      "loss": 0.6807,
+      "step": 11381
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2504830469524728,
+      "learning_rate": 1.8117621075493979e-06,
+      "loss": 0.6076,
+      "step": 11382
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.32592572215024,
+      "learning_rate": 1.8109637407313313e-06,
+      "loss": 0.7226,
+      "step": 11383
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1397359953272022,
+      "learning_rate": 1.8101655109532552e-06,
+      "loss": 0.6697,
+      "step": 11384
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.174418721653531,
+      "learning_rate": 1.8093674182494698e-06,
+      "loss": 0.5778,
+      "step": 11385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6618236088739229,
+      "learning_rate": 1.8085694626542755e-06,
+      "loss": 0.6711,
+      "step": 11386
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9177351621099896,
+      "learning_rate": 1.8077716442019554e-06,
+      "loss": 0.792,
+      "step": 11387
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8776630369640974,
+      "learning_rate": 1.8069739629267991e-06,
+      "loss": 0.7616,
+      "step": 11388
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.558494092783764,
+      "learning_rate": 1.8061764188630831e-06,
+      "loss": 0.7698,
+      "step": 11389
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1535155276901063,
+      "learning_rate": 1.8053790120450793e-06,
+      "loss": 0.6907,
+      "step": 11390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7446886101210628,
+      "learning_rate": 1.8045817425070544e-06,
+      "loss": 0.6484,
+      "step": 11391
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0905874591068745,
+      "learning_rate": 1.8037846102832667e-06,
+      "loss": 0.5795,
+      "step": 11392
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.007635628045368,
+      "learning_rate": 1.8029876154079755e-06,
+      "loss": 0.7185,
+      "step": 11393
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9032470499495044,
+      "learning_rate": 1.8021907579154257e-06,
+      "loss": 0.724,
+      "step": 11394
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4844017979795712,
+      "learning_rate": 1.8013940378398593e-06,
+      "loss": 0.5793,
+      "step": 11395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.806888680672446,
+      "learning_rate": 1.8005974552155158e-06,
+      "loss": 0.8406,
+      "step": 11396
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7337203708544149,
+      "learning_rate": 1.7998010100766255e-06,
+      "loss": 0.8342,
+      "step": 11397
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.881721567355076,
+      "learning_rate": 1.799004702457413e-06,
+      "loss": 0.8107,
+      "step": 11398
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0242515630791604,
+      "learning_rate": 1.7982085323920973e-06,
+      "loss": 0.6185,
+      "step": 11399
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5369307088368926,
+      "learning_rate": 1.7974124999148918e-06,
+      "loss": 0.7241,
+      "step": 11400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6264397437644589,
+      "learning_rate": 1.7966166050600037e-06,
+      "loss": 0.7345,
+      "step": 11401
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6255148404149637,
+      "learning_rate": 1.795820847861633e-06,
+      "loss": 0.7594,
+      "step": 11402
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4644693067762833,
+      "learning_rate": 1.7950252283539776e-06,
+      "loss": 0.7171,
+      "step": 11403
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5726441960326858,
+      "learning_rate": 1.7942297465712282e-06,
+      "loss": 0.7199,
+      "step": 11404
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.76562368601498,
+      "learning_rate": 1.7934344025475625e-06,
+      "loss": 0.7528,
+      "step": 11405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5783272392183134,
+      "learning_rate": 1.7926391963171642e-06,
+      "loss": 0.8008,
+      "step": 11406
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8853301674320502,
+      "learning_rate": 1.7918441279142025e-06,
+      "loss": 0.7434,
+      "step": 11407
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9594315673598657,
+      "learning_rate": 1.7910491973728433e-06,
+      "loss": 0.7033,
+      "step": 11408
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7002659000627425,
+      "learning_rate": 1.7902544047272468e-06,
+      "loss": 0.6598,
+      "step": 11409
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7886311428103778,
+      "learning_rate": 1.7894597500115657e-06,
+      "loss": 0.6645,
+      "step": 11410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3468259745106785,
+      "learning_rate": 1.788665233259953e-06,
+      "loss": 0.5657,
+      "step": 11411
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6687553760823846,
+      "learning_rate": 1.7878708545065433e-06,
+      "loss": 0.7764,
+      "step": 11412
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6094225263251087,
+      "learning_rate": 1.7870766137854784e-06,
+      "loss": 0.5683,
+      "step": 11413
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3679378284592707,
+      "learning_rate": 1.7862825111308873e-06,
+      "loss": 0.5897,
+      "step": 11414
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1529865891687971,
+      "learning_rate": 1.7854885465768939e-06,
+      "loss": 0.6502,
+      "step": 11415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9405441570572097,
+      "learning_rate": 1.7846947201576171e-06,
+      "loss": 0.8137,
+      "step": 11416
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.577706898417918,
+      "learning_rate": 1.7839010319071687e-06,
+      "loss": 0.7788,
+      "step": 11417
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6981943436675166,
+      "learning_rate": 1.7831074818596555e-06,
+      "loss": 0.7182,
+      "step": 11418
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6601704434400608,
+      "learning_rate": 1.7823140700491786e-06,
+      "loss": 0.6019,
+      "step": 11419
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6293165550927347,
+      "learning_rate": 1.7815207965098303e-06,
+      "loss": 0.79,
+      "step": 11420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8882692644383416,
+      "learning_rate": 1.780727661275703e-06,
+      "loss": 0.9011,
+      "step": 11421
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7020104402824665,
+      "learning_rate": 1.7799346643808796e-06,
+      "loss": 0.6942,
+      "step": 11422
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7992333623673884,
+      "learning_rate": 1.7791418058594318e-06,
+      "loss": 0.7591,
+      "step": 11423
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.513221703388717,
+      "learning_rate": 1.7783490857454354e-06,
+      "loss": 0.6899,
+      "step": 11424
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5005537693194482,
+      "learning_rate": 1.7775565040729536e-06,
+      "loss": 0.6974,
+      "step": 11425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4656990319254957,
+      "learning_rate": 1.776764060876046e-06,
+      "loss": 0.6729,
+      "step": 11426
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9997419100279397,
+      "learning_rate": 1.7759717561887652e-06,
+      "loss": 0.6415,
+      "step": 11427
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8924808806673162,
+      "learning_rate": 1.7751795900451562e-06,
+      "loss": 0.5839,
+      "step": 11428
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5941025224522996,
+      "learning_rate": 1.7743875624792662e-06,
+      "loss": 0.5789,
+      "step": 11429
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0485357369770814,
+      "learning_rate": 1.7735956735251224e-06,
+      "loss": 0.583,
+      "step": 11430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.755461632753523,
+      "learning_rate": 1.7728039232167603e-06,
+      "loss": 0.6762,
+      "step": 11431
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0813001379399847,
+      "learning_rate": 1.7720123115882009e-06,
+      "loss": 0.5442,
+      "step": 11432
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8672941389102915,
+      "learning_rate": 1.771220838673462e-06,
+      "loss": 0.737,
+      "step": 11433
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5578324957622325,
+      "learning_rate": 1.770429504506554e-06,
+      "loss": 0.6576,
+      "step": 11434
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7428867058342639,
+      "learning_rate": 1.769638309121483e-06,
+      "loss": 0.7231,
+      "step": 11435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6862327309497473,
+      "learning_rate": 1.7688472525522481e-06,
+      "loss": 0.6362,
+      "step": 11436
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6790703760204357,
+      "learning_rate": 1.7680563348328433e-06,
+      "loss": 0.7029,
+      "step": 11437
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6345863948156258,
+      "learning_rate": 1.7672655559972535e-06,
+      "loss": 0.7995,
+      "step": 11438
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.524602014287223,
+      "learning_rate": 1.7664749160794642e-06,
+      "loss": 0.6327,
+      "step": 11439
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.997890342490487,
+      "learning_rate": 1.7656844151134495e-06,
+      "loss": 0.696,
+      "step": 11440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5863131989621955,
+      "learning_rate": 1.7648940531331787e-06,
+      "loss": 0.7111,
+      "step": 11441
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.78310396217392,
+      "learning_rate": 1.7641038301726155e-06,
+      "loss": 0.8599,
+      "step": 11442
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7020893798010983,
+      "learning_rate": 1.763313746265715e-06,
+      "loss": 0.6804,
+      "step": 11443
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5657459761785846,
+      "learning_rate": 1.7625238014464358e-06,
+      "loss": 0.7139,
+      "step": 11444
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8609415339233353,
+      "learning_rate": 1.7617339957487167e-06,
+      "loss": 0.6942,
+      "step": 11445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5938170337924547,
+      "learning_rate": 1.760944329206498e-06,
+      "loss": 0.6836,
+      "step": 11446
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6933496338491327,
+      "learning_rate": 1.760154801853719e-06,
+      "loss": 0.6957,
+      "step": 11447
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7077939868246312,
+      "learning_rate": 1.7593654137243005e-06,
+      "loss": 0.7432,
+      "step": 11448
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5266886200813814,
+      "learning_rate": 1.7585761648521688e-06,
+      "loss": 0.6245,
+      "step": 11449
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1182585969804566,
+      "learning_rate": 1.7577870552712378e-06,
+      "loss": 0.7435,
+      "step": 11450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9757502158220936,
+      "learning_rate": 1.756998085015419e-06,
+      "loss": 0.7842,
+      "step": 11451
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5999443749316047,
+      "learning_rate": 1.7562092541186144e-06,
+      "loss": 0.6428,
+      "step": 11452
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6903574424403107,
+      "learning_rate": 1.7554205626147202e-06,
+      "loss": 0.7136,
+      "step": 11453
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7303187773468618,
+      "learning_rate": 1.7546320105376346e-06,
+      "loss": 0.6406,
+      "step": 11454
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7431567132922219,
+      "learning_rate": 1.7538435979212376e-06,
+      "loss": 0.7805,
+      "step": 11455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4894890554495854,
+      "learning_rate": 1.7530553247994093e-06,
+      "loss": 0.6566,
+      "step": 11456
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.127706048195909,
+      "learning_rate": 1.7522671912060262e-06,
+      "loss": 0.687,
+      "step": 11457
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.552288786975181,
+      "learning_rate": 1.751479197174955e-06,
+      "loss": 0.7349,
+      "step": 11458
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6405767137565612,
+      "learning_rate": 1.750691342740058e-06,
+      "loss": 0.72,
+      "step": 11459
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.007193354698157,
+      "learning_rate": 1.7499036279351905e-06,
+      "loss": 0.819,
+      "step": 11460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2511152212064816,
+      "learning_rate": 1.7491160527942002e-06,
+      "loss": 0.8452,
+      "step": 11461
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.676973848614407,
+      "learning_rate": 1.7483286173509372e-06,
+      "loss": 0.8094,
+      "step": 11462
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6085732540235755,
+      "learning_rate": 1.7475413216392317e-06,
+      "loss": 0.6765,
+      "step": 11463
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0000033141923503,
+      "learning_rate": 1.746754165692921e-06,
+      "loss": 0.6915,
+      "step": 11464
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.72540804367352,
+      "learning_rate": 1.7459671495458308e-06,
+      "loss": 0.8925,
+      "step": 11465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8336879012289709,
+      "learning_rate": 1.7451802732317763e-06,
+      "loss": 0.7092,
+      "step": 11466
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7170938936632802,
+      "learning_rate": 1.7443935367845755e-06,
+      "loss": 0.8228,
+      "step": 11467
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6474377580927309,
+      "learning_rate": 1.7436069402380352e-06,
+      "loss": 0.6395,
+      "step": 11468
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8529813803839184,
+      "learning_rate": 1.742820483625957e-06,
+      "loss": 0.7403,
+      "step": 11469
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9657522113510812,
+      "learning_rate": 1.7420341669821373e-06,
+      "loss": 0.6943,
+      "step": 11470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5401038895977752,
+      "learning_rate": 1.7412479903403628e-06,
+      "loss": 0.7127,
+      "step": 11471
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8472311439352984,
+      "learning_rate": 1.7404619537344236e-06,
+      "loss": 0.6602,
+      "step": 11472
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.143047910491523,
+      "learning_rate": 1.7396760571980902e-06,
+      "loss": 0.5839,
+      "step": 11473
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7682724906372693,
+      "learning_rate": 1.7388903007651398e-06,
+      "loss": 0.6914,
+      "step": 11474
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6345880515510778,
+      "learning_rate": 1.7381046844693356e-06,
+      "loss": 0.8332,
+      "step": 11475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5955861267049773,
+      "learning_rate": 1.737319208344438e-06,
+      "loss": 0.6666,
+      "step": 11476
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.056853037642628,
+      "learning_rate": 1.7365338724242003e-06,
+      "loss": 0.6993,
+      "step": 11477
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1592604188884115,
+      "learning_rate": 1.7357486767423699e-06,
+      "loss": 0.7016,
+      "step": 11478
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7339307395050816,
+      "learning_rate": 1.7349636213326876e-06,
+      "loss": 0.6579,
+      "step": 11479
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1304955194795163,
+      "learning_rate": 1.7341787062288928e-06,
+      "loss": 0.729,
+      "step": 11480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8636301634477397,
+      "learning_rate": 1.7333939314647097e-06,
+      "loss": 0.7105,
+      "step": 11481
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7116374329657176,
+      "learning_rate": 1.7326092970738656e-06,
+      "loss": 0.6505,
+      "step": 11482
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.690006590716441,
+      "learning_rate": 1.7318248030900792e-06,
+      "loss": 0.7218,
+      "step": 11483
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8813489346799055,
+      "learning_rate": 1.7310404495470557e-06,
+      "loss": 0.6684,
+      "step": 11484
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7700397895248274,
+      "learning_rate": 1.7302562364785063e-06,
+      "loss": 0.6648,
+      "step": 11485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.93531794999445,
+      "learning_rate": 1.729472163918129e-06,
+      "loss": 0.6364,
+      "step": 11486
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.058093683596956,
+      "learning_rate": 1.7286882318996162e-06,
+      "loss": 0.6597,
+      "step": 11487
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8015805611619045,
+      "learning_rate": 1.7279044404566554e-06,
+      "loss": 0.7472,
+      "step": 11488
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.504533227397424,
+      "learning_rate": 1.727120789622927e-06,
+      "loss": 0.8626,
+      "step": 11489
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.81616507175443,
+      "learning_rate": 1.7263372794321104e-06,
+      "loss": 0.7101,
+      "step": 11490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7321339823453052,
+      "learning_rate": 1.7255539099178687e-06,
+      "loss": 0.723,
+      "step": 11491
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.854695000200052,
+      "learning_rate": 1.7247706811138692e-06,
+      "loss": 0.6414,
+      "step": 11492
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.702679891151684,
+      "learning_rate": 1.723987593053768e-06,
+      "loss": 0.6525,
+      "step": 11493
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5731926711005542,
+      "learning_rate": 1.7232046457712164e-06,
+      "loss": 0.7033,
+      "step": 11494
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5720687649395826,
+      "learning_rate": 1.7224218392998587e-06,
+      "loss": 0.6539,
+      "step": 11495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0188945843757993,
+      "learning_rate": 1.7216391736733345e-06,
+      "loss": 0.6927,
+      "step": 11496
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7824921099101643,
+      "learning_rate": 1.7208566489252742e-06,
+      "loss": 0.7341,
+      "step": 11497
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2661367744641205,
+      "learning_rate": 1.7200742650893105e-06,
+      "loss": 0.6169,
+      "step": 11498
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7117243848645438,
+      "learning_rate": 1.7192920221990566e-06,
+      "loss": 0.7736,
+      "step": 11499
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6509408374279562,
+      "learning_rate": 1.7185099202881328e-06,
+      "loss": 0.71,
+      "step": 11500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.132116580024883,
+      "learning_rate": 1.7177279593901463e-06,
+      "loss": 0.7197,
+      "step": 11501
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1910502019252243,
+      "learning_rate": 1.7169461395386988e-06,
+      "loss": 0.6508,
+      "step": 11502
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7838827775187838,
+      "learning_rate": 1.7161644607673884e-06,
+      "loss": 0.6776,
+      "step": 11503
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.831689922403292,
+      "learning_rate": 1.7153829231098018e-06,
+      "loss": 0.7637,
+      "step": 11504
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1346287370048247,
+      "learning_rate": 1.7146015265995302e-06,
+      "loss": 0.5399,
+      "step": 11505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7731296192529473,
+      "learning_rate": 1.7138202712701463e-06,
+      "loss": 0.7854,
+      "step": 11506
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5174450082837856,
+      "learning_rate": 1.7130391571552224e-06,
+      "loss": 0.6849,
+      "step": 11507
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.492389619937026,
+      "learning_rate": 1.712258184288328e-06,
+      "loss": 0.7555,
+      "step": 11508
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2047813730412171,
+      "learning_rate": 1.7114773527030215e-06,
+      "loss": 0.6826,
+      "step": 11509
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6860855776103973,
+      "learning_rate": 1.7106966624328575e-06,
+      "loss": 0.6872,
+      "step": 11510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5450564999336993,
+      "learning_rate": 1.7099161135113828e-06,
+      "loss": 0.7417,
+      "step": 11511
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6960491427844446,
+      "learning_rate": 1.7091357059721409e-06,
+      "loss": 0.7486,
+      "step": 11512
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7610223364761637,
+      "learning_rate": 1.7083554398486663e-06,
+      "loss": 0.6928,
+      "step": 11513
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0759390616571323,
+      "learning_rate": 1.7075753151744885e-06,
+      "loss": 0.7949,
+      "step": 11514
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.713812668779437,
+      "learning_rate": 1.7067953319831327e-06,
+      "loss": 0.7506,
+      "step": 11515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.516670904325601,
+      "learning_rate": 1.7060154903081177e-06,
+      "loss": 0.6863,
+      "step": 11516
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.116145106642938,
+      "learning_rate": 1.7052357901829502e-06,
+      "loss": 0.6635,
+      "step": 11517
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.680842448087354,
+      "learning_rate": 1.70445623164114e-06,
+      "loss": 0.6815,
+      "step": 11518
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6379165840693621,
+      "learning_rate": 1.7036768147161853e-06,
+      "loss": 0.7408,
+      "step": 11519
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.591455013478093,
+      "learning_rate": 1.7028975394415791e-06,
+      "loss": 0.7842,
+      "step": 11520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4930239342275922,
+      "learning_rate": 1.702118405850809e-06,
+      "loss": 0.6958,
+      "step": 11521
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8596785161305274,
+      "learning_rate": 1.7013394139773537e-06,
+      "loss": 0.7288,
+      "step": 11522
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0639560547455993,
+      "learning_rate": 1.700560563854694e-06,
+      "loss": 0.7556,
+      "step": 11523
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0598206086201918,
+      "learning_rate": 1.6997818555162915e-06,
+      "loss": 0.7009,
+      "step": 11524
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6888390675706073,
+      "learning_rate": 1.6990032889956142e-06,
+      "loss": 0.7188,
+      "step": 11525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9734092330608406,
+      "learning_rate": 1.6982248643261173e-06,
+      "loss": 0.8476,
+      "step": 11526
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.071489204784312,
+      "learning_rate": 1.6974465815412517e-06,
+      "loss": 0.6181,
+      "step": 11527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7486772868153575,
+      "learning_rate": 1.6966684406744615e-06,
+      "loss": 0.6139,
+      "step": 11528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.885266794719367,
+      "learning_rate": 1.6958904417591853e-06,
+      "loss": 0.7936,
+      "step": 11529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9755635693562905,
+      "learning_rate": 1.6951125848288558e-06,
+      "loss": 0.6586,
+      "step": 11530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.419041169219675,
+      "learning_rate": 1.6943348699168981e-06,
+      "loss": 0.805,
+      "step": 11531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7069056311255464,
+      "learning_rate": 1.6935572970567321e-06,
+      "loss": 0.6848,
+      "step": 11532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5336656492844802,
+      "learning_rate": 1.6927798662817745e-06,
+      "loss": 0.7018,
+      "step": 11533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.681431855108199,
+      "learning_rate": 1.6920025776254334e-06,
+      "loss": 0.6992,
+      "step": 11534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0806045557649138,
+      "learning_rate": 1.6912254311211052e-06,
+      "loss": 0.7579,
+      "step": 11535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.782769468876289,
+      "learning_rate": 1.6904484268021915e-06,
+      "loss": 0.6925,
+      "step": 11536
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5311911875858795,
+      "learning_rate": 1.689671564702079e-06,
+      "loss": 0.7355,
+      "step": 11537
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7466234493411048,
+      "learning_rate": 1.688894844854152e-06,
+      "loss": 0.7027,
+      "step": 11538
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8425702136897337,
+      "learning_rate": 1.6881182672917879e-06,
+      "loss": 0.6627,
+      "step": 11539
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2736530919825986,
+      "learning_rate": 1.6873418320483565e-06,
+      "loss": 0.622,
+      "step": 11540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6715042431908933,
+      "learning_rate": 1.6865655391572266e-06,
+      "loss": 0.5921,
+      "step": 11541
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1462930585322477,
+      "learning_rate": 1.6857893886517523e-06,
+      "loss": 0.5616,
+      "step": 11542
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8583814645321115,
+      "learning_rate": 1.6850133805652907e-06,
+      "loss": 0.6273,
+      "step": 11543
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6976207761466988,
+      "learning_rate": 1.6842375149311868e-06,
+      "loss": 0.7337,
+      "step": 11544
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4664068930567704,
+      "learning_rate": 1.6834617917827817e-06,
+      "loss": 0.8166,
+      "step": 11545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.530435365726149,
+      "learning_rate": 1.6826862111534098e-06,
+      "loss": 0.7756,
+      "step": 11546
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6794210208342117,
+      "learning_rate": 1.6819107730763996e-06,
+      "loss": 0.6642,
+      "step": 11547
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6293020148654425,
+      "learning_rate": 1.681135477585073e-06,
+      "loss": 0.6403,
+      "step": 11548
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7766846665871994,
+      "learning_rate": 1.680360324712746e-06,
+      "loss": 0.7304,
+      "step": 11549
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.793995382776511,
+      "learning_rate": 1.6795853144927282e-06,
+      "loss": 0.6598,
+      "step": 11550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8270863785543368,
+      "learning_rate": 1.6788104469583256e-06,
+      "loss": 0.7686,
+      "step": 11551
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6383868198466134,
+      "learning_rate": 1.678035722142835e-06,
+      "loss": 0.6797,
+      "step": 11552
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.819406335284538,
+      "learning_rate": 1.6772611400795475e-06,
+      "loss": 0.7095,
+      "step": 11553
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.025443464509837,
+      "learning_rate": 1.6764867008017493e-06,
+      "loss": 0.8224,
+      "step": 11554
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9714524616952263,
+      "learning_rate": 1.6757124043427193e-06,
+      "loss": 0.73,
+      "step": 11555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7502596578789695,
+      "learning_rate": 1.674938250735731e-06,
+      "loss": 0.8025,
+      "step": 11556
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.04690410466406,
+      "learning_rate": 1.6741642400140513e-06,
+      "loss": 0.6698,
+      "step": 11557
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0246888746964853,
+      "learning_rate": 1.6733903722109391e-06,
+      "loss": 0.7486,
+      "step": 11558
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4094418261330963,
+      "learning_rate": 1.672616647359655e-06,
+      "loss": 0.6886,
+      "step": 11559
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6388887490951574,
+      "learning_rate": 1.6718430654934408e-06,
+      "loss": 0.8221,
+      "step": 11560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6314176138408532,
+      "learning_rate": 1.671069626645544e-06,
+      "loss": 0.7921,
+      "step": 11561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.735547469173329,
+      "learning_rate": 1.6702963308491994e-06,
+      "loss": 0.6492,
+      "step": 11562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5103851056352473,
+      "learning_rate": 1.669523178137637e-06,
+      "loss": 0.7363,
+      "step": 11563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7409137616083064,
+      "learning_rate": 1.668750168544081e-06,
+      "loss": 0.6242,
+      "step": 11564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6921500108504643,
+      "learning_rate": 1.6679773021017475e-06,
+      "loss": 0.6952,
+      "step": 11565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.052987394538072,
+      "learning_rate": 1.6672045788438534e-06,
+      "loss": 0.602,
+      "step": 11566
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7995040218630436,
+      "learning_rate": 1.6664319988036e-06,
+      "loss": 0.5464,
+      "step": 11567
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6865778003020926,
+      "learning_rate": 1.6656595620141857e-06,
+      "loss": 0.6872,
+      "step": 11568
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6714487774857598,
+      "learning_rate": 1.664887268508808e-06,
+      "loss": 0.6809,
+      "step": 11569
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9400280926291797,
+      "learning_rate": 1.6641151183206516e-06,
+      "loss": 0.7328,
+      "step": 11570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7110076203482896,
+      "learning_rate": 1.663343111482898e-06,
+      "loss": 0.7009,
+      "step": 11571
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.663504222964053,
+      "learning_rate": 1.6625712480287225e-06,
+      "loss": 0.6522,
+      "step": 11572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6912615709085836,
+      "learning_rate": 1.661799527991293e-06,
+      "loss": 0.7739,
+      "step": 11573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.474198270876817,
+      "learning_rate": 1.6610279514037725e-06,
+      "loss": 0.7083,
+      "step": 11574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.549864445393239,
+      "learning_rate": 1.6602565182993152e-06,
+      "loss": 0.753,
+      "step": 11575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6353073401609244,
+      "learning_rate": 1.6594852287110758e-06,
+      "loss": 0.7446,
+      "step": 11576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9249858394282837,
+      "learning_rate": 1.6587140826721964e-06,
+      "loss": 0.6709,
+      "step": 11577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8238257058601384,
+      "learning_rate": 1.657943080215812e-06,
+      "loss": 0.7556,
+      "step": 11578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6329459834540163,
+      "learning_rate": 1.657172221375058e-06,
+      "loss": 0.7111,
+      "step": 11579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.797824078667247,
+      "learning_rate": 1.656401506183058e-06,
+      "loss": 0.6502,
+      "step": 11580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9985817053063335,
+      "learning_rate": 1.6556309346729322e-06,
+      "loss": 0.71,
+      "step": 11581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7923569710014204,
+      "learning_rate": 1.6548605068777935e-06,
+      "loss": 0.7122,
+      "step": 11582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8627781111203525,
+      "learning_rate": 1.6540902228307471e-06,
+      "loss": 0.6703,
+      "step": 11583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9847809033160773,
+      "learning_rate": 1.6533200825648993e-06,
+      "loss": 0.7659,
+      "step": 11584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5926585217094216,
+      "learning_rate": 1.6525500861133386e-06,
+      "loss": 0.7561,
+      "step": 11585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7988807836058047,
+      "learning_rate": 1.6517802335091549e-06,
+      "loss": 0.7451,
+      "step": 11586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.259128313576922,
+      "learning_rate": 1.6510105247854325e-06,
+      "loss": 0.6709,
+      "step": 11587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6214953436900241,
+      "learning_rate": 1.6502409599752468e-06,
+      "loss": 0.7337,
+      "step": 11588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8386930972734334,
+      "learning_rate": 1.6494715391116671e-06,
+      "loss": 0.6634,
+      "step": 11589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4620318734054636,
+      "learning_rate": 1.6487022622277582e-06,
+      "loss": 0.7047,
+      "step": 11590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.909314190953337,
+      "learning_rate": 1.6479331293565758e-06,
+      "loss": 0.6777,
+      "step": 11591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.146780492391168,
+      "learning_rate": 1.6471641405311727e-06,
+      "loss": 0.8809,
+      "step": 11592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.686903947844292,
+      "learning_rate": 1.6463952957845924e-06,
+      "loss": 0.6529,
+      "step": 11593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6332690036270479,
+      "learning_rate": 1.6456265951498763e-06,
+      "loss": 0.7203,
+      "step": 11594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8012493965316219,
+      "learning_rate": 1.6448580386600576e-06,
+      "loss": 0.8478,
+      "step": 11595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.892091299215211,
+      "learning_rate": 1.6440896263481582e-06,
+      "loss": 0.8348,
+      "step": 11596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7300816311383036,
+      "learning_rate": 1.6433213582472036e-06,
+      "loss": 0.8204,
+      "step": 11597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6197444239822436,
+      "learning_rate": 1.6425532343902051e-06,
+      "loss": 0.6958,
+      "step": 11598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5957234266684241,
+      "learning_rate": 1.641785254810172e-06,
+      "loss": 0.6269,
+      "step": 11599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8287027644674216,
+      "learning_rate": 1.6410174195401057e-06,
+      "loss": 0.7264,
+      "step": 11600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6570867199278987,
+      "learning_rate": 1.6402497286129998e-06,
+      "loss": 0.7058,
+      "step": 11601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6906653528645643,
+      "learning_rate": 1.6394821820618496e-06,
+      "loss": 0.6709,
+      "step": 11602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.85832690922226,
+      "learning_rate": 1.63871477991963e-06,
+      "loss": 0.7601,
+      "step": 11603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1678756145625235,
+      "learning_rate": 1.6379475222193248e-06,
+      "loss": 0.7175,
+      "step": 11604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8066413545032234,
+      "learning_rate": 1.637180408993902e-06,
+      "loss": 0.7275,
+      "step": 11605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9770228102865208,
+      "learning_rate": 1.636413440276326e-06,
+      "loss": 0.6344,
+      "step": 11606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9154309303690773,
+      "learning_rate": 1.635646616099556e-06,
+      "loss": 0.6934,
+      "step": 11607
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7312120500166297,
+      "learning_rate": 1.634879936496544e-06,
+      "loss": 0.7935,
+      "step": 11608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8648255403345597,
+      "learning_rate": 1.6341134015002352e-06,
+      "loss": 0.7972,
+      "step": 11609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.541792208855032,
+      "learning_rate": 1.63334701114357e-06,
+      "loss": 0.6148,
+      "step": 11610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7372172039625189,
+      "learning_rate": 1.63258076545948e-06,
+      "loss": 0.6326,
+      "step": 11611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1390755082217852,
+      "learning_rate": 1.6318146644808952e-06,
+      "loss": 0.6404,
+      "step": 11612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1657658432129785,
+      "learning_rate": 1.631048708240736e-06,
+      "loss": 0.6976,
+      "step": 11613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7787871528365502,
+      "learning_rate": 1.6302828967719175e-06,
+      "loss": 0.6686,
+      "step": 11614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.461306282115623,
+      "learning_rate": 1.6295172301073469e-06,
+      "loss": 0.6752,
+      "step": 11615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7830303069398514,
+      "learning_rate": 1.6287517082799265e-06,
+      "loss": 0.763,
+      "step": 11616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7687284911142158,
+      "learning_rate": 1.6279863313225559e-06,
+      "loss": 0.7333,
+      "step": 11617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7712364276914438,
+      "learning_rate": 1.6272210992681213e-06,
+      "loss": 0.6654,
+      "step": 11618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.568644715733486,
+      "learning_rate": 1.626456012149506e-06,
+      "loss": 0.7338,
+      "step": 11619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9733882101335998,
+      "learning_rate": 1.6256910699995921e-06,
+      "loss": 0.7387,
+      "step": 11620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.477102646914293,
+      "learning_rate": 1.624926272851245e-06,
+      "loss": 0.671,
+      "step": 11621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9623257327445975,
+      "learning_rate": 1.6241616207373339e-06,
+      "loss": 0.7154,
+      "step": 11622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6700904126604321,
+      "learning_rate": 1.6233971136907168e-06,
+      "loss": 0.6509,
+      "step": 11623
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5281723588998462,
+      "learning_rate": 1.6226327517442453e-06,
+      "loss": 0.6453,
+      "step": 11624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0332767426415566,
+      "learning_rate": 1.6218685349307667e-06,
+      "loss": 0.7686,
+      "step": 11625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7251550821243533,
+      "learning_rate": 1.6211044632831185e-06,
+      "loss": 0.7901,
+      "step": 11626
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6535970154965873,
+      "learning_rate": 1.620340536834139e-06,
+      "loss": 0.7588,
+      "step": 11627
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8073801885326113,
+      "learning_rate": 1.6195767556166548e-06,
+      "loss": 0.6834,
+      "step": 11628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6141095160968815,
+      "learning_rate": 1.6188131196634827e-06,
+      "loss": 0.6391,
+      "step": 11629
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5511082539425036,
+      "learning_rate": 1.6180496290074432e-06,
+      "loss": 0.7816,
+      "step": 11630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7159295457229666,
+      "learning_rate": 1.617286283681343e-06,
+      "loss": 0.6905,
+      "step": 11631
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7733762359220004,
+      "learning_rate": 1.6165230837179846e-06,
+      "loss": 0.7072,
+      "step": 11632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2298061786936665,
+      "learning_rate": 1.6157600291501646e-06,
+      "loss": 0.6158,
+      "step": 11633
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7875290542728368,
+      "learning_rate": 1.6149971200106723e-06,
+      "loss": 0.6978,
+      "step": 11634
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8237087134679397,
+      "learning_rate": 1.6142343563322955e-06,
+      "loss": 0.7433,
+      "step": 11635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2426909400301622,
+      "learning_rate": 1.613471738147807e-06,
+      "loss": 0.6291,
+      "step": 11636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.663494699224434,
+      "learning_rate": 1.6127092654899784e-06,
+      "loss": 0.6647,
+      "step": 11637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1249219287454764,
+      "learning_rate": 1.6119469383915804e-06,
+      "loss": 0.6771,
+      "step": 11638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.201273962782296,
+      "learning_rate": 1.6111847568853645e-06,
+      "loss": 0.6197,
+      "step": 11639
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.71030633330625,
+      "learning_rate": 1.6104227210040884e-06,
+      "loss": 0.7896,
+      "step": 11640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8298888514772107,
+      "learning_rate": 1.6096608307804973e-06,
+      "loss": 0.7716,
+      "step": 11641
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6709477378224766,
+      "learning_rate": 1.608899086247331e-06,
+      "loss": 0.7103,
+      "step": 11642
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9278902934643176,
+      "learning_rate": 1.608137487437323e-06,
+      "loss": 0.7344,
+      "step": 11643
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7229101393872883,
+      "learning_rate": 1.6073760343831996e-06,
+      "loss": 0.7439,
+      "step": 11644
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2431124672029976,
+      "learning_rate": 1.6066147271176857e-06,
+      "loss": 0.7148,
+      "step": 11645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6749498604474735,
+      "learning_rate": 1.6058535656734958e-06,
+      "loss": 0.6543,
+      "step": 11646
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6735807898243038,
+      "learning_rate": 1.6050925500833342e-06,
+      "loss": 0.8276,
+      "step": 11647
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5772584927805195,
+      "learning_rate": 1.604331680379908e-06,
+      "loss": 0.6658,
+      "step": 11648
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0214731161335564,
+      "learning_rate": 1.603570956595913e-06,
+      "loss": 0.6998,
+      "step": 11649
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7771952308848722,
+      "learning_rate": 1.602810378764037e-06,
+      "loss": 0.7002,
+      "step": 11650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.771136269106022,
+      "learning_rate": 1.6020499469169659e-06,
+      "loss": 0.6658,
+      "step": 11651
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6282498897881257,
+      "learning_rate": 1.601289661087374e-06,
+      "loss": 0.773,
+      "step": 11652
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5471272291854345,
+      "learning_rate": 1.600529521307938e-06,
+      "loss": 0.7994,
+      "step": 11653
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0827615623999947,
+      "learning_rate": 1.5997695276113168e-06,
+      "loss": 0.6613,
+      "step": 11654
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9300434141054952,
+      "learning_rate": 1.599009680030173e-06,
+      "loss": 0.7844,
+      "step": 11655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7350432597657492,
+      "learning_rate": 1.598249978597159e-06,
+      "loss": 0.8021,
+      "step": 11656
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9698436261376608,
+      "learning_rate": 1.5974904233449167e-06,
+      "loss": 0.6847,
+      "step": 11657
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7628449046038046,
+      "learning_rate": 1.59673101430609e-06,
+      "loss": 0.7445,
+      "step": 11658
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8700912028728536,
+      "learning_rate": 1.595971751513311e-06,
+      "loss": 0.6352,
+      "step": 11659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8735716923839725,
+      "learning_rate": 1.5952126349992076e-06,
+      "loss": 0.6959,
+      "step": 11660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.938589629302944,
+      "learning_rate": 1.5944536647963998e-06,
+      "loss": 0.5623,
+      "step": 11661
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.852095750577303,
+      "learning_rate": 1.5936948409375007e-06,
+      "loss": 0.8079,
+      "step": 11662
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5112765319903618,
+      "learning_rate": 1.5929361634551227e-06,
+      "loss": 0.7089,
+      "step": 11663
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4664881331340358,
+      "learning_rate": 1.5921776323818655e-06,
+      "loss": 0.6384,
+      "step": 11664
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1109362830778635,
+      "learning_rate": 1.5914192477503248e-06,
+      "loss": 0.5908,
+      "step": 11665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.591677287020695,
+      "learning_rate": 1.5906610095930908e-06,
+      "loss": 0.71,
+      "step": 11666
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5354999681906845,
+      "learning_rate": 1.589902917942746e-06,
+      "loss": 0.6854,
+      "step": 11667
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7403472657703862,
+      "learning_rate": 1.5891449728318675e-06,
+      "loss": 0.6832,
+      "step": 11668
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8059329727321771,
+      "learning_rate": 1.5883871742930257e-06,
+      "loss": 0.6079,
+      "step": 11669
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.903601541498295,
+      "learning_rate": 1.5876295223587835e-06,
+      "loss": 0.7117,
+      "step": 11670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1105019310231103,
+      "learning_rate": 1.586872017061704e-06,
+      "loss": 0.7794,
+      "step": 11671
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6455150701547323,
+      "learning_rate": 1.5861146584343318e-06,
+      "loss": 0.7025,
+      "step": 11672
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.355273601467168,
+      "learning_rate": 1.585357446509218e-06,
+      "loss": 0.6975,
+      "step": 11673
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9243719590567585,
+      "learning_rate": 1.5846003813188993e-06,
+      "loss": 0.6714,
+      "step": 11674
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6759089770080988,
+      "learning_rate": 1.5838434628959081e-06,
+      "loss": 0.7535,
+      "step": 11675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1210740814601878,
+      "learning_rate": 1.5830866912727722e-06,
+      "loss": 0.6468,
+      "step": 11676
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8033962582792746,
+      "learning_rate": 1.582330066482009e-06,
+      "loss": 0.6086,
+      "step": 11677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7114933025314285,
+      "learning_rate": 1.5815735885561385e-06,
+      "loss": 0.6657,
+      "step": 11678
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.051778557748685,
+      "learning_rate": 1.5808172575276615e-06,
+      "loss": 0.767,
+      "step": 11679
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.850372843631604,
+      "learning_rate": 1.5800610734290805e-06,
+      "loss": 0.7247,
+      "step": 11680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6120816642918288,
+      "learning_rate": 1.5793050362928935e-06,
+      "loss": 0.6946,
+      "step": 11681
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.406739121553296,
+      "learning_rate": 1.5785491461515878e-06,
+      "loss": 0.7422,
+      "step": 11682
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.679306935277525,
+      "learning_rate": 1.5777934030376445e-06,
+      "loss": 0.7001,
+      "step": 11683
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8943004378863406,
+      "learning_rate": 1.5770378069835412e-06,
+      "loss": 0.7427,
+      "step": 11684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.70329437641159,
+      "learning_rate": 1.5762823580217463e-06,
+      "loss": 0.7345,
+      "step": 11685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.717195552295003,
+      "learning_rate": 1.5755270561847236e-06,
+      "loss": 0.7407,
+      "step": 11686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8964209497117621,
+      "learning_rate": 1.5747719015049294e-06,
+      "loss": 0.7197,
+      "step": 11687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7901669767772117,
+      "learning_rate": 1.5740168940148137e-06,
+      "loss": 0.7938,
+      "step": 11688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3637787382048843,
+      "learning_rate": 1.5732620337468258e-06,
+      "loss": 0.6577,
+      "step": 11689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0628262675829856,
+      "learning_rate": 1.5725073207333963e-06,
+      "loss": 0.6274,
+      "step": 11690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0810490056816242,
+      "learning_rate": 1.5717527550069628e-06,
+      "loss": 0.5794,
+      "step": 11691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0205620423178374,
+      "learning_rate": 1.5709983365999481e-06,
+      "loss": 0.6208,
+      "step": 11692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.565416589010137,
+      "learning_rate": 1.570244065544772e-06,
+      "loss": 0.6511,
+      "step": 11693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.696369615448815,
+      "learning_rate": 1.5694899418738462e-06,
+      "loss": 0.7351,
+      "step": 11694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5954636390097454,
+      "learning_rate": 1.568735965619577e-06,
+      "loss": 0.6765,
+      "step": 11695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.624599647749384,
+      "learning_rate": 1.5679821368143678e-06,
+      "loss": 0.7004,
+      "step": 11696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.126188654928838,
+      "learning_rate": 1.5672284554906087e-06,
+      "loss": 0.7023,
+      "step": 11697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4128760401428644,
+      "learning_rate": 1.5664749216806858e-06,
+      "loss": 0.7873,
+      "step": 11698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.879057479029206,
+      "learning_rate": 1.5657215354169841e-06,
+      "loss": 0.6519,
+      "step": 11699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1295730954693397,
+      "learning_rate": 1.564968296731877e-06,
+      "loss": 0.6281,
+      "step": 11700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.889960927939516,
+      "learning_rate": 1.564215205657732e-06,
+      "loss": 0.6556,
+      "step": 11701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0559966036625883,
+      "learning_rate": 1.563462262226912e-06,
+      "loss": 0.6187,
+      "step": 11702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2253073990671408,
+      "learning_rate": 1.5627094664717719e-06,
+      "loss": 0.6792,
+      "step": 11703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7694561535906113,
+      "learning_rate": 1.561956818424661e-06,
+      "loss": 0.7323,
+      "step": 11704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.925614907407532,
+      "learning_rate": 1.561204318117921e-06,
+      "loss": 0.7678,
+      "step": 11705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6976884743603446,
+      "learning_rate": 1.5604519655838918e-06,
+      "loss": 0.6403,
+      "step": 11706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.601326842151683,
+      "learning_rate": 1.5596997608549036e-06,
+      "loss": 0.7056,
+      "step": 11707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.515820735655555,
+      "learning_rate": 1.5589477039632755e-06,
+      "loss": 0.7235,
+      "step": 11708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8171880631260664,
+      "learning_rate": 1.5581957949413295e-06,
+      "loss": 0.6364,
+      "step": 11709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7369667518755285,
+      "learning_rate": 1.557444033821376e-06,
+      "loss": 0.6791,
+      "step": 11710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8764555311717992,
+      "learning_rate": 1.5566924206357187e-06,
+      "loss": 0.7702,
+      "step": 11711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5682913026194607,
+      "learning_rate": 1.5559409554166572e-06,
+      "loss": 0.673,
+      "step": 11712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.008179734725648,
+      "learning_rate": 1.5551896381964814e-06,
+      "loss": 0.7306,
+      "step": 11713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5676035690135723,
+      "learning_rate": 1.554438469007482e-06,
+      "loss": 0.8055,
+      "step": 11714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7416499738219375,
+      "learning_rate": 1.5536874478819325e-06,
+      "loss": 0.7456,
+      "step": 11715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6820477116894275,
+      "learning_rate": 1.5529365748521097e-06,
+      "loss": 0.6858,
+      "step": 11716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6330349802669306,
+      "learning_rate": 1.5521858499502794e-06,
+      "loss": 0.6639,
+      "step": 11717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5058213685944974,
+      "learning_rate": 1.5514352732087024e-06,
+      "loss": 0.6804,
+      "step": 11718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6489144413744556,
+      "learning_rate": 1.5506848446596317e-06,
+      "loss": 0.767,
+      "step": 11719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5435138566055782,
+      "learning_rate": 1.5499345643353154e-06,
+      "loss": 0.6796,
+      "step": 11720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1323752449161533,
+      "learning_rate": 1.5491844322679944e-06,
+      "loss": 0.6158,
+      "step": 11721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7494096252284461,
+      "learning_rate": 1.5484344484899038e-06,
+      "loss": 0.6344,
+      "step": 11722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7200073423731739,
+      "learning_rate": 1.5476846130332702e-06,
+      "loss": 0.7351,
+      "step": 11723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4489664118299768,
+      "learning_rate": 1.546934925930319e-06,
+      "loss": 0.6069,
+      "step": 11724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.625575810583876,
+      "learning_rate": 1.5461853872132648e-06,
+      "loss": 0.6322,
+      "step": 11725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7259917830539815,
+      "learning_rate": 1.5454359969143161e-06,
+      "loss": 0.8051,
+      "step": 11726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0035148307573847,
+      "learning_rate": 1.544686755065677e-06,
+      "loss": 0.7728,
+      "step": 11727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.298635289492555,
+      "learning_rate": 1.5439376616995432e-06,
+      "loss": 0.7397,
+      "step": 11728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1860873884250602,
+      "learning_rate": 1.5431887168481051e-06,
+      "loss": 0.6837,
+      "step": 11729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.120108040770344,
+      "learning_rate": 1.5424399205435464e-06,
+      "loss": 0.6615,
+      "step": 11730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.720300472894294,
+      "learning_rate": 1.5416912728180433e-06,
+      "loss": 0.7077,
+      "step": 11731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.713938567051576,
+      "learning_rate": 1.5409427737037713e-06,
+      "loss": 0.6999,
+      "step": 11732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.570999371013059,
+      "learning_rate": 1.540194423232889e-06,
+      "loss": 0.7955,
+      "step": 11733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0950429050085706,
+      "learning_rate": 1.5394462214375593e-06,
+      "loss": 0.667,
+      "step": 11734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7480751042860427,
+      "learning_rate": 1.5386981683499325e-06,
+      "loss": 0.7178,
+      "step": 11735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5396805995081277,
+      "learning_rate": 1.5379502640021544e-06,
+      "loss": 0.6527,
+      "step": 11736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5463972279837856,
+      "learning_rate": 1.537202508426363e-06,
+      "loss": 0.6841,
+      "step": 11737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.860081005470229,
+      "learning_rate": 1.5364549016546925e-06,
+      "loss": 0.7308,
+      "step": 11738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6809955568922335,
+      "learning_rate": 1.5357074437192688e-06,
+      "loss": 0.8356,
+      "step": 11739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2009269712190727,
+      "learning_rate": 1.5349601346522115e-06,
+      "loss": 0.5996,
+      "step": 11740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.731586020569824,
+      "learning_rate": 1.5342129744856327e-06,
+      "loss": 0.7807,
+      "step": 11741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0679654671745757,
+      "learning_rate": 1.5334659632516424e-06,
+      "loss": 0.7491,
+      "step": 11742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6501046647969584,
+      "learning_rate": 1.5327191009823405e-06,
+      "loss": 0.7133,
+      "step": 11743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5967182371855255,
+      "learning_rate": 1.5319723877098202e-06,
+      "loss": 0.8057,
+      "step": 11744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.542961795907747,
+      "learning_rate": 1.531225823466171e-06,
+      "loss": 0.6593,
+      "step": 11745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6681356759915027,
+      "learning_rate": 1.5304794082834713e-06,
+      "loss": 0.6707,
+      "step": 11746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.622823191846691,
+      "learning_rate": 1.5297331421938017e-06,
+      "loss": 0.7015,
+      "step": 11747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0891589667156514,
+      "learning_rate": 1.5289870252292265e-06,
+      "loss": 0.5449,
+      "step": 11748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.878344777634884,
+      "learning_rate": 1.5282410574218072e-06,
+      "loss": 0.63,
+      "step": 11749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1043761057685293,
+      "learning_rate": 1.5274952388036052e-06,
+      "loss": 0.6469,
+      "step": 11750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9144373879278673,
+      "learning_rate": 1.5267495694066631e-06,
+      "loss": 0.711,
+      "step": 11751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.882356311487524,
+      "learning_rate": 1.5260040492630286e-06,
+      "loss": 0.7846,
+      "step": 11752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.626568263408783,
+      "learning_rate": 1.5252586784047374e-06,
+      "loss": 0.7892,
+      "step": 11753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0773819990246647,
+      "learning_rate": 1.5245134568638197e-06,
+      "loss": 0.7023,
+      "step": 11754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7027389822010706,
+      "learning_rate": 1.523768384672299e-06,
+      "loss": 0.7243,
+      "step": 11755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8315382585792808,
+      "learning_rate": 1.5230234618621908e-06,
+      "loss": 0.8154,
+      "step": 11756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6366922489736497,
+      "learning_rate": 1.5222786884655117e-06,
+      "loss": 0.7198,
+      "step": 11757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.921161869124895,
+      "learning_rate": 1.5215340645142607e-06,
+      "loss": 0.499,
+      "step": 11758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7655257933858206,
+      "learning_rate": 1.5207895900404363e-06,
+      "loss": 0.691,
+      "step": 11759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4275435008941448,
+      "learning_rate": 1.520045265076034e-06,
+      "loss": 0.7922,
+      "step": 11760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8624722061745818,
+      "learning_rate": 1.5193010896530364e-06,
+      "loss": 0.6473,
+      "step": 11761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.330461310938107,
+      "learning_rate": 1.518557063803423e-06,
+      "loss": 0.6866,
+      "step": 11762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.130913888395994,
+      "learning_rate": 1.5178131875591662e-06,
+      "loss": 0.6789,
+      "step": 11763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7732507553439232,
+      "learning_rate": 1.5170694609522306e-06,
+      "loss": 0.6746,
+      "step": 11764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.142791575105262,
+      "learning_rate": 1.51632588401458e-06,
+      "loss": 0.6788,
+      "step": 11765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2547501018737268,
+      "learning_rate": 1.5155824567781618e-06,
+      "loss": 0.5717,
+      "step": 11766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7132607576318246,
+      "learning_rate": 1.5148391792749272e-06,
+      "loss": 0.6358,
+      "step": 11767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8043993866570178,
+      "learning_rate": 1.5140960515368165e-06,
+      "loss": 0.7878,
+      "step": 11768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9520293833865272,
+      "learning_rate": 1.5133530735957586e-06,
+      "loss": 0.6403,
+      "step": 11769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.280725373143093,
+      "learning_rate": 1.512610245483686e-06,
+      "loss": 0.5455,
+      "step": 11770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0929431573285733,
+      "learning_rate": 1.5118675672325184e-06,
+      "loss": 0.6677,
+      "step": 11771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9081137709017022,
+      "learning_rate": 1.5111250388741693e-06,
+      "loss": 0.8317,
+      "step": 11772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9581299318399825,
+      "learning_rate": 1.510382660440548e-06,
+      "loss": 0.7621,
+      "step": 11773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.81664472916902,
+      "learning_rate": 1.5096404319635533e-06,
+      "loss": 0.8695,
+      "step": 11774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6244279441364218,
+      "learning_rate": 1.5088983534750867e-06,
+      "loss": 0.6553,
+      "step": 11775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7615832899571189,
+      "learning_rate": 1.5081564250070291e-06,
+      "loss": 0.6538,
+      "step": 11776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6967038335348676,
+      "learning_rate": 1.5074146465912687e-06,
+      "loss": 0.6612,
+      "step": 11777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9202774460541894,
+      "learning_rate": 1.506673018259679e-06,
+      "loss": 0.6205,
+      "step": 11778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.701742282772854,
+      "learning_rate": 1.50593154004413e-06,
+      "loss": 0.6684,
+      "step": 11779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.679473016499785,
+      "learning_rate": 1.505190211976484e-06,
+      "loss": 0.7188,
+      "step": 11780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.825115383832085,
+      "learning_rate": 1.5044490340885987e-06,
+      "loss": 0.7179,
+      "step": 11781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5300201687791126,
+      "learning_rate": 1.5037080064123212e-06,
+      "loss": 0.729,
+      "step": 11782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7923592278425495,
+      "learning_rate": 1.5029671289795007e-06,
+      "loss": 0.6818,
+      "step": 11783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7836586161423962,
+      "learning_rate": 1.502226401821968e-06,
+      "loss": 0.6666,
+      "step": 11784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7204562323397228,
+      "learning_rate": 1.501485824971558e-06,
+      "loss": 0.7172,
+      "step": 11785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6087195800135543,
+      "learning_rate": 1.5007453984600956e-06,
+      "loss": 0.68,
+      "step": 11786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9748580609259232,
+      "learning_rate": 1.5000051223193928e-06,
+      "loss": 0.8616,
+      "step": 11787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5322420439882358,
+      "learning_rate": 1.4992649965812673e-06,
+      "loss": 0.7148,
+      "step": 11788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.126563075917564,
+      "learning_rate": 1.498525021277521e-06,
+      "loss": 0.7687,
+      "step": 11789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9813037805976295,
+      "learning_rate": 1.497785196439953e-06,
+      "loss": 0.7658,
+      "step": 11790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7333944908415857,
+      "learning_rate": 1.497045522100355e-06,
+      "loss": 0.7072,
+      "step": 11791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5231507161632105,
+      "learning_rate": 1.4963059982905104e-06,
+      "loss": 0.6431,
+      "step": 11792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7121243735573315,
+      "learning_rate": 1.4955666250422035e-06,
+      "loss": 0.6343,
+      "step": 11793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6463435229143382,
+      "learning_rate": 1.4948274023872005e-06,
+      "loss": 0.6454,
+      "step": 11794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.509538954976683,
+      "learning_rate": 1.4940883303572724e-06,
+      "loss": 0.6601,
+      "step": 11795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.62417286174877,
+      "learning_rate": 1.493349408984176e-06,
+      "loss": 0.6367,
+      "step": 11796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3626510443126256,
+      "learning_rate": 1.4926106382996658e-06,
+      "loss": 0.7113,
+      "step": 11797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6080429542306844,
+      "learning_rate": 1.491872018335488e-06,
+      "loss": 0.7348,
+      "step": 11798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7356911354232492,
+      "learning_rate": 1.4911335491233818e-06,
+      "loss": 0.7988,
+      "step": 11799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6594691119042433,
+      "learning_rate": 1.4903952306950808e-06,
+      "loss": 0.7496,
+      "step": 11800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8244006636683274,
+      "learning_rate": 1.489657063082316e-06,
+      "loss": 0.6171,
+      "step": 11801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6345574359699822,
+      "learning_rate": 1.4889190463168019e-06,
+      "loss": 0.593,
+      "step": 11802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.086579121870483,
+      "learning_rate": 1.4881811804302577e-06,
+      "loss": 0.6498,
+      "step": 11803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8233628750019388,
+      "learning_rate": 1.487443465454389e-06,
+      "loss": 0.7279,
+      "step": 11804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5705617511660186,
+      "learning_rate": 1.4867059014208979e-06,
+      "loss": 0.7477,
+      "step": 11805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.178680460286634,
+      "learning_rate": 1.4859684883614788e-06,
+      "loss": 0.6455,
+      "step": 11806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2063862624160926,
+      "learning_rate": 1.4852312263078183e-06,
+      "loss": 0.6546,
+      "step": 11807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8032634690070994,
+      "learning_rate": 1.4844941152916026e-06,
+      "loss": 0.7683,
+      "step": 11808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6889624357394295,
+      "learning_rate": 1.483757155344503e-06,
+      "loss": 0.6812,
+      "step": 11809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5631577428978152,
+      "learning_rate": 1.4830203464981879e-06,
+      "loss": 0.7188,
+      "step": 11810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0269146237840296,
+      "learning_rate": 1.4822836887843244e-06,
+      "loss": 0.6937,
+      "step": 11811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8687838891108557,
+      "learning_rate": 1.4815471822345618e-06,
+      "loss": 0.6807,
+      "step": 11812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7249978329182307,
+      "learning_rate": 1.4808108268805549e-06,
+      "loss": 0.7175,
+      "step": 11813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.704020547735568,
+      "learning_rate": 1.4800746227539437e-06,
+      "loss": 0.7476,
+      "step": 11814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.666672009383773,
+      "learning_rate": 1.479338569886366e-06,
+      "loss": 0.6531,
+      "step": 11815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1874562012819982,
+      "learning_rate": 1.47860266830945e-06,
+      "loss": 0.7577,
+      "step": 11816
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7111898831928252,
+      "learning_rate": 1.4778669180548195e-06,
+      "loss": 0.7335,
+      "step": 11817
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7542668425517103,
+      "learning_rate": 1.4771313191540927e-06,
+      "loss": 0.6914,
+      "step": 11818
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1467964854780546,
+      "learning_rate": 1.4763958716388798e-06,
+      "loss": 0.6124,
+      "step": 11819
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8503501481534745,
+      "learning_rate": 1.4756605755407816e-06,
+      "loss": 0.7009,
+      "step": 11820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1502917236114127,
+      "learning_rate": 1.474925430891399e-06,
+      "loss": 0.6759,
+      "step": 11821
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.18707450297034,
+      "learning_rate": 1.4741904377223215e-06,
+      "loss": 0.6216,
+      "step": 11822
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.116874629784628,
+      "learning_rate": 1.473455596065133e-06,
+      "loss": 0.6152,
+      "step": 11823
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6594519353055213,
+      "learning_rate": 1.4727209059514114e-06,
+      "loss": 0.6864,
+      "step": 11824
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.845590375628907,
+      "learning_rate": 1.4719863674127272e-06,
+      "loss": 0.6986,
+      "step": 11825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8484270567333907,
+      "learning_rate": 1.4712519804806485e-06,
+      "loss": 0.723,
+      "step": 11826
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.685095299165987,
+      "learning_rate": 1.4705177451867287e-06,
+      "loss": 0.7101,
+      "step": 11827
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6223891991626762,
+      "learning_rate": 1.4697836615625233e-06,
+      "loss": 0.6481,
+      "step": 11828
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6395164936780662,
+      "learning_rate": 1.4690497296395773e-06,
+      "loss": 0.755,
+      "step": 11829
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.007842288136002,
+      "learning_rate": 1.4683159494494259e-06,
+      "loss": 0.7632,
+      "step": 11830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7244616008051237,
+      "learning_rate": 1.4675823210236046e-06,
+      "loss": 0.689,
+      "step": 11831
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4207900195597258,
+      "learning_rate": 1.4668488443936385e-06,
+      "loss": 0.6416,
+      "step": 11832
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.668789345069953,
+      "learning_rate": 1.466115519591046e-06,
+      "loss": 0.699,
+      "step": 11833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5668875922985512,
+      "learning_rate": 1.46538234664734e-06,
+      "loss": 0.7114,
+      "step": 11834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6939658420224257,
+      "learning_rate": 1.4646493255940253e-06,
+      "loss": 0.6921,
+      "step": 11835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6660525944381708,
+      "learning_rate": 1.4639164564626041e-06,
+      "loss": 0.7347,
+      "step": 11836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4874496184036015,
+      "learning_rate": 1.4631837392845694e-06,
+      "loss": 0.7701,
+      "step": 11837
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8190737185935877,
+      "learning_rate": 1.4624511740914032e-06,
+      "loss": 0.7582,
+      "step": 11838
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1801446325515175,
+      "learning_rate": 1.4617187609145906e-06,
+      "loss": 0.6596,
+      "step": 11839
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.643512944824804,
+      "learning_rate": 1.4609864997856026e-06,
+      "loss": 0.7936,
+      "step": 11840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7352868401110528,
+      "learning_rate": 1.4602543907359068e-06,
+      "loss": 0.6991,
+      "step": 11841
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2837483445836402,
+      "learning_rate": 1.4595224337969633e-06,
+      "loss": 0.7158,
+      "step": 11842
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4409326699498894,
+      "learning_rate": 1.4587906290002242e-06,
+      "loss": 0.5881,
+      "step": 11843
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7641247760130243,
+      "learning_rate": 1.4580589763771413e-06,
+      "loss": 0.5718,
+      "step": 11844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5127175217801112,
+      "learning_rate": 1.4573274759591499e-06,
+      "loss": 0.6265,
+      "step": 11845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5495964604225667,
+      "learning_rate": 1.456596127777688e-06,
+      "loss": 0.7558,
+      "step": 11846
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6729777460121078,
+      "learning_rate": 1.4558649318641837e-06,
+      "loss": 0.8515,
+      "step": 11847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.776850716690227,
+      "learning_rate": 1.4551338882500538e-06,
+      "loss": 0.6269,
+      "step": 11848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5285002986894025,
+      "learning_rate": 1.4544029969667167e-06,
+      "loss": 0.6316,
+      "step": 11849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4430845404027104,
+      "learning_rate": 1.4536722580455792e-06,
+      "loss": 0.7136,
+      "step": 11850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2171829369956337,
+      "learning_rate": 1.4529416715180434e-06,
+      "loss": 0.6182,
+      "step": 11851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8705265169816003,
+      "learning_rate": 1.452211237415503e-06,
+      "loss": 0.657,
+      "step": 11852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.711148253052554,
+      "learning_rate": 1.451480955769346e-06,
+      "loss": 0.6564,
+      "step": 11853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7536474583145247,
+      "learning_rate": 1.4507508266109565e-06,
+      "loss": 0.7405,
+      "step": 11854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.531216350827763,
+      "learning_rate": 1.4500208499717088e-06,
+      "loss": 0.6346,
+      "step": 11855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.568003289053091,
+      "learning_rate": 1.4492910258829717e-06,
+      "loss": 0.7761,
+      "step": 11856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.646845508013944,
+      "learning_rate": 1.4485613543761068e-06,
+      "loss": 0.7788,
+      "step": 11857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.46424693153163,
+      "learning_rate": 1.44783183548247e-06,
+      "loss": 0.7251,
+      "step": 11858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1386722273517906,
+      "learning_rate": 1.4471024692334101e-06,
+      "loss": 0.5586,
+      "step": 11859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0993184334377135,
+      "learning_rate": 1.44637325566027e-06,
+      "loss": 0.7659,
+      "step": 11860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.859443642459071,
+      "learning_rate": 1.445644194794384e-06,
+      "loss": 0.7609,
+      "step": 11861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8069481679980957,
+      "learning_rate": 1.4449152866670858e-06,
+      "loss": 0.7232,
+      "step": 11862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.042296259243307,
+      "learning_rate": 1.4441865313096926e-06,
+      "loss": 0.6474,
+      "step": 11863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5161767294019544,
+      "learning_rate": 1.4434579287535244e-06,
+      "loss": 0.6625,
+      "step": 11864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6197545742042696,
+      "learning_rate": 1.4427294790298902e-06,
+      "loss": 0.6987,
+      "step": 11865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.788018580916695,
+      "learning_rate": 1.4420011821700924e-06,
+      "loss": 0.6481,
+      "step": 11866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9750058318770554,
+      "learning_rate": 1.4412730382054285e-06,
+      "loss": 0.7642,
+      "step": 11867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8137725452665274,
+      "learning_rate": 1.4405450471671861e-06,
+      "loss": 0.7141,
+      "step": 11868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6495365668460005,
+      "learning_rate": 1.439817209086653e-06,
+      "loss": 0.6378,
+      "step": 11869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0748771240497568,
+      "learning_rate": 1.4390895239951025e-06,
+      "loss": 0.6501,
+      "step": 11870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8721457024477202,
+      "learning_rate": 1.4383619919238035e-06,
+      "loss": 0.689,
+      "step": 11871
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.574893382584542,
+      "learning_rate": 1.4376346129040243e-06,
+      "loss": 0.7363,
+      "step": 11872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6710620424850522,
+      "learning_rate": 1.4369073869670186e-06,
+      "loss": 0.7265,
+      "step": 11873
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7118275871151196,
+      "learning_rate": 1.4361803141440384e-06,
+      "loss": 0.8013,
+      "step": 11874
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.158072383753148,
+      "learning_rate": 1.4354533944663274e-06,
+      "loss": 0.7018,
+      "step": 11875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8034380486541326,
+      "learning_rate": 1.4347266279651224e-06,
+      "loss": 0.7363,
+      "step": 11876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8547875858631084,
+      "learning_rate": 1.4340000146716544e-06,
+      "loss": 0.8315,
+      "step": 11877
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6328375593663504,
+      "learning_rate": 1.4332735546171466e-06,
+      "loss": 0.6768,
+      "step": 11878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6894388292356977,
+      "learning_rate": 1.432547247832819e-06,
+      "loss": 0.8304,
+      "step": 11879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1766034637855223,
+      "learning_rate": 1.431821094349883e-06,
+      "loss": 0.6437,
+      "step": 11880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5854986616473408,
+      "learning_rate": 1.4310950941995378e-06,
+      "loss": 0.6405,
+      "step": 11881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5956211723656253,
+      "learning_rate": 1.4303692474129866e-06,
+      "loss": 0.6915,
+      "step": 11882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9765351425935447,
+      "learning_rate": 1.4296435540214181e-06,
+      "loss": 0.7948,
+      "step": 11883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5785937857446517,
+      "learning_rate": 1.4289180140560189e-06,
+      "loss": 0.6501,
+      "step": 11884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1205786878896062,
+      "learning_rate": 1.428192627547965e-06,
+      "loss": 0.6806,
+      "step": 11885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8102842718708814,
+      "learning_rate": 1.4274673945284278e-06,
+      "loss": 0.6383,
+      "step": 11886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.732857442386427,
+      "learning_rate": 1.4267423150285758e-06,
+      "loss": 0.6565,
+      "step": 11887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6603244293644666,
+      "learning_rate": 1.4260173890795631e-06,
+      "loss": 0.7699,
+      "step": 11888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8953978824382185,
+      "learning_rate": 1.4252926167125413e-06,
+      "loss": 0.7038,
+      "step": 11889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9291108853414078,
+      "learning_rate": 1.4245679979586591e-06,
+      "loss": 0.6931,
+      "step": 11890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5758269414140536,
+      "learning_rate": 1.4238435328490525e-06,
+      "loss": 0.6543,
+      "step": 11891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5729985485937026,
+      "learning_rate": 1.4231192214148538e-06,
+      "loss": 0.7744,
+      "step": 11892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.145685129885805,
+      "learning_rate": 1.422395063687188e-06,
+      "loss": 0.7386,
+      "step": 11893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7044218655397934,
+      "learning_rate": 1.421671059697175e-06,
+      "loss": 0.6232,
+      "step": 11894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6109757577053905,
+      "learning_rate": 1.4209472094759252e-06,
+      "loss": 0.7135,
+      "step": 11895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.77784420345275,
+      "learning_rate": 1.4202235130545432e-06,
+      "loss": 0.665,
+      "step": 11896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6548946533034574,
+      "learning_rate": 1.4194999704641315e-06,
+      "loss": 0.7755,
+      "step": 11897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5262062851956089,
+      "learning_rate": 1.4187765817357812e-06,
+      "loss": 0.7605,
+      "step": 11898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8573647634615367,
+      "learning_rate": 1.418053346900574e-06,
+      "loss": 0.8912,
+      "step": 11899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5601226882582384,
+      "learning_rate": 1.4173302659895938e-06,
+      "loss": 0.7146,
+      "step": 11900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8535721588150758,
+      "learning_rate": 1.4166073390339113e-06,
+      "loss": 0.7432,
+      "step": 11901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.836532156675521,
+      "learning_rate": 1.4158845660645915e-06,
+      "loss": 0.7328,
+      "step": 11902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5199977622739358,
+      "learning_rate": 1.4151619471126943e-06,
+      "loss": 0.6703,
+      "step": 11903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5962663790074063,
+      "learning_rate": 1.4144394822092712e-06,
+      "loss": 0.7224,
+      "step": 11904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5676810046792904,
+      "learning_rate": 1.4137171713853714e-06,
+      "loss": 0.6858,
+      "step": 11905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.849038963761894,
+      "learning_rate": 1.412995014672029e-06,
+      "loss": 0.6701,
+      "step": 11906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6450724224826199,
+      "learning_rate": 1.4122730121002808e-06,
+      "loss": 0.6849,
+      "step": 11907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2467551520781817,
+      "learning_rate": 1.4115511637011514e-06,
+      "loss": 0.5571,
+      "step": 11908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.712542534603733,
+      "learning_rate": 1.4108294695056606e-06,
+      "loss": 0.7217,
+      "step": 11909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8128307046541188,
+      "learning_rate": 1.4101079295448217e-06,
+      "loss": 0.726,
+      "step": 11910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1449324763852222,
+      "learning_rate": 1.4093865438496395e-06,
+      "loss": 0.6303,
+      "step": 11911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6050517772566057,
+      "learning_rate": 1.408665312451114e-06,
+      "loss": 0.6731,
+      "step": 11912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.650273327892686,
+      "learning_rate": 1.4079442353802392e-06,
+      "loss": 0.636,
+      "step": 11913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6016551271541506,
+      "learning_rate": 1.4072233126679985e-06,
+      "loss": 0.7123,
+      "step": 11914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8049656827191047,
+      "learning_rate": 1.406502544345375e-06,
+      "loss": 0.6575,
+      "step": 11915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.549353672119022,
+      "learning_rate": 1.40578193044334e-06,
+      "loss": 0.7415,
+      "step": 11916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2199287206295149,
+      "learning_rate": 1.4050614709928612e-06,
+      "loss": 0.6829,
+      "step": 11917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0141568315609777,
+      "learning_rate": 1.4043411660248967e-06,
+      "loss": 0.7583,
+      "step": 11918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6176107587410558,
+      "learning_rate": 1.4036210155703989e-06,
+      "loss": 0.7158,
+      "step": 11919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5203870022938857,
+      "learning_rate": 1.4029010196603182e-06,
+      "loss": 0.7316,
+      "step": 11920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5000823293152563,
+      "learning_rate": 1.4021811783255912e-06,
+      "loss": 0.6909,
+      "step": 11921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1369900797809633,
+      "learning_rate": 1.4014614915971498e-06,
+      "loss": 0.6163,
+      "step": 11922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.769622854550624,
+      "learning_rate": 1.4007419595059252e-06,
+      "loss": 0.7779,
+      "step": 11923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8749889314759225,
+      "learning_rate": 1.4000225820828317e-06,
+      "loss": 0.6484,
+      "step": 11924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.527429373470426,
+      "learning_rate": 1.3993033593587874e-06,
+      "loss": 0.7102,
+      "step": 11925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3385839400708164,
+      "learning_rate": 1.3985842913646963e-06,
+      "loss": 0.6754,
+      "step": 11926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7335577069876569,
+      "learning_rate": 1.3978653781314595e-06,
+      "loss": 0.7302,
+      "step": 11927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7653458924899685,
+      "learning_rate": 1.3971466196899697e-06,
+      "loss": 0.7597,
+      "step": 11928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6240580626044465,
+      "learning_rate": 1.3964280160711119e-06,
+      "loss": 0.6953,
+      "step": 11929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2055560962636696,
+      "learning_rate": 1.3957095673057708e-06,
+      "loss": 0.6836,
+      "step": 11930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1320707234214393,
+      "learning_rate": 1.3949912734248156e-06,
+      "loss": 0.7034,
+      "step": 11931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8558773963086101,
+      "learning_rate": 1.3942731344591127e-06,
+      "loss": 0.73,
+      "step": 11932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7669025589647909,
+      "learning_rate": 1.3935551504395246e-06,
+      "loss": 0.7995,
+      "step": 11933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8418010491458963,
+      "learning_rate": 1.3928373213969038e-06,
+      "loss": 0.7809,
+      "step": 11934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5536780178675058,
+      "learning_rate": 1.3921196473620975e-06,
+      "loss": 0.6525,
+      "step": 11935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8533276351184278,
+      "learning_rate": 1.3914021283659445e-06,
+      "loss": 0.7212,
+      "step": 11936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0105208592459318,
+      "learning_rate": 1.3906847644392773e-06,
+      "loss": 0.6605,
+      "step": 11937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8278251268363057,
+      "learning_rate": 1.3899675556129272e-06,
+      "loss": 0.7074,
+      "step": 11938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7324071333741802,
+      "learning_rate": 1.38925050191771e-06,
+      "loss": 0.7123,
+      "step": 11939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8203206610832905,
+      "learning_rate": 1.3885336033844388e-06,
+      "loss": 0.6823,
+      "step": 11940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.502018650775922,
+      "learning_rate": 1.3878168600439246e-06,
+      "loss": 0.7583,
+      "step": 11941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7441361309143177,
+      "learning_rate": 1.3871002719269616e-06,
+      "loss": 0.7131,
+      "step": 11942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.458576470367799,
+      "learning_rate": 1.386383839064348e-06,
+      "loss": 0.6715,
+      "step": 11943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.804461008687341,
+      "learning_rate": 1.3856675614868687e-06,
+      "loss": 0.6982,
+      "step": 11944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8797623368899696,
+      "learning_rate": 1.3849514392253038e-06,
+      "loss": 0.7649,
+      "step": 11945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.24724617923923,
+      "learning_rate": 1.3842354723104267e-06,
+      "loss": 0.6427,
+      "step": 11946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.115739480539167,
+      "learning_rate": 1.3835196607730028e-06,
+      "loss": 0.7456,
+      "step": 11947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6233770617895436,
+      "learning_rate": 1.3828040046437963e-06,
+      "loss": 0.6638,
+      "step": 11948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.574671536124627,
+      "learning_rate": 1.3820885039535564e-06,
+      "loss": 0.6358,
+      "step": 11949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.589052573880993,
+      "learning_rate": 1.3813731587330291e-06,
+      "loss": 0.7552,
+      "step": 11950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1615694609836216,
+      "learning_rate": 1.380657969012958e-06,
+      "loss": 0.7744,
+      "step": 11951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8753685239096485,
+      "learning_rate": 1.3799429348240746e-06,
+      "loss": 0.7497,
+      "step": 11952
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6079206197939049,
+      "learning_rate": 1.3792280561971056e-06,
+      "loss": 0.7381,
+      "step": 11953
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0180472176177418,
+      "learning_rate": 1.378513333162771e-06,
+      "loss": 0.6124,
+      "step": 11954
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9346888620820526,
+      "learning_rate": 1.3777987657517822e-06,
+      "loss": 0.8446,
+      "step": 11955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6351376308053698,
+      "learning_rate": 1.3770843539948508e-06,
+      "loss": 0.7883,
+      "step": 11956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9591760308841408,
+      "learning_rate": 1.3763700979226701e-06,
+      "loss": 0.8068,
+      "step": 11957
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.840323118579226,
+      "learning_rate": 1.3756559975659385e-06,
+      "loss": 0.6752,
+      "step": 11958
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7039123875610698,
+      "learning_rate": 1.3749420529553414e-06,
+      "loss": 0.7438,
+      "step": 11959
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5256561770773607,
+      "learning_rate": 1.3742282641215549e-06,
+      "loss": 0.6879,
+      "step": 11960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.90236696758476,
+      "learning_rate": 1.3735146310952568e-06,
+      "loss": 0.6974,
+      "step": 11961
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9185736069291253,
+      "learning_rate": 1.3728011539071112e-06,
+      "loss": 0.7198,
+      "step": 11962
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.107580525750726,
+      "learning_rate": 1.3720878325877785e-06,
+      "loss": 0.7406,
+      "step": 11963
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.628681582161571,
+      "learning_rate": 1.3713746671679112e-06,
+      "loss": 0.7377,
+      "step": 11964
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.965755150033726,
+      "learning_rate": 1.3706616576781545e-06,
+      "loss": 0.5982,
+      "step": 11965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7191406060971133,
+      "learning_rate": 1.3699488041491526e-06,
+      "loss": 0.7857,
+      "step": 11966
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6512832583198005,
+      "learning_rate": 1.3692361066115318e-06,
+      "loss": 0.6901,
+      "step": 11967
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6970778268335514,
+      "learning_rate": 1.3685235650959228e-06,
+      "loss": 0.6297,
+      "step": 11968
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5857781612458575,
+      "learning_rate": 1.3678111796329446e-06,
+      "loss": 0.6681,
+      "step": 11969
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8524415772259346,
+      "learning_rate": 1.3670989502532089e-06,
+      "loss": 0.729,
+      "step": 11970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4319726670735824,
+      "learning_rate": 1.3663868769873223e-06,
+      "loss": 0.6791,
+      "step": 11971
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5884015112212901,
+      "learning_rate": 1.365674959865884e-06,
+      "loss": 0.6974,
+      "step": 11972
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6943779726348707,
+      "learning_rate": 1.3649631989194856e-06,
+      "loss": 0.6876,
+      "step": 11973
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6423908192557621,
+      "learning_rate": 1.3642515941787171e-06,
+      "loss": 0.6322,
+      "step": 11974
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.59579694700765,
+      "learning_rate": 1.363540145674152e-06,
+      "loss": 0.6835,
+      "step": 11975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6655274008269663,
+      "learning_rate": 1.3628288534363664e-06,
+      "loss": 0.6609,
+      "step": 11976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8989018206506123,
+      "learning_rate": 1.362117717495926e-06,
+      "loss": 0.7173,
+      "step": 11977
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7388497767572126,
+      "learning_rate": 1.3614067378833895e-06,
+      "loss": 0.6593,
+      "step": 11978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.244093045952615,
+      "learning_rate": 1.3606959146293086e-06,
+      "loss": 0.8447,
+      "step": 11979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8637226485976601,
+      "learning_rate": 1.3599852477642283e-06,
+      "loss": 0.708,
+      "step": 11980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7703711880278032,
+      "learning_rate": 1.3592747373186915e-06,
+      "loss": 0.7095,
+      "step": 11981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9843617977230785,
+      "learning_rate": 1.358564383323226e-06,
+      "loss": 0.7085,
+      "step": 11982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6253720633919098,
+      "learning_rate": 1.357854185808357e-06,
+      "loss": 0.5861,
+      "step": 11983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.257363458734188,
+      "learning_rate": 1.3571441448046086e-06,
+      "loss": 0.6954,
+      "step": 11984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7429217485980801,
+      "learning_rate": 1.3564342603424864e-06,
+      "loss": 0.8177,
+      "step": 11985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7036320710824213,
+      "learning_rate": 1.3557245324524999e-06,
+      "loss": 0.7463,
+      "step": 11986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5929376019741877,
+      "learning_rate": 1.3550149611651463e-06,
+      "loss": 0.6808,
+      "step": 11987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2217801928578458,
+      "learning_rate": 1.3543055465109183e-06,
+      "loss": 0.5262,
+      "step": 11988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7949099083966942,
+      "learning_rate": 1.3535962885202997e-06,
+      "loss": 0.7357,
+      "step": 11989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5303207875410298,
+      "learning_rate": 1.3528871872237687e-06,
+      "loss": 0.682,
+      "step": 11990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3386903780524546,
+      "learning_rate": 1.3521782426517988e-06,
+      "loss": 0.6567,
+      "step": 11991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6386320230891072,
+      "learning_rate": 1.3514694548348556e-06,
+      "loss": 0.6946,
+      "step": 11992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7963465701926142,
+      "learning_rate": 1.3507608238033927e-06,
+      "loss": 0.5956,
+      "step": 11993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8947920657285175,
+      "learning_rate": 1.350052349587866e-06,
+      "loss": 0.6817,
+      "step": 11994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.628390156422328,
+      "learning_rate": 1.3493440322187185e-06,
+      "loss": 0.8452,
+      "step": 11995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.671534789463041,
+      "learning_rate": 1.3486358717263887e-06,
+      "loss": 0.6752,
+      "step": 11996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9863547677454285,
+      "learning_rate": 1.3479278681413072e-06,
+      "loss": 0.87,
+      "step": 11997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9513903513238382,
+      "learning_rate": 1.3472200214938974e-06,
+      "loss": 0.7023,
+      "step": 11998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.623366943299237,
+      "learning_rate": 1.3465123318145817e-06,
+      "loss": 0.6236,
+      "step": 11999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1477155111921564,
+      "learning_rate": 1.3458047991337664e-06,
+      "loss": 0.6077,
+      "step": 12000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.672005416502843,
+      "learning_rate": 1.3450974234818554e-06,
+      "loss": 0.7526,
+      "step": 12001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8392264439551445,
+      "learning_rate": 1.34439020488925e-06,
+      "loss": 0.742,
+      "step": 12002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9316721604323932,
+      "learning_rate": 1.3436831433863384e-06,
+      "loss": 0.7128,
+      "step": 12003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1831792530294962,
+      "learning_rate": 1.342976239003505e-06,
+      "loss": 0.6276,
+      "step": 12004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9860924843391943,
+      "learning_rate": 1.3422694917711276e-06,
+      "loss": 0.6401,
+      "step": 12005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.79665352618551,
+      "learning_rate": 1.3415629017195763e-06,
+      "loss": 0.8088,
+      "step": 12006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7900020017432676,
+      "learning_rate": 1.3408564688792153e-06,
+      "loss": 0.8801,
+      "step": 12007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6804344957116617,
+      "learning_rate": 1.340150193280399e-06,
+      "loss": 0.6204,
+      "step": 12008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7553655645084285,
+      "learning_rate": 1.339444074953482e-06,
+      "loss": 0.808,
+      "step": 12009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3759079278236277,
+      "learning_rate": 1.3387381139288064e-06,
+      "loss": 0.6382,
+      "step": 12010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8682504949651808,
+      "learning_rate": 1.3380323102367055e-06,
+      "loss": 0.7514,
+      "step": 12011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.935839191946656,
+      "learning_rate": 1.3373266639075134e-06,
+      "loss": 0.6945,
+      "step": 12012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7346197650608626,
+      "learning_rate": 1.3366211749715519e-06,
+      "loss": 0.7031,
+      "step": 12013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9370533675743933,
+      "learning_rate": 1.335915843459137e-06,
+      "loss": 0.6765,
+      "step": 12014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.464853701673546,
+      "learning_rate": 1.335210669400579e-06,
+      "loss": 0.6819,
+      "step": 12015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0223780470398975,
+      "learning_rate": 1.3345056528261784e-06,
+      "loss": 0.6501,
+      "step": 12016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.74763662630206,
+      "learning_rate": 1.3338007937662367e-06,
+      "loss": 0.8253,
+      "step": 12017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6388170814170566,
+      "learning_rate": 1.3330960922510373e-06,
+      "loss": 0.5958,
+      "step": 12018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4158959743241446,
+      "learning_rate": 1.3323915483108662e-06,
+      "loss": 0.5908,
+      "step": 12019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1787262875745506,
+      "learning_rate": 1.3316871619759992e-06,
+      "loss": 0.6779,
+      "step": 12020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1176495227470287,
+      "learning_rate": 1.3309829332767037e-06,
+      "loss": 0.6888,
+      "step": 12021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5598630790068528,
+      "learning_rate": 1.330278862243244e-06,
+      "loss": 0.7908,
+      "step": 12022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7941283357839393,
+      "learning_rate": 1.3295749489058741e-06,
+      "loss": 0.7155,
+      "step": 12023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.121171424726944,
+      "learning_rate": 1.3288711932948427e-06,
+      "loss": 0.8246,
+      "step": 12024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6511502951957073,
+      "learning_rate": 1.3281675954403923e-06,
+      "loss": 0.7793,
+      "step": 12025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.671788707808332,
+      "learning_rate": 1.3274641553727568e-06,
+      "loss": 0.634,
+      "step": 12026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.459293850356576,
+      "learning_rate": 1.3267608731221671e-06,
+      "loss": 0.6421,
+      "step": 12027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.912223798728156,
+      "learning_rate": 1.3260577487188431e-06,
+      "loss": 0.718,
+      "step": 12028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.098537121063058,
+      "learning_rate": 1.3253547821930002e-06,
+      "loss": 0.6563,
+      "step": 12029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.73764687552057,
+      "learning_rate": 1.3246519735748465e-06,
+      "loss": 0.9196,
+      "step": 12030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9035326646287924,
+      "learning_rate": 1.3239493228945821e-06,
+      "loss": 0.7607,
+      "step": 12031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8069635477706212,
+      "learning_rate": 1.323246830182403e-06,
+      "loss": 0.6908,
+      "step": 12032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7123494254796272,
+      "learning_rate": 1.3225444954684962e-06,
+      "loss": 0.7067,
+      "step": 12033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6695674051991134,
+      "learning_rate": 1.3218423187830409e-06,
+      "loss": 0.7278,
+      "step": 12034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.629266645392056,
+      "learning_rate": 1.3211403001562155e-06,
+      "loss": 0.6442,
+      "step": 12035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0928845891620917,
+      "learning_rate": 1.3204384396181819e-06,
+      "loss": 0.638,
+      "step": 12036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7808217413157785,
+      "learning_rate": 1.3197367371991048e-06,
+      "loss": 0.72,
+      "step": 12037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0731751493413415,
+      "learning_rate": 1.3190351929291367e-06,
+      "loss": 0.7498,
+      "step": 12038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.839455320779782,
+      "learning_rate": 1.3183338068384243e-06,
+      "loss": 0.7162,
+      "step": 12039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.468502566652883,
+      "learning_rate": 1.3176325789571075e-06,
+      "loss": 0.6033,
+      "step": 12040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8747337930755745,
+      "learning_rate": 1.316931509315319e-06,
+      "loss": 0.7069,
+      "step": 12041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6420225211891923,
+      "learning_rate": 1.3162305979431883e-06,
+      "loss": 0.6896,
+      "step": 12042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.01909731425546,
+      "learning_rate": 1.315529844870832e-06,
+      "loss": 0.6961,
+      "step": 12043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3118381827605834,
+      "learning_rate": 1.3148292501283627e-06,
+      "loss": 0.699,
+      "step": 12044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9209288181727089,
+      "learning_rate": 1.314128813745889e-06,
+      "loss": 0.7673,
+      "step": 12045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8114664420329616,
+      "learning_rate": 1.3134285357535093e-06,
+      "loss": 0.7331,
+      "step": 12046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6285188072495755,
+      "learning_rate": 1.3127284161813153e-06,
+      "loss": 0.6607,
+      "step": 12047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8870358610750033,
+      "learning_rate": 1.3120284550593936e-06,
+      "loss": 0.6568,
+      "step": 12048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8892376802822384,
+      "learning_rate": 1.3113286524178232e-06,
+      "loss": 0.8036,
+      "step": 12049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8936520981862857,
+      "learning_rate": 1.310629008286675e-06,
+      "loss": 0.6545,
+      "step": 12050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0839422869088626,
+      "learning_rate": 1.3099295226960151e-06,
+      "loss": 0.6777,
+      "step": 12051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.515014101034557,
+      "learning_rate": 1.3092301956759008e-06,
+      "loss": 0.713,
+      "step": 12052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.342699095785085,
+      "learning_rate": 1.3085310272563872e-06,
+      "loss": 0.661,
+      "step": 12053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.913730515265672,
+      "learning_rate": 1.3078320174675141e-06,
+      "loss": 0.8101,
+      "step": 12054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5788203671025225,
+      "learning_rate": 1.3071331663393233e-06,
+      "loss": 0.5674,
+      "step": 12055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9943336149188542,
+      "learning_rate": 1.3064344739018452e-06,
+      "loss": 0.7364,
+      "step": 12056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6557471703629647,
+      "learning_rate": 1.3057359401851034e-06,
+      "loss": 0.6883,
+      "step": 12057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5082103616837328,
+      "learning_rate": 1.305037565219116e-06,
+      "loss": 0.6247,
+      "step": 12058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8383537292695922,
+      "learning_rate": 1.3043393490338918e-06,
+      "loss": 0.8018,
+      "step": 12059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6580543572625557,
+      "learning_rate": 1.3036412916594398e-06,
+      "loss": 0.5459,
+      "step": 12060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9854301512230714,
+      "learning_rate": 1.3029433931257524e-06,
+      "loss": 0.6615,
+      "step": 12061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3073031964325401,
+      "learning_rate": 1.3022456534628191e-06,
+      "loss": 0.6114,
+      "step": 12062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7735613285448069,
+      "learning_rate": 1.3015480727006274e-06,
+      "loss": 0.684,
+      "step": 12063
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8132071976778195,
+      "learning_rate": 1.3008506508691516e-06,
+      "loss": 0.7102,
+      "step": 12064
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6627404911005053,
+      "learning_rate": 1.3001533879983618e-06,
+      "loss": 0.7755,
+      "step": 12065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1078511710317365,
+      "learning_rate": 1.2994562841182213e-06,
+      "loss": 0.805,
+      "step": 12066
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9012524899174752,
+      "learning_rate": 1.298759339258685e-06,
+      "loss": 0.719,
+      "step": 12067
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5185890758097227,
+      "learning_rate": 1.2980625534497037e-06,
+      "loss": 0.7231,
+      "step": 12068
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4640911872342373,
+      "learning_rate": 1.2973659267212173e-06,
+      "loss": 0.6965,
+      "step": 12069
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7665295140453647,
+      "learning_rate": 1.2966694591031653e-06,
+      "loss": 0.8458,
+      "step": 12070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.002635366775161,
+      "learning_rate": 1.2959731506254751e-06,
+      "loss": 0.659,
+      "step": 12071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6773567098795734,
+      "learning_rate": 1.2952770013180655e-06,
+      "loss": 0.894,
+      "step": 12072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6164838146504656,
+      "learning_rate": 1.2945810112108554e-06,
+      "loss": 0.7795,
+      "step": 12073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.561929109487986,
+      "learning_rate": 1.2938851803337516e-06,
+      "loss": 0.6166,
+      "step": 12074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5704508014276188,
+      "learning_rate": 1.2931895087166551e-06,
+      "loss": 0.7319,
+      "step": 12075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9139975087375782,
+      "learning_rate": 1.2924939963894617e-06,
+      "loss": 0.6558,
+      "step": 12076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6684115943717406,
+      "learning_rate": 1.2917986433820567e-06,
+      "loss": 0.6958,
+      "step": 12077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.078370156150722,
+      "learning_rate": 1.2911034497243251e-06,
+      "loss": 0.6439,
+      "step": 12078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6373527524613973,
+      "learning_rate": 1.290408415446136e-06,
+      "loss": 0.6949,
+      "step": 12079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7070957268491793,
+      "learning_rate": 1.2897135405773602e-06,
+      "loss": 0.807,
+      "step": 12080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1015830449590616,
+      "learning_rate": 1.2890188251478569e-06,
+      "loss": 0.6119,
+      "step": 12081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9610617635148417,
+      "learning_rate": 1.2883242691874792e-06,
+      "loss": 0.6333,
+      "step": 12082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5047272865744015,
+      "learning_rate": 1.2876298727260744e-06,
+      "loss": 0.6649,
+      "step": 12083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5694552664796841,
+      "learning_rate": 1.2869356357934815e-06,
+      "loss": 0.679,
+      "step": 12084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9788976131722165,
+      "learning_rate": 1.2862415584195338e-06,
+      "loss": 0.6407,
+      "step": 12085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8112325641594247,
+      "learning_rate": 1.2855476406340578e-06,
+      "loss": 0.7668,
+      "step": 12086
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7363760094543073,
+      "learning_rate": 1.28485388246687e-06,
+      "loss": 0.8055,
+      "step": 12087
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8315527276408823,
+      "learning_rate": 1.2841602839477868e-06,
+      "loss": 0.6453,
+      "step": 12088
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.995558688925542,
+      "learning_rate": 1.2834668451066118e-06,
+      "loss": 0.8377,
+      "step": 12089
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6122137699407462,
+      "learning_rate": 1.2827735659731438e-06,
+      "loss": 0.7268,
+      "step": 12090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.756609533512625,
+      "learning_rate": 1.2820804465771742e-06,
+      "loss": 0.7978,
+      "step": 12091
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6125217971252157,
+      "learning_rate": 1.2813874869484866e-06,
+      "loss": 0.8135,
+      "step": 12092
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6314837198732262,
+      "learning_rate": 1.2806946871168634e-06,
+      "loss": 0.5749,
+      "step": 12093
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5543403037334325,
+      "learning_rate": 1.2800020471120717e-06,
+      "loss": 0.6701,
+      "step": 12094
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8089550736387046,
+      "learning_rate": 1.279309566963875e-06,
+      "loss": 0.785,
+      "step": 12095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.930932709249818,
+      "learning_rate": 1.2786172467020357e-06,
+      "loss": 0.8365,
+      "step": 12096
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7345689275121583,
+      "learning_rate": 1.2779250863562987e-06,
+      "loss": 0.6042,
+      "step": 12097
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7044538846025195,
+      "learning_rate": 1.2772330859564115e-06,
+      "loss": 0.7084,
+      "step": 12098
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9948972120750692,
+      "learning_rate": 1.276541245532109e-06,
+      "loss": 0.744,
+      "step": 12099
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.782384415801023,
+      "learning_rate": 1.2758495651131226e-06,
+      "loss": 0.6559,
+      "step": 12100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2736451588590225,
+      "learning_rate": 1.275158044729174e-06,
+      "loss": 0.5757,
+      "step": 12101
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8371306208631575,
+      "learning_rate": 1.2744666844099801e-06,
+      "loss": 0.6003,
+      "step": 12102
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5499494799062263,
+      "learning_rate": 1.2737754841852501e-06,
+      "loss": 0.5704,
+      "step": 12103
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7907966400243385,
+      "learning_rate": 1.2730844440846862e-06,
+      "loss": 0.7402,
+      "step": 12104
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.785383462402314,
+      "learning_rate": 1.2723935641379826e-06,
+      "loss": 0.7003,
+      "step": 12105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6474802056818287,
+      "learning_rate": 1.2717028443748303e-06,
+      "loss": 0.6425,
+      "step": 12106
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0174507779528206,
+      "learning_rate": 1.2710122848249106e-06,
+      "loss": 0.6684,
+      "step": 12107
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0628787763554404,
+      "learning_rate": 1.2703218855178984e-06,
+      "loss": 0.661,
+      "step": 12108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.67088370605151,
+      "learning_rate": 1.2696316464834607e-06,
+      "loss": 0.7053,
+      "step": 12109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6629205437412717,
+      "learning_rate": 1.2689415677512574e-06,
+      "loss": 0.7503,
+      "step": 12110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5814838406166207,
+      "learning_rate": 1.2682516493509478e-06,
+      "loss": 0.6768,
+      "step": 12111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5934895522464088,
+      "learning_rate": 1.267561891312174e-06,
+      "loss": 0.7187,
+      "step": 12112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.280648952964166,
+      "learning_rate": 1.2668722936645773e-06,
+      "loss": 0.6845,
+      "step": 12113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.002801159663558,
+      "learning_rate": 1.2661828564377948e-06,
+      "loss": 0.638,
+      "step": 12114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4397232609401398,
+      "learning_rate": 1.2654935796614481e-06,
+      "loss": 0.764,
+      "step": 12115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0186343678290863,
+      "learning_rate": 1.2648044633651607e-06,
+      "loss": 0.7781,
+      "step": 12116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9938966799145372,
+      "learning_rate": 1.2641155075785444e-06,
+      "loss": 0.6946,
+      "step": 12117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9669629837765412,
+      "learning_rate": 1.2634267123312049e-06,
+      "loss": 0.7297,
+      "step": 12118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.832571032524733,
+      "learning_rate": 1.2627380776527415e-06,
+      "loss": 0.7236,
+      "step": 12119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6040582302308852,
+      "learning_rate": 1.2620496035727448e-06,
+      "loss": 0.7114,
+      "step": 12120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6257492609780027,
+      "learning_rate": 1.2613612901208022e-06,
+      "loss": 0.682,
+      "step": 12121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.294007889808752,
+      "learning_rate": 1.2606731373264936e-06,
+      "loss": 0.595,
+      "step": 12122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.745799569224936,
+      "learning_rate": 1.2599851452193851e-06,
+      "loss": 0.6847,
+      "step": 12123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6048188985367446,
+      "learning_rate": 1.259297313829046e-06,
+      "loss": 0.7166,
+      "step": 12124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5888725582527423,
+      "learning_rate": 1.258609643185032e-06,
+      "loss": 0.6279,
+      "step": 12125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6630613661867555,
+      "learning_rate": 1.2579221333168945e-06,
+      "loss": 0.7872,
+      "step": 12126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7706461090869183,
+      "learning_rate": 1.2572347842541766e-06,
+      "loss": 0.6298,
+      "step": 12127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9871709194026483,
+      "learning_rate": 1.2565475960264145e-06,
+      "loss": 0.7561,
+      "step": 12128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8620238422454043,
+      "learning_rate": 1.255860568663142e-06,
+      "loss": 0.6962,
+      "step": 12129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6830269375687321,
+      "learning_rate": 1.2551737021938765e-06,
+      "loss": 0.6295,
+      "step": 12130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0477850637573485,
+      "learning_rate": 1.2544869966481389e-06,
+      "loss": 0.6636,
+      "step": 12131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.677109414406624,
+      "learning_rate": 1.253800452055438e-06,
+      "loss": 0.704,
+      "step": 12132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4631491376635157,
+      "learning_rate": 1.2531140684452724e-06,
+      "loss": 0.6809,
+      "step": 12133
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7852111438050149,
+      "learning_rate": 1.2524278458471411e-06,
+      "loss": 0.6018,
+      "step": 12134
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5096450644176707,
+      "learning_rate": 1.2517417842905316e-06,
+      "loss": 0.5997,
+      "step": 12135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3470670404911496,
+      "learning_rate": 1.2510558838049258e-06,
+      "loss": 0.7184,
+      "step": 12136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.756868571636369,
+      "learning_rate": 1.2503701444197974e-06,
+      "loss": 0.6621,
+      "step": 12137
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6333677431017561,
+      "learning_rate": 1.249684566164614e-06,
+      "loss": 0.7027,
+      "step": 12138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6427200105634359,
+      "learning_rate": 1.248999149068838e-06,
+      "loss": 0.6675,
+      "step": 12139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6831119214612262,
+      "learning_rate": 1.2483138931619227e-06,
+      "loss": 0.8156,
+      "step": 12140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9144246160140652,
+      "learning_rate": 1.2476287984733154e-06,
+      "loss": 0.7429,
+      "step": 12141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7155837875419013,
+      "learning_rate": 1.246943865032455e-06,
+      "loss": 0.7077,
+      "step": 12142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.60812127900586,
+      "learning_rate": 1.246259092868775e-06,
+      "loss": 0.7706,
+      "step": 12143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.734125975068975,
+      "learning_rate": 1.2455744820117028e-06,
+      "loss": 0.7117,
+      "step": 12144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8244686190983055,
+      "learning_rate": 1.2448900324906559e-06,
+      "loss": 0.7128,
+      "step": 12145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.582951006619882,
+      "learning_rate": 1.2442057443350458e-06,
+      "loss": 0.7416,
+      "step": 12146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.91176538723806,
+      "learning_rate": 1.2435216175742826e-06,
+      "loss": 0.7243,
+      "step": 12147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6360806029032506,
+      "learning_rate": 1.242837652237759e-06,
+      "loss": 0.8172,
+      "step": 12148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7019846701560886,
+      "learning_rate": 1.2421538483548706e-06,
+      "loss": 0.6715,
+      "step": 12149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6366276457792364,
+      "learning_rate": 1.2414702059550016e-06,
+      "loss": 0.7267,
+      "step": 12150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7138104052896788,
+      "learning_rate": 1.2407867250675253e-06,
+      "loss": 0.6763,
+      "step": 12151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7609251387095248,
+      "learning_rate": 1.2401034057218181e-06,
+      "loss": 0.7343,
+      "step": 12152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5728985797010708,
+      "learning_rate": 1.2394202479472406e-06,
+      "loss": 0.7007,
+      "step": 12153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7277549200584412,
+      "learning_rate": 1.2387372517731505e-06,
+      "loss": 0.6004,
+      "step": 12154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7343098369564198,
+      "learning_rate": 1.238054417228897e-06,
+      "loss": 0.7921,
+      "step": 12155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.277094537332713,
+      "learning_rate": 1.2373717443438227e-06,
+      "loss": 0.7951,
+      "step": 12156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8847224017981368,
+      "learning_rate": 1.2366892331472657e-06,
+      "loss": 0.7268,
+      "step": 12157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6997421350862099,
+      "learning_rate": 1.2360068836685536e-06,
+      "loss": 0.6542,
+      "step": 12158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3257640462558997,
+      "learning_rate": 1.2353246959370086e-06,
+      "loss": 0.6512,
+      "step": 12159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6965747801875857,
+      "learning_rate": 1.234642669981946e-06,
+      "loss": 0.7356,
+      "step": 12160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6689145771425413,
+      "learning_rate": 1.2339608058326736e-06,
+      "loss": 0.8386,
+      "step": 12161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7015894838593935,
+      "learning_rate": 1.2332791035184926e-06,
+      "loss": 0.768,
+      "step": 12162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.687169231379623,
+      "learning_rate": 1.2325975630686976e-06,
+      "loss": 0.6674,
+      "step": 12163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7247803461444091,
+      "learning_rate": 1.2319161845125744e-06,
+      "loss": 0.6905,
+      "step": 12164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6982125212399641,
+      "learning_rate": 1.2312349678794072e-06,
+      "loss": 0.6433,
+      "step": 12165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6642269120560482,
+      "learning_rate": 1.2305539131984646e-06,
+      "loss": 0.7507,
+      "step": 12166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0775683313595428,
+      "learning_rate": 1.2298730204990162e-06,
+      "loss": 0.6246,
+      "step": 12167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.225860872341814,
+      "learning_rate": 1.229192289810321e-06,
+      "loss": 0.7108,
+      "step": 12168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.912421131515851,
+      "learning_rate": 1.228511721161631e-06,
+      "loss": 0.7334,
+      "step": 12169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6470252932915632,
+      "learning_rate": 1.2278313145821924e-06,
+      "loss": 0.6725,
+      "step": 12170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8815924006751532,
+      "learning_rate": 1.2271510701012412e-06,
+      "loss": 0.7411,
+      "step": 12171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8706892759770979,
+      "learning_rate": 1.2264709877480147e-06,
+      "loss": 0.8006,
+      "step": 12172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6087745143529284,
+      "learning_rate": 1.2257910675517315e-06,
+      "loss": 0.6008,
+      "step": 12173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6008531096200052,
+      "learning_rate": 1.2251113095416113e-06,
+      "loss": 0.6082,
+      "step": 12174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7385457184784159,
+      "learning_rate": 1.2244317137468658e-06,
+      "loss": 0.7125,
+      "step": 12175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.581450809031163,
+      "learning_rate": 1.223752280196699e-06,
+      "loss": 0.6696,
+      "step": 12176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8766800983288048,
+      "learning_rate": 1.2230730089203063e-06,
+      "loss": 0.7581,
+      "step": 12177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5725301593710597,
+      "learning_rate": 1.2223938999468788e-06,
+      "loss": 0.7132,
+      "step": 12178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.751884283332595,
+      "learning_rate": 1.2217149533055976e-06,
+      "loss": 0.7711,
+      "step": 12179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8545823931348393,
+      "learning_rate": 1.22103616902564e-06,
+      "loss": 0.7455,
+      "step": 12180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.189781346829359,
+      "learning_rate": 1.2203575471361733e-06,
+      "loss": 0.7243,
+      "step": 12181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7452205384273922,
+      "learning_rate": 1.2196790876663617e-06,
+      "loss": 0.7599,
+      "step": 12182
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.001980594070462,
+      "learning_rate": 1.2190007906453605e-06,
+      "loss": 0.6453,
+      "step": 12183
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8958647078813382,
+      "learning_rate": 1.2183226561023132e-06,
+      "loss": 0.7208,
+      "step": 12184
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7114968539343827,
+      "learning_rate": 1.2176446840663647e-06,
+      "loss": 0.7941,
+      "step": 12185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.101649369817233,
+      "learning_rate": 1.216966874566648e-06,
+      "loss": 0.7628,
+      "step": 12186
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.187694036372792,
+      "learning_rate": 1.21628922763229e-06,
+      "loss": 0.8258,
+      "step": 12187
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.62069189116503,
+      "learning_rate": 1.2156117432924108e-06,
+      "loss": 0.721,
+      "step": 12188
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6790535336120636,
+      "learning_rate": 1.2149344215761216e-06,
+      "loss": 0.713,
+      "step": 12189
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6573579656567108,
+      "learning_rate": 1.2142572625125332e-06,
+      "loss": 0.7183,
+      "step": 12190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1759593884265545,
+      "learning_rate": 1.2135802661307388e-06,
+      "loss": 0.6183,
+      "step": 12191
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6229883031661103,
+      "learning_rate": 1.2129034324598343e-06,
+      "loss": 0.6314,
+      "step": 12192
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6551568558024854,
+      "learning_rate": 1.212226761528904e-06,
+      "loss": 0.6679,
+      "step": 12193
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7692786518704031,
+      "learning_rate": 1.2115502533670253e-06,
+      "loss": 0.6987,
+      "step": 12194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8740349935544889,
+      "learning_rate": 1.2108739080032695e-06,
+      "loss": 0.8199,
+      "step": 12195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.672595791742777,
+      "learning_rate": 1.2101977254667007e-06,
+      "loss": 0.7399,
+      "step": 12196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8116497549762514,
+      "learning_rate": 1.2095217057863762e-06,
+      "loss": 0.6166,
+      "step": 12197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.614262808015533,
+      "learning_rate": 1.2088458489913458e-06,
+      "loss": 0.7551,
+      "step": 12198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9308530819214251,
+      "learning_rate": 1.2081701551106506e-06,
+      "loss": 0.7168,
+      "step": 12199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6329841278248274,
+      "learning_rate": 1.207494624173331e-06,
+      "loss": 0.605,
+      "step": 12200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7863275670444843,
+      "learning_rate": 1.2068192562084146e-06,
+      "loss": 0.6613,
+      "step": 12201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1152132340172003,
+      "learning_rate": 1.2061440512449196e-06,
+      "loss": 0.5395,
+      "step": 12202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6798540959451305,
+      "learning_rate": 1.205469009311866e-06,
+      "loss": 0.7526,
+      "step": 12203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8681286761222462,
+      "learning_rate": 1.20479413043826e-06,
+      "loss": 0.657,
+      "step": 12204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4808722620402126,
+      "learning_rate": 1.2041194146531026e-06,
+      "loss": 0.721,
+      "step": 12205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8354707460949986,
+      "learning_rate": 1.2034448619853873e-06,
+      "loss": 0.7412,
+      "step": 12206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7106316557729866,
+      "learning_rate": 1.202770472464101e-06,
+      "loss": 0.7477,
+      "step": 12207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6884092627924063,
+      "learning_rate": 1.2020962461182268e-06,
+      "loss": 0.7291,
+      "step": 12208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.520565600739326,
+      "learning_rate": 1.201422182976732e-06,
+      "loss": 0.6424,
+      "step": 12209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7116599554279506,
+      "learning_rate": 1.200748283068588e-06,
+      "loss": 0.7373,
+      "step": 12210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8014223270588416,
+      "learning_rate": 1.2000745464227515e-06,
+      "loss": 0.7244,
+      "step": 12211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0367005458478427,
+      "learning_rate": 1.1994009730681744e-06,
+      "loss": 0.6386,
+      "step": 12212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7292440554380486,
+      "learning_rate": 1.198727563033802e-06,
+      "loss": 0.6556,
+      "step": 12213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7961555908637519,
+      "learning_rate": 1.1980543163485726e-06,
+      "loss": 0.7447,
+      "step": 12214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7249076294481083,
+      "learning_rate": 1.1973812330414159e-06,
+      "loss": 0.6026,
+      "step": 12215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6097894789265677,
+      "learning_rate": 1.1967083131412566e-06,
+      "loss": 0.6928,
+      "step": 12216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.728047161335896,
+      "learning_rate": 1.1960355566770104e-06,
+      "loss": 0.7899,
+      "step": 12217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1891249237071475,
+      "learning_rate": 1.1953629636775893e-06,
+      "loss": 0.6457,
+      "step": 12218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8066731526677355,
+      "learning_rate": 1.1946905341718951e-06,
+      "loss": 0.6898,
+      "step": 12219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9293697768967424,
+      "learning_rate": 1.1940182681888235e-06,
+      "loss": 0.7487,
+      "step": 12220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1704519712054802,
+      "learning_rate": 1.1933461657572637e-06,
+      "loss": 0.6701,
+      "step": 12221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6645586211751522,
+      "learning_rate": 1.1926742269060965e-06,
+      "loss": 0.6282,
+      "step": 12222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6401903343834137,
+      "learning_rate": 1.1920024516641976e-06,
+      "loss": 0.6565,
+      "step": 12223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5227210776539173,
+      "learning_rate": 1.1913308400604339e-06,
+      "loss": 0.7297,
+      "step": 12224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6626956262386252,
+      "learning_rate": 1.1906593921236649e-06,
+      "loss": 0.7342,
+      "step": 12225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6863848120247975,
+      "learning_rate": 1.1899881078827485e-06,
+      "loss": 0.7335,
+      "step": 12226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7918140937262765,
+      "learning_rate": 1.189316987366526e-06,
+      "loss": 0.7489,
+      "step": 12227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0753456797920338,
+      "learning_rate": 1.1886460306038405e-06,
+      "loss": 0.5322,
+      "step": 12228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.106150409831473,
+      "learning_rate": 1.1879752376235231e-06,
+      "loss": 0.6234,
+      "step": 12229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9625882479136787,
+      "learning_rate": 1.1873046084544004e-06,
+      "loss": 0.6496,
+      "step": 12230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7455747444461094,
+      "learning_rate": 1.1866341431252892e-06,
+      "loss": 0.5982,
+      "step": 12231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.818847830812802,
+      "learning_rate": 1.185963841665001e-06,
+      "loss": 0.6891,
+      "step": 12232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7189206920294722,
+      "learning_rate": 1.1852937041023433e-06,
+      "loss": 0.7765,
+      "step": 12233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9590864959517817,
+      "learning_rate": 1.1846237304661095e-06,
+      "loss": 0.7854,
+      "step": 12234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9219323252451006,
+      "learning_rate": 1.1839539207850898e-06,
+      "loss": 0.6593,
+      "step": 12235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.673831416052096,
+      "learning_rate": 1.1832842750880702e-06,
+      "loss": 0.7334,
+      "step": 12236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.588064347099677,
+      "learning_rate": 1.1826147934038262e-06,
+      "loss": 0.7184,
+      "step": 12237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7753877121217971,
+      "learning_rate": 1.181945475761126e-06,
+      "loss": 0.7948,
+      "step": 12238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2231965796985311,
+      "learning_rate": 1.181276322188732e-06,
+      "loss": 0.6629,
+      "step": 12239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6820516633364522,
+      "learning_rate": 1.1806073327153988e-06,
+      "loss": 0.7199,
+      "step": 12240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.101608078041351,
+      "learning_rate": 1.1799385073698749e-06,
+      "loss": 0.6066,
+      "step": 12241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6364247449370373,
+      "learning_rate": 1.1792698461809e-06,
+      "loss": 0.7775,
+      "step": 12242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6271367688946015,
+      "learning_rate": 1.1786013491772103e-06,
+      "loss": 0.8469,
+      "step": 12243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6633054933231932,
+      "learning_rate": 1.1779330163875325e-06,
+      "loss": 0.7279,
+      "step": 12244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6582869886654157,
+      "learning_rate": 1.1772648478405824e-06,
+      "loss": 0.7552,
+      "step": 12245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5637951981480684,
+      "learning_rate": 1.176596843565077e-06,
+      "loss": 0.7058,
+      "step": 12246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5722122497261544,
+      "learning_rate": 1.1759290035897208e-06,
+      "loss": 0.7088,
+      "step": 12247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.768077797315996,
+      "learning_rate": 1.1752613279432112e-06,
+      "loss": 0.7135,
+      "step": 12248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8669284537043427,
+      "learning_rate": 1.1745938166542414e-06,
+      "loss": 0.768,
+      "step": 12249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1488024841605518,
+      "learning_rate": 1.173926469751493e-06,
+      "loss": 0.6854,
+      "step": 12250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.974293438220182,
+      "learning_rate": 1.1732592872636478e-06,
+      "loss": 0.9022,
+      "step": 12251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9850801754599476,
+      "learning_rate": 1.172592269219373e-06,
+      "loss": 0.7987,
+      "step": 12252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.239247789385002,
+      "learning_rate": 1.1719254156473302e-06,
+      "loss": 0.696,
+      "step": 12253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5723356353390476,
+      "learning_rate": 1.1712587265761799e-06,
+      "loss": 0.7376,
+      "step": 12254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8991472407264451,
+      "learning_rate": 1.1705922020345688e-06,
+      "loss": 0.7473,
+      "step": 12255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7269342076368108,
+      "learning_rate": 1.16992584205114e-06,
+      "loss": 0.693,
+      "step": 12256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.258848404589107,
+      "learning_rate": 1.1692596466545275e-06,
+      "loss": 0.6575,
+      "step": 12257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0768917495250945,
+      "learning_rate": 1.168593615873358e-06,
+      "loss": 0.6364,
+      "step": 12258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5590645132776226,
+      "learning_rate": 1.1679277497362563e-06,
+      "loss": 0.7034,
+      "step": 12259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.566698960695535,
+      "learning_rate": 1.167262048271831e-06,
+      "loss": 0.6467,
+      "step": 12260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.082090597105129,
+      "learning_rate": 1.166596511508693e-06,
+      "loss": 0.655,
+      "step": 12261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3093685840031775,
+      "learning_rate": 1.1659311394754418e-06,
+      "loss": 0.6553,
+      "step": 12262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9045794268668812,
+      "learning_rate": 1.1652659322006659e-06,
+      "loss": 0.736,
+      "step": 12263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5530580325478347,
+      "learning_rate": 1.1646008897129546e-06,
+      "loss": 0.7486,
+      "step": 12264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6157578246389122,
+      "learning_rate": 1.1639360120408843e-06,
+      "loss": 0.682,
+      "step": 12265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7925266310957124,
+      "learning_rate": 1.1632712992130278e-06,
+      "loss": 0.7069,
+      "step": 12266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.851852081707394,
+      "learning_rate": 1.1626067512579481e-06,
+      "loss": 0.6365,
+      "step": 12267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.596539955108654,
+      "learning_rate": 1.1619423682042008e-06,
+      "loss": 0.6466,
+      "step": 12268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.58645348124421,
+      "learning_rate": 1.161278150080341e-06,
+      "loss": 0.6134,
+      "step": 12269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7826466108337644,
+      "learning_rate": 1.160614096914906e-06,
+      "loss": 0.7222,
+      "step": 12270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.155594353892775,
+      "learning_rate": 1.1599502087364345e-06,
+      "loss": 0.6874,
+      "step": 12271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6634267100541709,
+      "learning_rate": 1.1592864855734553e-06,
+      "loss": 0.6714,
+      "step": 12272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.635617502212792,
+      "learning_rate": 1.1586229274544898e-06,
+      "loss": 0.6533,
+      "step": 12273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.573388134254371,
+      "learning_rate": 1.157959534408052e-06,
+      "loss": 0.6523,
+      "step": 12274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8370537387232926,
+      "learning_rate": 1.1572963064626497e-06,
+      "loss": 0.6188,
+      "step": 12275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.752053951989885,
+      "learning_rate": 1.1566332436467815e-06,
+      "loss": 0.7541,
+      "step": 12276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7557905180103561,
+      "learning_rate": 1.155970345988946e-06,
+      "loss": 0.7429,
+      "step": 12277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.841862373049728,
+      "learning_rate": 1.1553076135176222e-06,
+      "loss": 0.7913,
+      "step": 12278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.017146781424179,
+      "learning_rate": 1.1546450462612951e-06,
+      "loss": 0.7272,
+      "step": 12279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.700270803945642,
+      "learning_rate": 1.1539826442484337e-06,
+      "loss": 0.7627,
+      "step": 12280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0287224172672675,
+      "learning_rate": 1.1533204075075039e-06,
+      "loss": 0.763,
+      "step": 12281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8238960202461791,
+      "learning_rate": 1.1526583360669635e-06,
+      "loss": 0.659,
+      "step": 12282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6046830704196993,
+      "learning_rate": 1.1519964299552611e-06,
+      "loss": 0.6187,
+      "step": 12283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4711445603947908,
+      "learning_rate": 1.151334689200845e-06,
+      "loss": 0.611,
+      "step": 12284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8009414577244771,
+      "learning_rate": 1.1506731138321474e-06,
+      "loss": 0.8085,
+      "step": 12285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7425029892741322,
+      "learning_rate": 1.1500117038775977e-06,
+      "loss": 0.7673,
+      "step": 12286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3253184536989895,
+      "learning_rate": 1.149350459365622e-06,
+      "loss": 0.6805,
+      "step": 12287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.993745270460723,
+      "learning_rate": 1.1486893803246302e-06,
+      "loss": 0.5863,
+      "step": 12288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5964138030068198,
+      "learning_rate": 1.1480284667830343e-06,
+      "loss": 0.8043,
+      "step": 12289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8863421923408406,
+      "learning_rate": 1.1473677187692345e-06,
+      "loss": 0.6953,
+      "step": 12290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9177670333407142,
+      "learning_rate": 1.1467071363116234e-06,
+      "loss": 0.8589,
+      "step": 12291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6324464900153195,
+      "learning_rate": 1.1460467194385889e-06,
+      "loss": 0.8579,
+      "step": 12292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7151730130637484,
+      "learning_rate": 1.1453864681785087e-06,
+      "loss": 0.7602,
+      "step": 12293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6622366795622814,
+      "learning_rate": 1.1447263825597577e-06,
+      "loss": 0.7081,
+      "step": 12294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6950430015644438,
+      "learning_rate": 1.1440664626107017e-06,
+      "loss": 0.6122,
+      "step": 12295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.999612530767629,
+      "learning_rate": 1.143406708359695e-06,
+      "loss": 0.7241,
+      "step": 12296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8866995700694467,
+      "learning_rate": 1.142747119835092e-06,
+      "loss": 0.654,
+      "step": 12297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1235564124043336,
+      "learning_rate": 1.1420876970652361e-06,
+      "loss": 0.7335,
+      "step": 12298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.161353519685229,
+      "learning_rate": 1.1414284400784643e-06,
+      "loss": 0.5928,
+      "step": 12299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7381136864759683,
+      "learning_rate": 1.1407693489031052e-06,
+      "loss": 0.6841,
+      "step": 12300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.601662852198008,
+      "learning_rate": 1.1401104235674808e-06,
+      "loss": 0.6833,
+      "step": 12301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6100437555492677,
+      "learning_rate": 1.1394516640999114e-06,
+      "loss": 0.6965,
+      "step": 12302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7025930558908897,
+      "learning_rate": 1.1387930705286998e-06,
+      "loss": 0.8435,
+      "step": 12303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4915206089555606,
+      "learning_rate": 1.1381346428821482e-06,
+      "loss": 0.646,
+      "step": 12304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8779992457201224,
+      "learning_rate": 1.1374763811885547e-06,
+      "loss": 0.7281,
+      "step": 12305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1939367410819983,
+      "learning_rate": 1.1368182854762005e-06,
+      "loss": 0.6939,
+      "step": 12306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7452726895911135,
+      "learning_rate": 1.1361603557733698e-06,
+      "loss": 0.6753,
+      "step": 12307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7044342755944462,
+      "learning_rate": 1.135502592108334e-06,
+      "loss": 0.6943,
+      "step": 12308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.597469405989656,
+      "learning_rate": 1.134844994509358e-06,
+      "loss": 0.672,
+      "step": 12309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9459480617643778,
+      "learning_rate": 1.1341875630047012e-06,
+      "loss": 0.5871,
+      "step": 12310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.567017370574516,
+      "learning_rate": 1.1335302976226132e-06,
+      "loss": 0.7285,
+      "step": 12311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4615797344332013,
+      "learning_rate": 1.1328731983913404e-06,
+      "loss": 0.6922,
+      "step": 12312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5066056916201527,
+      "learning_rate": 1.13221626533912e-06,
+      "loss": 0.7057,
+      "step": 12313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7688279686679809,
+      "learning_rate": 1.1315594984941786e-06,
+      "loss": 0.7483,
+      "step": 12314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4211157132303174,
+      "learning_rate": 1.1309028978847419e-06,
+      "loss": 0.6789,
+      "step": 12315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0067808407491885,
+      "learning_rate": 1.1302464635390247e-06,
+      "loss": 0.693,
+      "step": 12316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8537246714430422,
+      "learning_rate": 1.1295901954852356e-06,
+      "loss": 0.6663,
+      "step": 12317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7741871400225502,
+      "learning_rate": 1.1289340937515753e-06,
+      "loss": 0.664,
+      "step": 12318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6838409894005089,
+      "learning_rate": 1.1282781583662372e-06,
+      "loss": 0.7022,
+      "step": 12319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0475920128278517,
+      "learning_rate": 1.1276223893574123e-06,
+      "loss": 0.6545,
+      "step": 12320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6021948468259835,
+      "learning_rate": 1.1269667867532746e-06,
+      "loss": 0.6953,
+      "step": 12321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6472976918137519,
+      "learning_rate": 1.1263113505820012e-06,
+      "loss": 0.6693,
+      "step": 12322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7699204694498374,
+      "learning_rate": 1.1256560808717576e-06,
+      "loss": 0.7779,
+      "step": 12323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6884809689600013,
+      "learning_rate": 1.1250009776506982e-06,
+      "loss": 0.7395,
+      "step": 12324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8955144774974388,
+      "learning_rate": 1.124346040946978e-06,
+      "loss": 0.6783,
+      "step": 12325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2498240018390843,
+      "learning_rate": 1.1236912707887404e-06,
+      "loss": 0.6781,
+      "step": 12326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2625177501540028,
+      "learning_rate": 1.1230366672041216e-06,
+      "loss": 0.6458,
+      "step": 12327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1534861988121596,
+      "learning_rate": 1.1223822302212517e-06,
+      "loss": 0.6356,
+      "step": 12328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5666673374935272,
+      "learning_rate": 1.1217279598682518e-06,
+      "loss": 0.6091,
+      "step": 12329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1363211508136692,
+      "learning_rate": 1.1210738561732405e-06,
+      "loss": 0.7913,
+      "step": 12330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4404293245620836,
+      "learning_rate": 1.120419919164324e-06,
+      "loss": 0.6635,
+      "step": 12331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1787186269754035,
+      "learning_rate": 1.1197661488696043e-06,
+      "loss": 0.651,
+      "step": 12332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1514549658450062,
+      "learning_rate": 1.1191125453171748e-06,
+      "loss": 0.6905,
+      "step": 12333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7233087137401668,
+      "learning_rate": 1.118459108535122e-06,
+      "loss": 0.6771,
+      "step": 12334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6619797266320404,
+      "learning_rate": 1.1178058385515268e-06,
+      "loss": 0.6265,
+      "step": 12335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5347489816818447,
+      "learning_rate": 1.1171527353944601e-06,
+      "loss": 0.72,
+      "step": 12336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6645400024613561,
+      "learning_rate": 1.1164997990919868e-06,
+      "loss": 0.6804,
+      "step": 12337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8044593115912986,
+      "learning_rate": 1.1158470296721691e-06,
+      "loss": 0.7681,
+      "step": 12338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.634464551580171,
+      "learning_rate": 1.1151944271630517e-06,
+      "loss": 0.7116,
+      "step": 12339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.63318980075215,
+      "learning_rate": 1.1145419915926836e-06,
+      "loss": 0.7677,
+      "step": 12340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6790126023649357,
+      "learning_rate": 1.1138897229890995e-06,
+      "loss": 0.6584,
+      "step": 12341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2520702062927351,
+      "learning_rate": 1.1132376213803286e-06,
+      "loss": 0.6903,
+      "step": 12342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.87338219869429,
+      "learning_rate": 1.1125856867943936e-06,
+      "loss": 0.6595,
+      "step": 12343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7594290311053693,
+      "learning_rate": 1.1119339192593077e-06,
+      "loss": 0.6182,
+      "step": 12344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.699834350266058,
+      "learning_rate": 1.111282318803083e-06,
+      "loss": 0.6975,
+      "step": 12345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0282713156320282,
+      "learning_rate": 1.1106308854537167e-06,
+      "loss": 0.7391,
+      "step": 12346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5769266425625714,
+      "learning_rate": 1.1099796192392021e-06,
+      "loss": 0.7417,
+      "step": 12347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7528555284855607,
+      "learning_rate": 1.109328520187528e-06,
+      "loss": 0.6905,
+      "step": 12348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7589737267028591,
+      "learning_rate": 1.1086775883266725e-06,
+      "loss": 0.6967,
+      "step": 12349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.616706383554804,
+      "learning_rate": 1.1080268236846076e-06,
+      "loss": 0.7926,
+      "step": 12350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5676023456278283,
+      "learning_rate": 1.1073762262892979e-06,
+      "loss": 0.7151,
+      "step": 12351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.153153345334844,
+      "learning_rate": 1.1067257961687005e-06,
+      "loss": 0.953,
+      "step": 12352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6381357785712793,
+      "learning_rate": 1.1060755333507673e-06,
+      "loss": 0.6436,
+      "step": 12353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9443241873588043,
+      "learning_rate": 1.1054254378634399e-06,
+      "loss": 0.6628,
+      "step": 12354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9946967149996728,
+      "learning_rate": 1.1047755097346541e-06,
+      "loss": 0.7284,
+      "step": 12355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6853960894634834,
+      "learning_rate": 1.1041257489923429e-06,
+      "loss": 0.7571,
+      "step": 12356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.582626008827606,
+      "learning_rate": 1.1034761556644214e-06,
+      "loss": 0.6154,
+      "step": 12357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.871928506677979,
+      "learning_rate": 1.1028267297788092e-06,
+      "loss": 0.6799,
+      "step": 12358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.78060457092216,
+      "learning_rate": 1.102177471363412e-06,
+      "loss": 0.7219,
+      "step": 12359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6399222003983962,
+      "learning_rate": 1.1015283804461297e-06,
+      "loss": 0.6835,
+      "step": 12360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5551991522786741,
+      "learning_rate": 1.1008794570548554e-06,
+      "loss": 0.6069,
+      "step": 12361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1058162680306793,
+      "learning_rate": 1.100230701217473e-06,
+      "loss": 0.6609,
+      "step": 12362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.008028650891276,
+      "learning_rate": 1.0995821129618656e-06,
+      "loss": 0.7733,
+      "step": 12363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0679115015572296,
+      "learning_rate": 1.0989336923158999e-06,
+      "loss": 0.6747,
+      "step": 12364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3124891337120843,
+      "learning_rate": 1.0982854393074399e-06,
+      "loss": 0.7631,
+      "step": 12365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.789325904294627,
+      "learning_rate": 1.0976373539643465e-06,
+      "loss": 0.6262,
+      "step": 12366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6875077556074425,
+      "learning_rate": 1.0969894363144661e-06,
+      "loss": 0.7523,
+      "step": 12367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7497146132775914,
+      "learning_rate": 1.0963416863856425e-06,
+      "loss": 0.8007,
+      "step": 12368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5221021113850417,
+      "learning_rate": 1.0956941042057106e-06,
+      "loss": 0.6758,
+      "step": 12369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4745490350871748,
+      "learning_rate": 1.095046689802498e-06,
+      "loss": 0.6511,
+      "step": 12370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0206426188102413,
+      "learning_rate": 1.0943994432038262e-06,
+      "loss": 0.5345,
+      "step": 12371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.424453466551243,
+      "learning_rate": 1.0937523644375076e-06,
+      "loss": 0.7852,
+      "step": 12372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6938532706889522,
+      "learning_rate": 1.0931054535313507e-06,
+      "loss": 0.7994,
+      "step": 12373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7041535908790333,
+      "learning_rate": 1.0924587105131546e-06,
+      "loss": 0.6444,
+      "step": 12374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0993338695228485,
+      "learning_rate": 1.0918121354107076e-06,
+      "loss": 0.827,
+      "step": 12375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.919916976657535,
+      "learning_rate": 1.091165728251799e-06,
+      "loss": 0.5584,
+      "step": 12376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.717140501583914,
+      "learning_rate": 1.090519489064204e-06,
+      "loss": 0.7678,
+      "step": 12377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7871429233453326,
+      "learning_rate": 1.0898734178756937e-06,
+      "loss": 0.7546,
+      "step": 12378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5883738719143896,
+      "learning_rate": 1.0892275147140307e-06,
+      "loss": 0.6544,
+      "step": 12379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7484187821519055,
+      "learning_rate": 1.0885817796069697e-06,
+      "loss": 0.7979,
+      "step": 12380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8763206233078826,
+      "learning_rate": 1.087936212582264e-06,
+      "loss": 0.7002,
+      "step": 12381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.040578468677363,
+      "learning_rate": 1.0872908136676491e-06,
+      "loss": 0.7048,
+      "step": 12382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6901678696670213,
+      "learning_rate": 1.0866455828908634e-06,
+      "loss": 0.7385,
+      "step": 12383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2368657740635145,
+      "learning_rate": 1.086000520279632e-06,
+      "loss": 0.7466,
+      "step": 12384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9848251152586764,
+      "learning_rate": 1.0853556258616755e-06,
+      "loss": 0.7642,
+      "step": 12385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9963413669450973,
+      "learning_rate": 1.0847108996647066e-06,
+      "loss": 0.7408,
+      "step": 12386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4638281172769514,
+      "learning_rate": 1.0840663417164293e-06,
+      "loss": 0.7448,
+      "step": 12387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5750580007785444,
+      "learning_rate": 1.0834219520445432e-06,
+      "loss": 0.7696,
+      "step": 12388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8901613791707812,
+      "learning_rate": 1.0827777306767384e-06,
+      "loss": 0.7086,
+      "step": 12389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9858649443675127,
+      "learning_rate": 1.082133677640697e-06,
+      "loss": 0.583,
+      "step": 12390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7410347390618863,
+      "learning_rate": 1.0814897929640984e-06,
+      "loss": 0.8273,
+      "step": 12391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8331173618888508,
+      "learning_rate": 1.0808460766746104e-06,
+      "loss": 0.6357,
+      "step": 12392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0413652785086511,
+      "learning_rate": 1.0802025287998953e-06,
+      "loss": 0.7046,
+      "step": 12393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9563546181124818,
+      "learning_rate": 1.0795591493676072e-06,
+      "loss": 0.6483,
+      "step": 12394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6890409170967116,
+      "learning_rate": 1.078915938405392e-06,
+      "loss": 0.6566,
+      "step": 12395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4113615742453856,
+      "learning_rate": 1.078272895940895e-06,
+      "loss": 0.6658,
+      "step": 12396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.534638344626512,
+      "learning_rate": 1.0776300220017437e-06,
+      "loss": 0.6361,
+      "step": 12397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.782683661564746,
+      "learning_rate": 1.0769873166155653e-06,
+      "loss": 0.6883,
+      "step": 12398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6465938454355489,
+      "learning_rate": 1.0763447798099813e-06,
+      "loss": 0.6823,
+      "step": 12399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2102240321148274,
+      "learning_rate": 1.075702411612598e-06,
+      "loss": 0.6454,
+      "step": 12400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9316020501239048,
+      "learning_rate": 1.0750602120510233e-06,
+      "loss": 0.5961,
+      "step": 12401
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6000187917880753,
+      "learning_rate": 1.0744181811528526e-06,
+      "loss": 0.7246,
+      "step": 12402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.711872837417398,
+      "learning_rate": 1.073776318945675e-06,
+      "loss": 0.7465,
+      "step": 12403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8636847473212335,
+      "learning_rate": 1.0731346254570735e-06,
+      "loss": 0.6948,
+      "step": 12404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7368115517271758,
+      "learning_rate": 1.0724931007146227e-06,
+      "loss": 0.6585,
+      "step": 12405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6184873782585065,
+      "learning_rate": 1.0718517447458904e-06,
+      "loss": 0.6328,
+      "step": 12406
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6567784943066137,
+      "learning_rate": 1.071210557578437e-06,
+      "loss": 0.6891,
+      "step": 12407
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7082425239530639,
+      "learning_rate": 1.0705695392398146e-06,
+      "loss": 0.6505,
+      "step": 12408
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5318892989763353,
+      "learning_rate": 1.0699286897575718e-06,
+      "loss": 0.7431,
+      "step": 12409
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6941760329043767,
+      "learning_rate": 1.0692880091592457e-06,
+      "loss": 0.783,
+      "step": 12410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.371780506926364,
+      "learning_rate": 1.068647497472368e-06,
+      "loss": 0.6444,
+      "step": 12411
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.222750468210723,
+      "learning_rate": 1.0680071547244636e-06,
+      "loss": 0.6729,
+      "step": 12412
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5368983268181784,
+      "learning_rate": 1.0673669809430465e-06,
+      "loss": 0.637,
+      "step": 12413
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7004210359381244,
+      "learning_rate": 1.066726976155632e-06,
+      "loss": 0.7605,
+      "step": 12414
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8492401646864867,
+      "learning_rate": 1.0660871403897177e-06,
+      "loss": 0.7032,
+      "step": 12415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5799192694745428,
+      "learning_rate": 1.0654474736727988e-06,
+      "loss": 0.7464,
+      "step": 12416
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4066077766584577,
+      "learning_rate": 1.0648079760323675e-06,
+      "loss": 0.6418,
+      "step": 12417
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.457561923057859,
+      "learning_rate": 1.064168647495899e-06,
+      "loss": 0.6535,
+      "step": 12418
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.446735551623609,
+      "learning_rate": 1.0635294880908702e-06,
+      "loss": 0.6695,
+      "step": 12419
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6413684498007848,
+      "learning_rate": 1.0628904978447463e-06,
+      "loss": 0.6988,
+      "step": 12420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5997825101839973,
+      "learning_rate": 1.0622516767849867e-06,
+      "loss": 0.6552,
+      "step": 12421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5913997061592449,
+      "learning_rate": 1.0616130249390423e-06,
+      "loss": 0.7535,
+      "step": 12422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6872116826509778,
+      "learning_rate": 1.0609745423343553e-06,
+      "loss": 0.7232,
+      "step": 12423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0501889565152562,
+      "learning_rate": 1.0603362289983687e-06,
+      "loss": 0.7442,
+      "step": 12424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7749723400714346,
+      "learning_rate": 1.0596980849585065e-06,
+      "loss": 0.7733,
+      "step": 12425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9029062358924291,
+      "learning_rate": 1.0590601102421916e-06,
+      "loss": 0.7535,
+      "step": 12426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6329135409920708,
+      "learning_rate": 1.0584223048768427e-06,
+      "loss": 0.6711,
+      "step": 12427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.208960516063149,
+      "learning_rate": 1.0577846688898652e-06,
+      "loss": 0.7223,
+      "step": 12428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.016526496562999,
+      "learning_rate": 1.0571472023086604e-06,
+      "loss": 0.799,
+      "step": 12429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6540743918409782,
+      "learning_rate": 1.0565099051606214e-06,
+      "loss": 0.6916,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9697018458104268,
+      "learning_rate": 1.055872777473133e-06,
+      "loss": 0.6017,
+      "step": 12431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.613621069601389,
+      "learning_rate": 1.0552358192735784e-06,
+      "loss": 0.6975,
+      "step": 12432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.266758896384049,
+      "learning_rate": 1.0545990305893233e-06,
+      "loss": 0.7529,
+      "step": 12433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4104021367741497,
+      "learning_rate": 1.053962411447736e-06,
+      "loss": 0.6571,
+      "step": 12434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6971710241283213,
+      "learning_rate": 1.0533259618761738e-06,
+      "loss": 0.63,
+      "step": 12435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2430873003993574,
+      "learning_rate": 1.0526896819019817e-06,
+      "loss": 0.7026,
+      "step": 12436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7730597851722574,
+      "learning_rate": 1.0520535715525066e-06,
+      "loss": 0.6208,
+      "step": 12437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5025875406514924,
+      "learning_rate": 1.0514176308550817e-06,
+      "loss": 0.6576,
+      "step": 12438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.244343636316278,
+      "learning_rate": 1.0507818598370355e-06,
+      "loss": 0.7947,
+      "step": 12439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9875277091247573,
+      "learning_rate": 1.0501462585256883e-06,
+      "loss": 0.6535,
+      "step": 12440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.593824839528554,
+      "learning_rate": 1.0495108269483518e-06,
+      "loss": 0.7173,
+      "step": 12441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9341556734506655,
+      "learning_rate": 1.0488755651323358e-06,
+      "loss": 0.7081,
+      "step": 12442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8579957413863497,
+      "learning_rate": 1.048240473104934e-06,
+      "loss": 0.8038,
+      "step": 12443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.717285803207686,
+      "learning_rate": 1.0476055508934408e-06,
+      "loss": 0.6216,
+      "step": 12444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.777599935474516,
+      "learning_rate": 1.04697079852514e-06,
+      "loss": 0.7915,
+      "step": 12445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6680379994332626,
+      "learning_rate": 1.0463362160273076e-06,
+      "loss": 0.6595,
+      "step": 12446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9763751313525597,
+      "learning_rate": 1.0457018034272136e-06,
+      "loss": 0.8413,
+      "step": 12447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5674464310956857,
+      "learning_rate": 1.04506756075212e-06,
+      "loss": 0.6413,
+      "step": 12448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.699976372740454,
+      "learning_rate": 1.0444334880292794e-06,
+      "loss": 0.71,
+      "step": 12449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.654765015407761,
+      "learning_rate": 1.0437995852859445e-06,
+      "loss": 0.6132,
+      "step": 12450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6452395625816436,
+      "learning_rate": 1.0431658525493498e-06,
+      "loss": 0.7339,
+      "step": 12451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8294182843757445,
+      "learning_rate": 1.0425322898467315e-06,
+      "loss": 0.6866,
+      "step": 12452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7340384062574121,
+      "learning_rate": 1.0418988972053162e-06,
+      "loss": 0.8335,
+      "step": 12453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.578679585542662,
+      "learning_rate": 1.0412656746523182e-06,
+      "loss": 0.6634,
+      "step": 12454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5290397573850787,
+      "learning_rate": 1.0406326222149516e-06,
+      "loss": 0.6351,
+      "step": 12455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.655306875012225,
+      "learning_rate": 1.0399997399204186e-06,
+      "loss": 0.8615,
+      "step": 12456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3729518942148644,
+      "learning_rate": 1.0393670277959174e-06,
+      "loss": 0.6417,
+      "step": 12457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3330964239957783,
+      "learning_rate": 1.038734485868635e-06,
+      "loss": 0.6422,
+      "step": 12458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9960284791075502,
+      "learning_rate": 1.0381021141657526e-06,
+      "loss": 0.6633,
+      "step": 12459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6647101599829612,
+      "learning_rate": 1.037469912714449e-06,
+      "loss": 0.6139,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8545399226140444,
+      "learning_rate": 1.0368378815418856e-06,
+      "loss": 0.7641,
+      "step": 12461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.284395398897979,
+      "learning_rate": 1.036206020675226e-06,
+      "loss": 0.6381,
+      "step": 12462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5279126384001975,
+      "learning_rate": 1.0355743301416215e-06,
+      "loss": 0.6257,
+      "step": 12463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6767406685575816,
+      "learning_rate": 1.0349428099682173e-06,
+      "loss": 0.6642,
+      "step": 12464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8111757083823128,
+      "learning_rate": 1.0343114601821513e-06,
+      "loss": 0.7049,
+      "step": 12465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7495669132041838,
+      "learning_rate": 1.033680280810554e-06,
+      "loss": 0.7422,
+      "step": 12466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9153156913734526,
+      "learning_rate": 1.0330492718805469e-06,
+      "loss": 0.761,
+      "step": 12467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8409995599518234,
+      "learning_rate": 1.0324184334192505e-06,
+      "loss": 0.6478,
+      "step": 12468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5927796084127208,
+      "learning_rate": 1.0317877654537672e-06,
+      "loss": 0.6654,
+      "step": 12469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3883300020456952,
+      "learning_rate": 1.031157268011203e-06,
+      "loss": 0.67,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5128095040379557,
+      "learning_rate": 1.0305269411186502e-06,
+      "loss": 0.731,
+      "step": 12471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8155238890418617,
+      "learning_rate": 1.0298967848031948e-06,
+      "loss": 0.8128,
+      "step": 12472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5731573148813207,
+      "learning_rate": 1.0292667990919164e-06,
+      "loss": 0.7678,
+      "step": 12473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7120681063452121,
+      "learning_rate": 1.0286369840118859e-06,
+      "loss": 0.7949,
+      "step": 12474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8984801823194584,
+      "learning_rate": 1.0280073395901719e-06,
+      "loss": 0.6963,
+      "step": 12475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7902526292753416,
+      "learning_rate": 1.0273778658538263e-06,
+      "loss": 0.6797,
+      "step": 12476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8873442487557701,
+      "learning_rate": 1.0267485628299007e-06,
+      "loss": 0.7252,
+      "step": 12477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.733757422721628,
+      "learning_rate": 1.026119430545441e-06,
+      "loss": 0.7578,
+      "step": 12478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.531015792892741,
+      "learning_rate": 1.025490469027477e-06,
+      "loss": 0.7027,
+      "step": 12479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9105849794053569,
+      "learning_rate": 1.0248616783030403e-06,
+      "loss": 0.6322,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7443731893838852,
+      "learning_rate": 1.0242330583991507e-06,
+      "loss": 0.6873,
+      "step": 12481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4804050809404545,
+      "learning_rate": 1.0236046093428204e-06,
+      "loss": 0.7735,
+      "step": 12482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3695091215668254,
+      "learning_rate": 1.0229763311610563e-06,
+      "loss": 0.7087,
+      "step": 12483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9163808849245145,
+      "learning_rate": 1.0223482238808557e-06,
+      "loss": 0.7128,
+      "step": 12484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1670606148307303,
+      "learning_rate": 1.0217202875292115e-06,
+      "loss": 0.6583,
+      "step": 12485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5490286827689914,
+      "learning_rate": 1.021092522133108e-06,
+      "loss": 0.6966,
+      "step": 12486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6925990290785007,
+      "learning_rate": 1.0204649277195178e-06,
+      "loss": 0.8555,
+      "step": 12487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6803267285601369,
+      "learning_rate": 1.0198375043154142e-06,
+      "loss": 0.8276,
+      "step": 12488
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8047375977963158,
+      "learning_rate": 1.0192102519477565e-06,
+      "loss": 0.7083,
+      "step": 12489
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3813016551580324,
+      "learning_rate": 1.0185831706435007e-06,
+      "loss": 0.7166,
+      "step": 12490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5593558988262095,
+      "learning_rate": 1.0179562604295929e-06,
+      "loss": 0.7073,
+      "step": 12491
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7620796707836741,
+      "learning_rate": 1.0173295213329714e-06,
+      "loss": 0.7138,
+      "step": 12492
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6402906723028736,
+      "learning_rate": 1.0167029533805733e-06,
+      "loss": 0.7192,
+      "step": 12493
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.773495561558609,
+      "learning_rate": 1.016076556599318e-06,
+      "loss": 0.6863,
+      "step": 12494
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3979679818419646,
+      "learning_rate": 1.0154503310161269e-06,
+      "loss": 0.6353,
+      "step": 12495
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6153078610560079,
+      "learning_rate": 1.014824276657909e-06,
+      "loss": 0.6237,
+      "step": 12496
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9707502803419152,
+      "learning_rate": 1.0141983935515675e-06,
+      "loss": 0.5797,
+      "step": 12497
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6972548134441736,
+      "learning_rate": 1.0135726817239983e-06,
+      "loss": 0.6327,
+      "step": 12498
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5407883715555535,
+      "learning_rate": 1.0129471412020886e-06,
+      "loss": 0.6603,
+      "step": 12499
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.64456808793813,
+      "learning_rate": 1.0123217720127203e-06,
+      "loss": 0.8286,
+      "step": 12500
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6264527586729176,
+      "learning_rate": 1.011696574182766e-06,
+      "loss": 0.8178,
+      "step": 12501
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6157407391938479,
+      "learning_rate": 1.0110715477390915e-06,
+      "loss": 0.7372,
+      "step": 12502
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5939468937036225,
+      "learning_rate": 1.0104466927085577e-06,
+      "loss": 0.6223,
+      "step": 12503
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7321897622461808,
+      "learning_rate": 1.0098220091180145e-06,
+      "loss": 0.7145,
+      "step": 12504
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8712754037984114,
+      "learning_rate": 1.0091974969943064e-06,
+      "loss": 0.7271,
+      "step": 12505
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6132792507153961,
+      "learning_rate": 1.0085731563642697e-06,
+      "loss": 0.6492,
+      "step": 12506
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8058446087336282,
+      "learning_rate": 1.0079489872547338e-06,
+      "loss": 0.9046,
+      "step": 12507
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5787700540604517,
+      "learning_rate": 1.0073249896925208e-06,
+      "loss": 0.6336,
+      "step": 12508
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7835810948678013,
+      "learning_rate": 1.006701163704445e-06,
+      "loss": 0.6965,
+      "step": 12509
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5449137341175032,
+      "learning_rate": 1.0060775093173126e-06,
+      "loss": 0.6505,
+      "step": 12510
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6378293920095215,
+      "learning_rate": 1.005454026557927e-06,
+      "loss": 0.6277,
+      "step": 12511
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7176366490634718,
+      "learning_rate": 1.0048307154530763e-06,
+      "loss": 0.6976,
+      "step": 12512
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7936297999978348,
+      "learning_rate": 1.0042075760295477e-06,
+      "loss": 0.7327,
+      "step": 12513
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0513340073491741,
+      "learning_rate": 1.0035846083141193e-06,
+      "loss": 0.6411,
+      "step": 12514
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.349832033540828,
+      "learning_rate": 1.002961812333561e-06,
+      "loss": 0.8219,
+      "step": 12515
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.853935305697888,
+      "learning_rate": 1.0023391881146349e-06,
+      "loss": 0.7295,
+      "step": 12516
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8135712322665478,
+      "learning_rate": 1.0017167356840974e-06,
+      "loss": 0.6714,
+      "step": 12517
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8033972068460806,
+      "learning_rate": 1.0010944550686968e-06,
+      "loss": 0.8133,
+      "step": 12518
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6430180548379276,
+      "learning_rate": 1.0004723462951732e-06,
+      "loss": 0.7076,
+      "step": 12519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.134526512476235,
+      "learning_rate": 9.99850409390259e-07,
+      "loss": 0.5901,
+      "step": 12520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5132864971291882,
+      "learning_rate": 9.992286443806825e-07,
+      "loss": 0.6513,
+      "step": 12521
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0588082048751253,
+      "learning_rate": 9.986070512931616e-07,
+      "loss": 0.6555,
+      "step": 12522
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.050398317574661,
+      "learning_rate": 9.97985630154407e-07,
+      "loss": 0.8357,
+      "step": 12523
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.732334022159222,
+      "learning_rate": 9.973643809911238e-07,
+      "loss": 0.7472,
+      "step": 12524
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6688721127071608,
+      "learning_rate": 9.967433038300067e-07,
+      "loss": 0.6965,
+      "step": 12525
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6801480454879694,
+      "learning_rate": 9.96122398697746e-07,
+      "loss": 0.6538,
+      "step": 12526
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.473769748582389,
+      "learning_rate": 9.955016656210226e-07,
+      "loss": 0.5922,
+      "step": 12527
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8778935833847024,
+      "learning_rate": 9.948811046265095e-07,
+      "loss": 0.7416,
+      "step": 12528
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.823304496363253,
+      "learning_rate": 9.942607157408784e-07,
+      "loss": 0.7164,
+      "step": 12529
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7538672640321824,
+      "learning_rate": 9.936404989907828e-07,
+      "loss": 0.7206,
+      "step": 12530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1022639419652265,
+      "learning_rate": 9.930204544028787e-07,
+      "loss": 0.7338,
+      "step": 12531
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7582434668394564,
+      "learning_rate": 9.924005820038102e-07,
+      "loss": 0.7584,
+      "step": 12532
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5210361795241907,
+      "learning_rate": 9.917808818202135e-07,
+      "loss": 0.6586,
+      "step": 12533
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.532873660411348,
+      "learning_rate": 9.911613538787196e-07,
+      "loss": 0.5686,
+      "step": 12534
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5956847129981098,
+      "learning_rate": 9.905419982059488e-07,
+      "loss": 0.7428,
+      "step": 12535
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8400502580239786,
+      "learning_rate": 9.899228148285206e-07,
+      "loss": 0.6529,
+      "step": 12536
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0517793660345858,
+      "learning_rate": 9.89303803773039e-07,
+      "loss": 0.6657,
+      "step": 12537
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2927058227149308,
+      "learning_rate": 9.886849650661035e-07,
+      "loss": 0.6929,
+      "step": 12538
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9641459820989415,
+      "learning_rate": 9.880662987343103e-07,
+      "loss": 0.5682,
+      "step": 12539
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7780130389795146,
+      "learning_rate": 9.87447804804243e-07,
+      "loss": 0.6799,
+      "step": 12540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4613728804523929,
+      "learning_rate": 9.868294833024805e-07,
+      "loss": 0.7785,
+      "step": 12541
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8519318452422822,
+      "learning_rate": 9.862113342555924e-07,
+      "loss": 0.722,
+      "step": 12542
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.517802833934489,
+      "learning_rate": 9.855933576901428e-07,
+      "loss": 0.668,
+      "step": 12543
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.887272875057069,
+      "learning_rate": 9.849755536326866e-07,
+      "loss": 0.7168,
+      "step": 12544
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8638598514682414,
+      "learning_rate": 9.843579221097722e-07,
+      "loss": 0.6807,
+      "step": 12545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.077646373117632,
+      "learning_rate": 9.837404631479419e-07,
+      "loss": 0.5613,
+      "step": 12546
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2304163252924984,
+      "learning_rate": 9.831231767737305e-07,
+      "loss": 0.6755,
+      "step": 12547
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5487899147751805,
+      "learning_rate": 9.82506063013659e-07,
+      "loss": 0.6909,
+      "step": 12548
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.537093598392847,
+      "learning_rate": 9.818891218942511e-07,
+      "loss": 0.6528,
+      "step": 12549
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7263539837037485,
+      "learning_rate": 9.812723534420165e-07,
+      "loss": 0.7729,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9020637602927823,
+      "learning_rate": 9.806557576834591e-07,
+      "loss": 0.6409,
+      "step": 12551
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6471594505153002,
+      "learning_rate": 9.800393346450754e-07,
+      "loss": 0.7517,
+      "step": 12552
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8031689883551243,
+      "learning_rate": 9.794230843533525e-07,
+      "loss": 0.7166,
+      "step": 12553
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.158745523327748,
+      "learning_rate": 9.78807006834777e-07,
+      "loss": 0.7157,
+      "step": 12554
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7243652558264397,
+      "learning_rate": 9.781911021158181e-07,
+      "loss": 0.7899,
+      "step": 12555
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6280379910245077,
+      "learning_rate": 9.775753702229452e-07,
+      "loss": 0.6311,
+      "step": 12556
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0733718206105065,
+      "learning_rate": 9.769598111826166e-07,
+      "loss": 0.5846,
+      "step": 12557
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6197814018607914,
+      "learning_rate": 9.763444250212855e-07,
+      "loss": 0.6551,
+      "step": 12558
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.92550638015739,
+      "learning_rate": 9.757292117653955e-07,
+      "loss": 0.7409,
+      "step": 12559
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.948802892232534,
+      "learning_rate": 9.751141714413836e-07,
+      "loss": 0.827,
+      "step": 12560
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6743168099936923,
+      "learning_rate": 9.7449930407568e-07,
+      "loss": 0.6165,
+      "step": 12561
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8396475716117466,
+      "learning_rate": 9.738846096947064e-07,
+      "loss": 0.7181,
+      "step": 12562
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.654941016927703,
+      "learning_rate": 9.732700883248763e-07,
+      "loss": 0.7181,
+      "step": 12563
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.084771996717991,
+      "learning_rate": 9.726557399925995e-07,
+      "loss": 0.7456,
+      "step": 12564
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6132776730510474,
+      "learning_rate": 9.72041564724277e-07,
+      "loss": 0.6763,
+      "step": 12565
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6940088697565934,
+      "learning_rate": 9.71427562546296e-07,
+      "loss": 0.6563,
+      "step": 12566
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8265806361331947,
+      "learning_rate": 9.708137334850465e-07,
+      "loss": 0.8243,
+      "step": 12567
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1057994970745486,
+      "learning_rate": 9.702000775669041e-07,
+      "loss": 0.6306,
+      "step": 12568
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7970579127455664,
+      "learning_rate": 9.695865948182392e-07,
+      "loss": 0.6043,
+      "step": 12569
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.783082902228062,
+      "learning_rate": 9.689732852654143e-07,
+      "loss": 0.6961,
+      "step": 12570
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6877995991564896,
+      "learning_rate": 9.683601489347843e-07,
+      "loss": 0.7205,
+      "step": 12571
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.09091277512291,
+      "learning_rate": 9.677471858526998e-07,
+      "loss": 0.7539,
+      "step": 12572
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6241998858960907,
+      "learning_rate": 9.671343960454965e-07,
+      "loss": 0.7048,
+      "step": 12573
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0884577182874555,
+      "learning_rate": 9.66521779539511e-07,
+      "loss": 0.6582,
+      "step": 12574
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9899215008154252,
+      "learning_rate": 9.659093363610678e-07,
+      "loss": 0.6856,
+      "step": 12575
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0549469398481497,
+      "learning_rate": 9.652970665364847e-07,
+      "loss": 0.5488,
+      "step": 12576
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9828175520449635,
+      "learning_rate": 9.646849700920729e-07,
+      "loss": 0.5981,
+      "step": 12577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6531578384581156,
+      "learning_rate": 9.640730470541343e-07,
+      "loss": 0.7102,
+      "step": 12578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.763983629912063,
+      "learning_rate": 9.63461297448966e-07,
+      "loss": 0.7397,
+      "step": 12579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7085617628334733,
+      "learning_rate": 9.628497213028553e-07,
+      "loss": 0.7228,
+      "step": 12580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9113555049421425,
+      "learning_rate": 9.622383186420818e-07,
+      "loss": 0.7873,
+      "step": 12581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7969203371996532,
+      "learning_rate": 9.616270894929219e-07,
+      "loss": 0.7278,
+      "step": 12582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8113021150532846,
+      "learning_rate": 9.610160338816404e-07,
+      "loss": 0.7167,
+      "step": 12583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.544513627744651,
+      "learning_rate": 9.604051518344948e-07,
+      "loss": 0.6375,
+      "step": 12584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5875319717147698,
+      "learning_rate": 9.597944433777362e-07,
+      "loss": 0.7687,
+      "step": 12585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5727488190314904,
+      "learning_rate": 9.59183908537607e-07,
+      "loss": 0.6743,
+      "step": 12586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9808446867322856,
+      "learning_rate": 9.585735473403479e-07,
+      "loss": 0.5939,
+      "step": 12587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.823463439156527,
+      "learning_rate": 9.579633598121824e-07,
+      "loss": 0.6639,
+      "step": 12588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5378843402194433,
+      "learning_rate": 9.57353345979332e-07,
+      "loss": 0.7013,
+      "step": 12589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.784530292106963,
+      "learning_rate": 9.567435058680146e-07,
+      "loss": 0.7034,
+      "step": 12590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.825667360167674,
+      "learning_rate": 9.56133839504431e-07,
+      "loss": 0.7122,
+      "step": 12591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0311671516425474,
+      "learning_rate": 9.55524346914784e-07,
+      "loss": 0.6951,
+      "step": 12592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9361284702025554,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.7141,
+      "step": 12593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.675991306920749,
+      "learning_rate": 9.543058831620528e-07,
+      "loss": 0.7761,
+      "step": 12594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7547881302576447,
+      "learning_rate": 9.536969120513284e-07,
+      "loss": 0.7392,
+      "step": 12595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9244823881129671,
+      "learning_rate": 9.530881148192578e-07,
+      "loss": 0.7778,
+      "step": 12596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8901215437941752,
+      "learning_rate": 9.524794914920072e-07,
+      "loss": 0.6916,
+      "step": 12597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9300530442664092,
+      "learning_rate": 9.51871042095725e-07,
+      "loss": 0.7193,
+      "step": 12598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8932527096145269,
+      "learning_rate": 9.512627666565588e-07,
+      "loss": 0.633,
+      "step": 12599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7640575598411778,
+      "learning_rate": 9.506546652006504e-07,
+      "loss": 0.7927,
+      "step": 12600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6696166937187509,
+      "learning_rate": 9.500467377541289e-07,
+      "loss": 0.6453,
+      "step": 12601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7741609722245262,
+      "learning_rate": 9.49438984343119e-07,
+      "loss": 0.6524,
+      "step": 12602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1162646398193092,
+      "learning_rate": 9.48831404993737e-07,
+      "loss": 0.5886,
+      "step": 12603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1029490429611697,
+      "learning_rate": 9.482239997320903e-07,
+      "loss": 0.6664,
+      "step": 12604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6967588086888479,
+      "learning_rate": 9.476167685842852e-07,
+      "loss": 0.6941,
+      "step": 12605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6577309653048578,
+      "learning_rate": 9.470097115764099e-07,
+      "loss": 0.7757,
+      "step": 12606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7208209558056715,
+      "learning_rate": 9.464028287345551e-07,
+      "loss": 0.631,
+      "step": 12607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6902809142273298,
+      "learning_rate": 9.457961200847998e-07,
+      "loss": 0.6657,
+      "step": 12608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.779195370028801,
+      "learning_rate": 9.451895856532117e-07,
+      "loss": 0.6248,
+      "step": 12609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5996973967379502,
+      "learning_rate": 9.445832254658594e-07,
+      "loss": 0.7202,
+      "step": 12610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.026843171544581,
+      "learning_rate": 9.439770395487974e-07,
+      "loss": 0.7088,
+      "step": 12611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7310108687735029,
+      "learning_rate": 9.43371027928075e-07,
+      "loss": 0.7411,
+      "step": 12612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8659294478615118,
+      "learning_rate": 9.427651906297347e-07,
+      "loss": 0.7508,
+      "step": 12613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.99943799126951,
+      "learning_rate": 9.421595276798084e-07,
+      "loss": 0.7223,
+      "step": 12614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6663638666148393,
+      "learning_rate": 9.415540391043276e-07,
+      "loss": 0.6082,
+      "step": 12615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6338050622301137,
+      "learning_rate": 9.40948724929307e-07,
+      "loss": 0.6475,
+      "step": 12616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5920979504178507,
+      "learning_rate": 9.403435851807579e-07,
+      "loss": 0.7225,
+      "step": 12617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2037438452954123,
+      "learning_rate": 9.397386198846881e-07,
+      "loss": 0.6923,
+      "step": 12618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6686096432095099,
+      "learning_rate": 9.39133829067092e-07,
+      "loss": 0.6563,
+      "step": 12619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7892959617862896,
+      "learning_rate": 9.385292127539597e-07,
+      "loss": 0.7932,
+      "step": 12620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8437966028151445,
+      "learning_rate": 9.379247709712725e-07,
+      "loss": 0.8055,
+      "step": 12621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6742713191365897,
+      "learning_rate": 9.373205037450028e-07,
+      "loss": 0.745,
+      "step": 12622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5927495006344445,
+      "learning_rate": 9.367164111011223e-07,
+      "loss": 0.7907,
+      "step": 12623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4238320636919184,
+      "learning_rate": 9.361124930655841e-07,
+      "loss": 0.7089,
+      "step": 12624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6770475465208663,
+      "learning_rate": 9.355087496643444e-07,
+      "loss": 0.6641,
+      "step": 12625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5003493426230228,
+      "learning_rate": 9.349051809233472e-07,
+      "loss": 0.7204,
+      "step": 12626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6560920163796684,
+      "learning_rate": 9.34301786868525e-07,
+      "loss": 0.6849,
+      "step": 12627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8217820007152197,
+      "learning_rate": 9.336985675258109e-07,
+      "loss": 0.7851,
+      "step": 12628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8031129535094852,
+      "learning_rate": 9.330955229211259e-07,
+      "loss": 0.671,
+      "step": 12629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.156799058241438,
+      "learning_rate": 9.324926530803835e-07,
+      "loss": 0.772,
+      "step": 12630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6163139223003777,
+      "learning_rate": 9.31889958029491e-07,
+      "loss": 0.6767,
+      "step": 12631
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.640170449342623,
+      "learning_rate": 9.312874377943454e-07,
+      "loss": 0.6957,
+      "step": 12632
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7918585848701898,
+      "learning_rate": 9.306850924008415e-07,
+      "loss": 0.7717,
+      "step": 12633
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.565550812959664,
+      "learning_rate": 9.300829218748625e-07,
+      "loss": 0.7214,
+      "step": 12634
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7764771205473084,
+      "learning_rate": 9.294809262422838e-07,
+      "loss": 0.7212,
+      "step": 12635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.750441912527927,
+      "learning_rate": 9.288791055289759e-07,
+      "loss": 0.7451,
+      "step": 12636
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8347721133551116,
+      "learning_rate": 9.282774597607991e-07,
+      "loss": 0.6501,
+      "step": 12637
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5763747559485226,
+      "learning_rate": 9.276759889636084e-07,
+      "loss": 0.7769,
+      "step": 12638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.142977876698756,
+      "learning_rate": 9.270746931632501e-07,
+      "loss": 0.7065,
+      "step": 12639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1185249278473133,
+      "learning_rate": 9.264735723855617e-07,
+      "loss": 0.6976,
+      "step": 12640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6736657667870702,
+      "learning_rate": 9.258726266563789e-07,
+      "loss": 0.6379,
+      "step": 12641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7842713028711203,
+      "learning_rate": 9.2527185600152e-07,
+      "loss": 0.7594,
+      "step": 12642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8052830452673398,
+      "learning_rate": 9.246712604468061e-07,
+      "loss": 0.6854,
+      "step": 12643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7455419848536056,
+      "learning_rate": 9.240708400180437e-07,
+      "loss": 0.7619,
+      "step": 12644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2814309213756374,
+      "learning_rate": 9.234705947410355e-07,
+      "loss": 0.813,
+      "step": 12645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.521162919857947,
+      "learning_rate": 9.228705246415742e-07,
+      "loss": 0.6595,
+      "step": 12646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7168709047602329,
+      "learning_rate": 9.222706297454459e-07,
+      "loss": 0.8372,
+      "step": 12647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.568597540849575,
+      "learning_rate": 9.216709100784326e-07,
+      "loss": 0.682,
+      "step": 12648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.02830031401305,
+      "learning_rate": 9.210713656663023e-07,
+      "loss": 0.7188,
+      "step": 12649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0643536301876688,
+      "learning_rate": 9.20471996534818e-07,
+      "loss": 0.6787,
+      "step": 12650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5393285154281482,
+      "learning_rate": 9.198728027097386e-07,
+      "loss": 0.7209,
+      "step": 12651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1734457495812705,
+      "learning_rate": 9.192737842168126e-07,
+      "loss": 0.7053,
+      "step": 12652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7118019099928412,
+      "learning_rate": 9.186749410817797e-07,
+      "loss": 0.7562,
+      "step": 12653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0192545443328465,
+      "learning_rate": 9.180762733303745e-07,
+      "loss": 0.583,
+      "step": 12654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8222236270652568,
+      "learning_rate": 9.174777809883229e-07,
+      "loss": 0.7297,
+      "step": 12655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5082335983027768,
+      "learning_rate": 9.168794640813428e-07,
+      "loss": 0.6701,
+      "step": 12656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9750109952025339,
+      "learning_rate": 9.162813226351447e-07,
+      "loss": 0.5628,
+      "step": 12657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.030829073820417,
+      "learning_rate": 9.156833566754347e-07,
+      "loss": 0.692,
+      "step": 12658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6903272745117115,
+      "learning_rate": 9.150855662279079e-07,
+      "loss": 0.7321,
+      "step": 12659
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1442514212684043,
+      "learning_rate": 9.144879513182498e-07,
+      "loss": 0.6961,
+      "step": 12660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.687410551719353,
+      "learning_rate": 9.138905119721442e-07,
+      "loss": 0.6599,
+      "step": 12661
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6537525153497272,
+      "learning_rate": 9.13293248215264e-07,
+      "loss": 0.7159,
+      "step": 12662
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.880806720619183,
+      "learning_rate": 9.126961600732742e-07,
+      "loss": 0.6395,
+      "step": 12663
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0471968220210335,
+      "learning_rate": 9.120992475718333e-07,
+      "loss": 0.662,
+      "step": 12664
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7125539917399648,
+      "learning_rate": 9.115025107365904e-07,
+      "loss": 0.6451,
+      "step": 12665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9389716982306242,
+      "learning_rate": 9.109059495931932e-07,
+      "loss": 0.7192,
+      "step": 12666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5467432549866293,
+      "learning_rate": 9.103095641672732e-07,
+      "loss": 0.7255,
+      "step": 12667
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9471910392329175,
+      "learning_rate": 9.097133544844577e-07,
+      "loss": 0.6414,
+      "step": 12668
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2720021329917994,
+      "learning_rate": 9.091173205703708e-07,
+      "loss": 0.6508,
+      "step": 12669
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6468312480451577,
+      "learning_rate": 9.085214624506228e-07,
+      "loss": 0.7743,
+      "step": 12670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0533353214947607,
+      "learning_rate": 9.079257801508201e-07,
+      "loss": 0.8265,
+      "step": 12671
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.557749007817813,
+      "learning_rate": 9.0733027369656e-07,
+      "loss": 0.8019,
+      "step": 12672
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1748488314968217,
+      "learning_rate": 9.067349431134331e-07,
+      "loss": 0.5735,
+      "step": 12673
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8762483219865513,
+      "learning_rate": 9.061397884270217e-07,
+      "loss": 0.7772,
+      "step": 12674
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4768115756763514,
+      "learning_rate": 9.055448096628999e-07,
+      "loss": 0.7488,
+      "step": 12675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7310909839118858,
+      "learning_rate": 9.049500068466377e-07,
+      "loss": 0.6057,
+      "step": 12676
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7641692397629996,
+      "learning_rate": 9.043553800037952e-07,
+      "loss": 0.6567,
+      "step": 12677
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5771486588724846,
+      "learning_rate": 9.037609291599214e-07,
+      "loss": 0.5949,
+      "step": 12678
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5473593622943693,
+      "learning_rate": 9.031666543405637e-07,
+      "loss": 0.7728,
+      "step": 12679
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.446450818750784,
+      "learning_rate": 9.025725555712595e-07,
+      "loss": 0.7249,
+      "step": 12680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.033241086640504,
+      "learning_rate": 9.019786328775382e-07,
+      "loss": 0.7685,
+      "step": 12681
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2728061634259324,
+      "learning_rate": 9.013848862849217e-07,
+      "loss": 0.6748,
+      "step": 12682
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0664185186561699,
+      "learning_rate": 9.007913158189236e-07,
+      "loss": 0.6258,
+      "step": 12683
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.564923479117562,
+      "learning_rate": 9.001979215050544e-07,
+      "loss": 0.6855,
+      "step": 12684
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0892964174701574,
+      "learning_rate": 8.996047033688083e-07,
+      "loss": 0.763,
+      "step": 12685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6849872107903665,
+      "learning_rate": 8.990116614356819e-07,
+      "loss": 0.6603,
+      "step": 12686
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6624443529234003,
+      "learning_rate": 8.984187957311579e-07,
+      "loss": 0.7331,
+      "step": 12687
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0566284998012476,
+      "learning_rate": 8.978261062807131e-07,
+      "loss": 0.6322,
+      "step": 12688
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0189747303732317,
+      "learning_rate": 8.972335931098159e-07,
+      "loss": 0.8216,
+      "step": 12689
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0864489397422583,
+      "learning_rate": 8.966412562439291e-07,
+      "loss": 0.6219,
+      "step": 12690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.745347488282354,
+      "learning_rate": 8.960490957085061e-07,
+      "loss": 0.7356,
+      "step": 12691
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3888446328690036,
+      "learning_rate": 8.954571115289934e-07,
+      "loss": 0.7248,
+      "step": 12692
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.092073374367913,
+      "learning_rate": 8.948653037308286e-07,
+      "loss": 0.6785,
+      "step": 12693
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8036821974322534,
+      "learning_rate": 8.942736723394458e-07,
+      "loss": 0.5758,
+      "step": 12694
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6248322448969077,
+      "learning_rate": 8.936822173802667e-07,
+      "loss": 0.6859,
+      "step": 12695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2055754931716764,
+      "learning_rate": 8.930909388787084e-07,
+      "loss": 0.6198,
+      "step": 12696
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4179990115026133,
+      "learning_rate": 8.92499836860179e-07,
+      "loss": 0.6425,
+      "step": 12697
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7982663492567859,
+      "learning_rate": 8.919089113500795e-07,
+      "loss": 0.7599,
+      "step": 12698
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.059921454849142,
+      "learning_rate": 8.913181623738032e-07,
+      "loss": 0.705,
+      "step": 12699
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6320885342669287,
+      "learning_rate": 8.907275899567363e-07,
+      "loss": 0.609,
+      "step": 12700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.433552801933943,
+      "learning_rate": 8.901371941242554e-07,
+      "loss": 0.6087,
+      "step": 12701
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0782476213127983,
+      "learning_rate": 8.895469749017344e-07,
+      "loss": 0.6781,
+      "step": 12702
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9469301950049378,
+      "learning_rate": 8.889569323145325e-07,
+      "loss": 0.763,
+      "step": 12703
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.675525994756449,
+      "learning_rate": 8.883670663880078e-07,
+      "loss": 0.8233,
+      "step": 12704
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.088346787441878,
+      "learning_rate": 8.877773771475074e-07,
+      "loss": 0.7033,
+      "step": 12705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.865265085672549,
+      "learning_rate": 8.871878646183718e-07,
+      "loss": 0.7412,
+      "step": 12706
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.648913802425956,
+      "learning_rate": 8.865985288259332e-07,
+      "loss": 0.6767,
+      "step": 12707
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7021938302394772,
+      "learning_rate": 8.860093697955152e-07,
+      "loss": 0.6801,
+      "step": 12708
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.844993334530243,
+      "learning_rate": 8.854203875524403e-07,
+      "loss": 0.7572,
+      "step": 12709
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8266730918836138,
+      "learning_rate": 8.848315821220133e-07,
+      "loss": 0.8215,
+      "step": 12710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8594216590104744,
+      "learning_rate": 8.842429535295366e-07,
+      "loss": 0.7953,
+      "step": 12711
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.001646001097912,
+      "learning_rate": 8.836545018003084e-07,
+      "loss": 0.7222,
+      "step": 12712
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6130626362886624,
+      "learning_rate": 8.830662269596135e-07,
+      "loss": 0.7066,
+      "step": 12713
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.047591305802909,
+      "learning_rate": 8.824781290327317e-07,
+      "loss": 0.6954,
+      "step": 12714
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.01128164037424,
+      "learning_rate": 8.818902080449348e-07,
+      "loss": 0.6887,
+      "step": 12715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7118557373189738,
+      "learning_rate": 8.813024640214873e-07,
+      "loss": 0.6742,
+      "step": 12716
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.689217779851701,
+      "learning_rate": 8.807148969876455e-07,
+      "loss": 0.7048,
+      "step": 12717
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6889441627415893,
+      "learning_rate": 8.801275069686593e-07,
+      "loss": 0.6955,
+      "step": 12718
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.767872109128628,
+      "learning_rate": 8.795402939897679e-07,
+      "loss": 0.7235,
+      "step": 12719
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8547322477493597,
+      "learning_rate": 8.789532580762095e-07,
+      "loss": 0.784,
+      "step": 12720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5923325702539302,
+      "learning_rate": 8.783663992532048e-07,
+      "loss": 0.6361,
+      "step": 12721
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2429990241508406,
+      "learning_rate": 8.777797175459773e-07,
+      "loss": 0.6288,
+      "step": 12722
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6801924143388043,
+      "learning_rate": 8.771932129797356e-07,
+      "loss": 0.621,
+      "step": 12723
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8438233287972332,
+      "learning_rate": 8.766068855796833e-07,
+      "loss": 0.6613,
+      "step": 12724
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7070441512465455,
+      "learning_rate": 8.760207353710165e-07,
+      "loss": 0.6447,
+      "step": 12725
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1149892549232736,
+      "learning_rate": 8.754347623789222e-07,
+      "loss": 0.6506,
+      "step": 12726
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.797601538965545,
+      "learning_rate": 8.748489666285842e-07,
+      "loss": 0.7667,
+      "step": 12727
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7413359555320056,
+      "learning_rate": 8.742633481451728e-07,
+      "loss": 0.692,
+      "step": 12728
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7069170685811508,
+      "learning_rate": 8.736779069538521e-07,
+      "loss": 0.7117,
+      "step": 12729
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.535668856662135,
+      "learning_rate": 8.730926430797826e-07,
+      "loss": 0.7124,
+      "step": 12730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0283267251332178,
+      "learning_rate": 8.72507556548114e-07,
+      "loss": 0.809,
+      "step": 12731
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3787415640853051,
+      "learning_rate": 8.719226473839876e-07,
+      "loss": 0.6501,
+      "step": 12732
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5320418042783153,
+      "learning_rate": 8.713379156125385e-07,
+      "loss": 0.6948,
+      "step": 12733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2292913885460155,
+      "learning_rate": 8.707533612588948e-07,
+      "loss": 0.6397,
+      "step": 12734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9309219109538183,
+      "learning_rate": 8.701689843481753e-07,
+      "loss": 0.7843,
+      "step": 12735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9090582343467761,
+      "learning_rate": 8.695847849054906e-07,
+      "loss": 0.6293,
+      "step": 12736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5897758213379187,
+      "learning_rate": 8.690007629559482e-07,
+      "loss": 0.7528,
+      "step": 12737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.969525214755921,
+      "learning_rate": 8.684169185246444e-07,
+      "loss": 0.794,
+      "step": 12738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0754516767456797,
+      "learning_rate": 8.67833251636665e-07,
+      "loss": 0.6581,
+      "step": 12739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8111540121852086,
+      "learning_rate": 8.672497623170944e-07,
+      "loss": 0.6562,
+      "step": 12740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5838434591732558,
+      "learning_rate": 8.666664505910055e-07,
+      "loss": 0.6525,
+      "step": 12741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7988374059149514,
+      "learning_rate": 8.660833164834653e-07,
+      "loss": 0.7017,
+      "step": 12742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.043553390321054,
+      "learning_rate": 8.65500360019531e-07,
+      "loss": 0.5793,
+      "step": 12743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5390395957759797,
+      "learning_rate": 8.649175812242532e-07,
+      "loss": 0.5987,
+      "step": 12744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5172820717026068,
+      "learning_rate": 8.643349801226791e-07,
+      "loss": 0.7288,
+      "step": 12745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7500626203732097,
+      "learning_rate": 8.637525567398392e-07,
+      "loss": 0.7417,
+      "step": 12746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7565041630405627,
+      "learning_rate": 8.631703111007645e-07,
+      "loss": 0.6681,
+      "step": 12747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6534643446187394,
+      "learning_rate": 8.625882432304749e-07,
+      "loss": 0.6433,
+      "step": 12748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.58085857571529,
+      "learning_rate": 8.62006353153983e-07,
+      "loss": 0.6433,
+      "step": 12749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7424760462193012,
+      "learning_rate": 8.61424640896294e-07,
+      "loss": 0.6901,
+      "step": 12750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7260080629585517,
+      "learning_rate": 8.608431064824052e-07,
+      "loss": 0.7559,
+      "step": 12751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7808007884696613,
+      "learning_rate": 8.602617499373056e-07,
+      "loss": 0.7241,
+      "step": 12752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6965680520675235,
+      "learning_rate": 8.596805712859807e-07,
+      "loss": 0.6842,
+      "step": 12753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6683108437202023,
+      "learning_rate": 8.590995705533994e-07,
+      "loss": 0.7616,
+      "step": 12754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.748124649029358,
+      "learning_rate": 8.58518747764534e-07,
+      "loss": 0.7137,
+      "step": 12755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.76091082734396,
+      "learning_rate": 8.579381029443412e-07,
+      "loss": 0.5821,
+      "step": 12756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9118988824686278,
+      "learning_rate": 8.573576361177733e-07,
+      "loss": 0.7501,
+      "step": 12757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0194295878981412,
+      "learning_rate": 8.567773473097735e-07,
+      "loss": 0.6114,
+      "step": 12758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7928680863898638,
+      "learning_rate": 8.561972365452775e-07,
+      "loss": 0.7041,
+      "step": 12759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2687749608009902,
+      "learning_rate": 8.556173038492171e-07,
+      "loss": 0.6405,
+      "step": 12760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.746058143228866,
+      "learning_rate": 8.550375492465102e-07,
+      "loss": 0.767,
+      "step": 12761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0651312239355217,
+      "learning_rate": 8.5445797276207e-07,
+      "loss": 0.7276,
+      "step": 12762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.63330390623966,
+      "learning_rate": 8.538785744208062e-07,
+      "loss": 0.8317,
+      "step": 12763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6391207580512166,
+      "learning_rate": 8.532993542476108e-07,
+      "loss": 0.844,
+      "step": 12764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6703724368171535,
+      "learning_rate": 8.527203122673789e-07,
+      "loss": 0.8318,
+      "step": 12765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6901332547624963,
+      "learning_rate": 8.521414485049917e-07,
+      "loss": 0.7799,
+      "step": 12766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7577259456927568,
+      "learning_rate": 8.51562762985324e-07,
+      "loss": 0.8717,
+      "step": 12767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9297376236683499,
+      "learning_rate": 8.509842557332437e-07,
+      "loss": 0.6443,
+      "step": 12768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6176870903468799,
+      "learning_rate": 8.504059267736097e-07,
+      "loss": 0.7011,
+      "step": 12769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.098331351824939,
+      "learning_rate": 8.49827776131274e-07,
+      "loss": 0.795,
+      "step": 12770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8083160595211007,
+      "learning_rate": 8.492498038310843e-07,
+      "loss": 0.7072,
+      "step": 12771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0193668836997163,
+      "learning_rate": 8.486720098978718e-07,
+      "loss": 0.7055,
+      "step": 12772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2622120003565116,
+      "learning_rate": 8.480943943564701e-07,
+      "loss": 0.749,
+      "step": 12773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1916541115860197,
+      "learning_rate": 8.475169572316988e-07,
+      "loss": 0.7415,
+      "step": 12774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7594436360584533,
+      "learning_rate": 8.469396985483724e-07,
+      "loss": 0.6941,
+      "step": 12775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6501858012794342,
+      "learning_rate": 8.463626183312962e-07,
+      "loss": 0.637,
+      "step": 12776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5845890238962226,
+      "learning_rate": 8.457857166052674e-07,
+      "loss": 0.6498,
+      "step": 12777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1177169378874208,
+      "learning_rate": 8.452089933950813e-07,
+      "loss": 0.6357,
+      "step": 12778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.543368823729768,
+      "learning_rate": 8.446324487255164e-07,
+      "loss": 0.7417,
+      "step": 12779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7550105154987865,
+      "learning_rate": 8.440560826213485e-07,
+      "loss": 0.6307,
+      "step": 12780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1506407326940953,
+      "learning_rate": 8.434798951073492e-07,
+      "loss": 0.6477,
+      "step": 12781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.366743884219293,
+      "learning_rate": 8.429038862082734e-07,
+      "loss": 0.6886,
+      "step": 12782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.699142457805042,
+      "learning_rate": 8.423280559488767e-07,
+      "loss": 0.6417,
+      "step": 12783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.645020340818598,
+      "learning_rate": 8.417524043539038e-07,
+      "loss": 0.6249,
+      "step": 12784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.515488767489869,
+      "learning_rate": 8.411769314480905e-07,
+      "loss": 0.6048,
+      "step": 12785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8014069904345746,
+      "learning_rate": 8.406016372561665e-07,
+      "loss": 0.6695,
+      "step": 12786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.781738377824596,
+      "learning_rate": 8.400265218028525e-07,
+      "loss": 0.7029,
+      "step": 12787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8667218458534467,
+      "learning_rate": 8.394515851128654e-07,
+      "loss": 0.6807,
+      "step": 12788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6134059179960827,
+      "learning_rate": 8.388768272109105e-07,
+      "loss": 0.628,
+      "step": 12789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5680567972907014,
+      "learning_rate": 8.383022481216829e-07,
+      "loss": 0.8134,
+      "step": 12790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8007285004703848,
+      "learning_rate": 8.37727847869878e-07,
+      "loss": 0.7037,
+      "step": 12791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7447981651679356,
+      "learning_rate": 8.371536264801772e-07,
+      "loss": 0.6851,
+      "step": 12792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3064937522448175,
+      "learning_rate": 8.365795839772561e-07,
+      "loss": 0.7677,
+      "step": 12793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.551272986153029,
+      "learning_rate": 8.36005720385783e-07,
+      "loss": 0.7183,
+      "step": 12794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0014361881216252,
+      "learning_rate": 8.354320357304163e-07,
+      "loss": 0.6466,
+      "step": 12795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.829332691587707,
+      "learning_rate": 8.34858530035813e-07,
+      "loss": 0.5437,
+      "step": 12796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5022960524795694,
+      "learning_rate": 8.342852033266119e-07,
+      "loss": 0.6577,
+      "step": 12797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5130950104395706,
+      "learning_rate": 8.337120556274548e-07,
+      "loss": 0.7176,
+      "step": 12798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.034113988493985,
+      "learning_rate": 8.331390869629702e-07,
+      "loss": 0.8291,
+      "step": 12799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6156754244652154,
+      "learning_rate": 8.32566297357777e-07,
+      "loss": 0.6094,
+      "step": 12800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8594295816570388,
+      "learning_rate": 8.319936868364925e-07,
+      "loss": 0.6917,
+      "step": 12801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.819260235323442,
+      "learning_rate": 8.314212554237222e-07,
+      "loss": 0.7154,
+      "step": 12802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7641979488799397,
+      "learning_rate": 8.308490031440641e-07,
+      "loss": 0.7005,
+      "step": 12803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1508463559801125,
+      "learning_rate": 8.302769300221098e-07,
+      "loss": 0.5613,
+      "step": 12804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5817005499116918,
+      "learning_rate": 8.29705036082441e-07,
+      "loss": 0.6779,
+      "step": 12805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.400753070653093,
+      "learning_rate": 8.291333213496355e-07,
+      "loss": 0.6901,
+      "step": 12806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5551467409343607,
+      "learning_rate": 8.285617858482609e-07,
+      "loss": 0.7107,
+      "step": 12807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5584897309758636,
+      "learning_rate": 8.279904296028757e-07,
+      "loss": 0.7051,
+      "step": 12808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3742689465283613,
+      "learning_rate": 8.274192526380337e-07,
+      "loss": 0.5652,
+      "step": 12809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.543947248706248,
+      "learning_rate": 8.268482549782797e-07,
+      "loss": 0.6892,
+      "step": 12810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.778350276666105,
+      "learning_rate": 8.262774366481496e-07,
+      "loss": 0.667,
+      "step": 12811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5383009987945926,
+      "learning_rate": 8.257067976721739e-07,
+      "loss": 0.6094,
+      "step": 12812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6707705142495173,
+      "learning_rate": 8.251363380748723e-07,
+      "loss": 0.8102,
+      "step": 12813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.537436380054552,
+      "learning_rate": 8.24566057880763e-07,
+      "loss": 0.7241,
+      "step": 12814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9143722438788897,
+      "learning_rate": 8.239959571143463e-07,
+      "loss": 0.7622,
+      "step": 12815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.83480783010793,
+      "learning_rate": 8.234260358001256e-07,
+      "loss": 0.6819,
+      "step": 12816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7658476667112,
+      "learning_rate": 8.228562939625906e-07,
+      "loss": 0.7109,
+      "step": 12817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2580677823401054,
+      "learning_rate": 8.222867316262217e-07,
+      "loss": 0.7991,
+      "step": 12818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.774868989521529,
+      "learning_rate": 8.217173488154972e-07,
+      "loss": 0.7327,
+      "step": 12819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7427265825544769,
+      "learning_rate": 8.211481455548837e-07,
+      "loss": 0.7174,
+      "step": 12820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.645324173080836,
+      "learning_rate": 8.205791218688414e-07,
+      "loss": 0.7364,
+      "step": 12821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.837949169230332,
+      "learning_rate": 8.20010277781822e-07,
+      "loss": 0.7468,
+      "step": 12822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3078969011833455,
+      "learning_rate": 8.194416133182693e-07,
+      "loss": 0.667,
+      "step": 12823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.223722147844546,
+      "learning_rate": 8.188731285026219e-07,
+      "loss": 0.6975,
+      "step": 12824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.622426586680079,
+      "learning_rate": 8.183048233593083e-07,
+      "loss": 0.6643,
+      "step": 12825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.72735504358278,
+      "learning_rate": 8.177366979127499e-07,
+      "loss": 0.7868,
+      "step": 12826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.918156961374051,
+      "learning_rate": 8.171687521873595e-07,
+      "loss": 0.6908,
+      "step": 12827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8634412059151675,
+      "learning_rate": 8.166009862075435e-07,
+      "loss": 0.6439,
+      "step": 12828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5503126490222883,
+      "learning_rate": 8.160333999977004e-07,
+      "loss": 0.7219,
+      "step": 12829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.425971129416011,
+      "learning_rate": 8.154659935822201e-07,
+      "loss": 0.6573,
+      "step": 12830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.541226274207035,
+      "learning_rate": 8.148987669854846e-07,
+      "loss": 0.6885,
+      "step": 12831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.554713676207537,
+      "learning_rate": 8.143317202318718e-07,
+      "loss": 0.6143,
+      "step": 12832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6429438183248941,
+      "learning_rate": 8.137648533457448e-07,
+      "loss": 0.7196,
+      "step": 12833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6098661301206334,
+      "learning_rate": 8.131981663514665e-07,
+      "loss": 0.728,
+      "step": 12834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6603163033696835,
+      "learning_rate": 8.126316592733874e-07,
+      "loss": 0.7298,
+      "step": 12835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9041274383319264,
+      "learning_rate": 8.120653321358519e-07,
+      "loss": 0.6927,
+      "step": 12836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5938670900702319,
+      "learning_rate": 8.114991849631953e-07,
+      "loss": 0.7559,
+      "step": 12837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8779411575605895,
+      "learning_rate": 8.109332177797469e-07,
+      "loss": 0.7423,
+      "step": 12838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7410006838423526,
+      "learning_rate": 8.103674306098291e-07,
+      "loss": 0.7494,
+      "step": 12839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9360610600100463,
+      "learning_rate": 8.098018234777533e-07,
+      "loss": 0.7959,
+      "step": 12840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6859790435137572,
+      "learning_rate": 8.092363964078237e-07,
+      "loss": 0.6644,
+      "step": 12841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.830109084731601,
+      "learning_rate": 8.086711494243405e-07,
+      "loss": 0.7709,
+      "step": 12842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7662192903768779,
+      "learning_rate": 8.081060825515924e-07,
+      "loss": 0.5801,
+      "step": 12843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5214747938527837,
+      "learning_rate": 8.075411958138623e-07,
+      "loss": 0.6707,
+      "step": 12844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6582595488218643,
+      "learning_rate": 8.069764892354237e-07,
+      "loss": 0.6877,
+      "step": 12845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4232085416478315,
+      "learning_rate": 8.064119628405443e-07,
+      "loss": 0.7665,
+      "step": 12846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4054298687128606,
+      "learning_rate": 8.058476166534818e-07,
+      "loss": 0.6529,
+      "step": 12847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1190004174679946,
+      "learning_rate": 8.052834506984874e-07,
+      "loss": 0.6309,
+      "step": 12848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8963112692514974,
+      "learning_rate": 8.047194649998063e-07,
+      "loss": 0.906,
+      "step": 12849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8667718492040788,
+      "learning_rate": 8.041556595816741e-07,
+      "loss": 0.6401,
+      "step": 12850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1667235008473957,
+      "learning_rate": 8.035920344683157e-07,
+      "loss": 0.6345,
+      "step": 12851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.62799963917866,
+      "learning_rate": 8.030285896839546e-07,
+      "loss": 0.6429,
+      "step": 12852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6961383377387604,
+      "learning_rate": 8.024653252528025e-07,
+      "loss": 0.6759,
+      "step": 12853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.556184023120913,
+      "learning_rate": 8.019022411990634e-07,
+      "loss": 0.6058,
+      "step": 12854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7919750524047449,
+      "learning_rate": 8.013393375469347e-07,
+      "loss": 0.6973,
+      "step": 12855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8495412257396142,
+      "learning_rate": 8.007766143206047e-07,
+      "loss": 0.6927,
+      "step": 12856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5725078999951048,
+      "learning_rate": 8.00214071544258e-07,
+      "loss": 0.7175,
+      "step": 12857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6481874070776457,
+      "learning_rate": 7.996517092420636e-07,
+      "loss": 0.7739,
+      "step": 12858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5278341537846007,
+      "learning_rate": 7.99089527438191e-07,
+      "loss": 0.7905,
+      "step": 12859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9137796210225393,
+      "learning_rate": 7.985275261567971e-07,
+      "loss": 0.7924,
+      "step": 12860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7617125480118823,
+      "learning_rate": 7.97965705422033e-07,
+      "loss": 0.6569,
+      "step": 12861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.761560153841894,
+      "learning_rate": 7.974040652580412e-07,
+      "loss": 0.7941,
+      "step": 12862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6790580184361377,
+      "learning_rate": 7.968426056889561e-07,
+      "loss": 0.7304,
+      "step": 12863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7485266551889092,
+      "learning_rate": 7.962813267389052e-07,
+      "loss": 0.5806,
+      "step": 12864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6098386070406359,
+      "learning_rate": 7.957202284320076e-07,
+      "loss": 0.7554,
+      "step": 12865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.580535996347156,
+      "learning_rate": 7.951593107923744e-07,
+      "loss": 0.6564,
+      "step": 12866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7355915790204692,
+      "learning_rate": 7.945985738441114e-07,
+      "loss": 0.6761,
+      "step": 12867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7910084124428423,
+      "learning_rate": 7.940380176113149e-07,
+      "loss": 0.6977,
+      "step": 12868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8516173578040513,
+      "learning_rate": 7.93477642118069e-07,
+      "loss": 0.7691,
+      "step": 12869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6371076380713474,
+      "learning_rate": 7.929174473884593e-07,
+      "loss": 0.6757,
+      "step": 12870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6355660850423481,
+      "learning_rate": 7.923574334465562e-07,
+      "loss": 0.6272,
+      "step": 12871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7684144758575708,
+      "learning_rate": 7.917976003164246e-07,
+      "loss": 0.799,
+      "step": 12872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2278845569772852,
+      "learning_rate": 7.912379480221228e-07,
+      "loss": 0.6205,
+      "step": 12873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.820437821763086,
+      "learning_rate": 7.906784765876985e-07,
+      "loss": 0.7182,
+      "step": 12874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5962496025668873,
+      "learning_rate": 7.90119186037197e-07,
+      "loss": 0.6596,
+      "step": 12875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.754958275853414,
+      "learning_rate": 7.895600763946482e-07,
+      "loss": 0.7568,
+      "step": 12876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9485652996723737,
+      "learning_rate": 7.890011476840809e-07,
+      "loss": 0.7137,
+      "step": 12877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.101948258146142,
+      "learning_rate": 7.884423999295122e-07,
+      "loss": 0.6098,
+      "step": 12878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6555300020333963,
+      "learning_rate": 7.878838331549538e-07,
+      "loss": 0.6161,
+      "step": 12879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8560593085912584,
+      "learning_rate": 7.873254473844077e-07,
+      "loss": 0.7349,
+      "step": 12880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7653369680844022,
+      "learning_rate": 7.867672426418693e-07,
+      "loss": 0.679,
+      "step": 12881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0930704152923254,
+      "learning_rate": 7.86209218951326e-07,
+      "loss": 0.7588,
+      "step": 12882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6472114309792742,
+      "learning_rate": 7.856513763367569e-07,
+      "loss": 0.703,
+      "step": 12883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7824773995616832,
+      "learning_rate": 7.850937148221332e-07,
+      "loss": 0.6845,
+      "step": 12884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2568979437965753,
+      "learning_rate": 7.845362344314206e-07,
+      "loss": 0.7025,
+      "step": 12885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6761287044564528,
+      "learning_rate": 7.839789351885746e-07,
+      "loss": 0.6735,
+      "step": 12886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0599880072774013,
+      "learning_rate": 7.834218171175428e-07,
+      "loss": 0.6547,
+      "step": 12887
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5485068193705973,
+      "learning_rate": 7.828648802422667e-07,
+      "loss": 0.6302,
+      "step": 12888
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.007946643032436,
+      "learning_rate": 7.823081245866776e-07,
+      "loss": 0.7989,
+      "step": 12889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9796132099949693,
+      "learning_rate": 7.817515501747041e-07,
+      "loss": 0.6912,
+      "step": 12890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7599281824021935,
+      "learning_rate": 7.811951570302594e-07,
+      "loss": 0.7461,
+      "step": 12891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3374656529593654,
+      "learning_rate": 7.806389451772539e-07,
+      "loss": 0.5504,
+      "step": 12892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6227812159825035,
+      "learning_rate": 7.800829146395922e-07,
+      "loss": 0.7084,
+      "step": 12893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6775055370201604,
+      "learning_rate": 7.795270654411635e-07,
+      "loss": 0.6737,
+      "step": 12894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.088040723194621,
+      "learning_rate": 7.789713976058578e-07,
+      "loss": 0.6872,
+      "step": 12895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7320746774582674,
+      "learning_rate": 7.784159111575512e-07,
+      "loss": 0.8661,
+      "step": 12896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4520963032908674,
+      "learning_rate": 7.778606061201154e-07,
+      "loss": 0.6693,
+      "step": 12897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.725555833991763,
+      "learning_rate": 7.773054825174126e-07,
+      "loss": 0.6788,
+      "step": 12898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9260343152838044,
+      "learning_rate": 7.767505403732961e-07,
+      "loss": 0.7221,
+      "step": 12899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8291623056269215,
+      "learning_rate": 7.761957797116171e-07,
+      "loss": 0.7279,
+      "step": 12900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8651851460234943,
+      "learning_rate": 7.756412005562114e-07,
+      "loss": 0.7623,
+      "step": 12901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.640730048119487,
+      "learning_rate": 7.750868029309099e-07,
+      "loss": 0.6596,
+      "step": 12902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7798153160811232,
+      "learning_rate": 7.745325868595394e-07,
+      "loss": 0.6471,
+      "step": 12903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.732771368939862,
+      "learning_rate": 7.739785523659144e-07,
+      "loss": 0.6907,
+      "step": 12904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8806401795305698,
+      "learning_rate": 7.734246994738426e-07,
+      "loss": 0.6698,
+      "step": 12905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7783612216233213,
+      "learning_rate": 7.728710282071244e-07,
+      "loss": 0.8522,
+      "step": 12906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.932908266078203,
+      "learning_rate": 7.723175385895509e-07,
+      "loss": 0.7491,
+      "step": 12907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.801946390726688,
+      "learning_rate": 7.717642306449113e-07,
+      "loss": 0.7674,
+      "step": 12908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0729942293255983,
+      "learning_rate": 7.712111043969772e-07,
+      "loss": 0.6169,
+      "step": 12909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.491970007041833,
+      "learning_rate": 7.706581598695207e-07,
+      "loss": 0.6799,
+      "step": 12910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6385529753330164,
+      "learning_rate": 7.701053970863037e-07,
+      "loss": 0.6755,
+      "step": 12911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4259715041187953,
+      "learning_rate": 7.695528160710764e-07,
+      "loss": 0.7244,
+      "step": 12912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3065219877242265,
+      "learning_rate": 7.690004168475868e-07,
+      "loss": 0.6191,
+      "step": 12913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8288288177810283,
+      "learning_rate": 7.684481994395726e-07,
+      "loss": 0.6738,
+      "step": 12914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.585249161991323,
+      "learning_rate": 7.678961638707633e-07,
+      "loss": 0.6397,
+      "step": 12915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.218928447027698,
+      "learning_rate": 7.673443101648808e-07,
+      "loss": 0.7122,
+      "step": 12916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.98823620329753,
+      "learning_rate": 7.667926383456392e-07,
+      "loss": 0.776,
+      "step": 12917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7600191189147272,
+      "learning_rate": 7.662411484367482e-07,
+      "loss": 0.7527,
+      "step": 12918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6877383746663213,
+      "learning_rate": 7.656898404619029e-07,
+      "loss": 0.7652,
+      "step": 12919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7088829478435112,
+      "learning_rate": 7.651387144447942e-07,
+      "loss": 0.7541,
+      "step": 12920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9492917506997065,
+      "learning_rate": 7.645877704091082e-07,
+      "loss": 0.6747,
+      "step": 12921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7609221667255457,
+      "learning_rate": 7.640370083785175e-07,
+      "loss": 0.6587,
+      "step": 12922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9420024131697353,
+      "learning_rate": 7.634864283766913e-07,
+      "loss": 0.8187,
+      "step": 12923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8131742335977088,
+      "learning_rate": 7.629360304272882e-07,
+      "loss": 0.7631,
+      "step": 12924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.55559426976853,
+      "learning_rate": 7.623858145539592e-07,
+      "loss": 0.6337,
+      "step": 12925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.217656858913942,
+      "learning_rate": 7.618357807803517e-07,
+      "loss": 0.7526,
+      "step": 12926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1275914506610576,
+      "learning_rate": 7.612859291300973e-07,
+      "loss": 0.6537,
+      "step": 12927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6335880283456856,
+      "learning_rate": 7.607362596268281e-07,
+      "loss": 0.86,
+      "step": 12928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1582379299157555,
+      "learning_rate": 7.601867722941642e-07,
+      "loss": 0.6568,
+      "step": 12929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8019894864312371,
+      "learning_rate": 7.596374671557144e-07,
+      "loss": 0.7123,
+      "step": 12930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9968976296524437,
+      "learning_rate": 7.590883442350883e-07,
+      "loss": 0.7292,
+      "step": 12931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.67516529866836,
+      "learning_rate": 7.585394035558807e-07,
+      "loss": 0.7347,
+      "step": 12932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6514362057369802,
+      "learning_rate": 7.579906451416813e-07,
+      "loss": 0.6761,
+      "step": 12933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0906353292999853,
+      "learning_rate": 7.57442069016071e-07,
+      "loss": 0.6607,
+      "step": 12934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0700918626647464,
+      "learning_rate": 7.568936752026229e-07,
+      "loss": 0.6198,
+      "step": 12935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6658307930110612,
+      "learning_rate": 7.563454637249056e-07,
+      "loss": 0.6638,
+      "step": 12936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5959189335208859,
+      "learning_rate": 7.557974346064728e-07,
+      "loss": 0.6543,
+      "step": 12937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6558497183510341,
+      "learning_rate": 7.552495878708776e-07,
+      "loss": 0.7582,
+      "step": 12938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.534027565109878,
+      "learning_rate": 7.547019235416609e-07,
+      "loss": 0.7702,
+      "step": 12939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7738078055913653,
+      "learning_rate": 7.541544416423574e-07,
+      "loss": 0.6346,
+      "step": 12940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8729696742304693,
+      "learning_rate": 7.536071421964941e-07,
+      "loss": 0.636,
+      "step": 12941
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6905431441816778,
+      "learning_rate": 7.530600252275888e-07,
+      "loss": 0.6431,
+      "step": 12942
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9918672507039933,
+      "learning_rate": 7.52513090759151e-07,
+      "loss": 0.5368,
+      "step": 12943
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9889546747188023,
+      "learning_rate": 7.519663388146886e-07,
+      "loss": 0.8198,
+      "step": 12944
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6218392653250462,
+      "learning_rate": 7.514197694176905e-07,
+      "loss": 0.653,
+      "step": 12945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6399036801319982,
+      "learning_rate": 7.50873382591648e-07,
+      "loss": 0.7216,
+      "step": 12946
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8110659569131065,
+      "learning_rate": 7.503271783600402e-07,
+      "loss": 0.8156,
+      "step": 12947
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.666860633724129,
+      "learning_rate": 7.497811567463381e-07,
+      "loss": 0.8054,
+      "step": 12948
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.727484808541349,
+      "learning_rate": 7.492353177740047e-07,
+      "loss": 0.6809,
+      "step": 12949
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.792742844833414,
+      "learning_rate": 7.486896614664962e-07,
+      "loss": 0.6353,
+      "step": 12950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2366450295482483,
+      "learning_rate": 7.481441878472629e-07,
+      "loss": 0.6581,
+      "step": 12951
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7240499214203548,
+      "learning_rate": 7.475988969397424e-07,
+      "loss": 0.6486,
+      "step": 12952
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.606521423120446,
+      "learning_rate": 7.470537887673667e-07,
+      "loss": 0.686,
+      "step": 12953
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1742452610001546,
+      "learning_rate": 7.465088633535639e-07,
+      "loss": 0.7822,
+      "step": 12954
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6725270166882962,
+      "learning_rate": 7.459641207217461e-07,
+      "loss": 0.6672,
+      "step": 12955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2685877967396635,
+      "learning_rate": 7.454195608953252e-07,
+      "loss": 0.7835,
+      "step": 12956
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0682259376977172,
+      "learning_rate": 7.448751838977014e-07,
+      "loss": 0.6779,
+      "step": 12957
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2065373405744582,
+      "learning_rate": 7.443309897522671e-07,
+      "loss": 0.5982,
+      "step": 12958
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5124025781750072,
+      "learning_rate": 7.437869784824086e-07,
+      "loss": 0.7923,
+      "step": 12959
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.600247585385879,
+      "learning_rate": 7.432431501115012e-07,
+      "loss": 0.7255,
+      "step": 12960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7725061991652629,
+      "learning_rate": 7.426995046629176e-07,
+      "loss": 0.7052,
+      "step": 12961
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4011902350222174,
+      "learning_rate": 7.421560421600182e-07,
+      "loss": 0.6436,
+      "step": 12962
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.191939781120955,
+      "learning_rate": 7.416127626261549e-07,
+      "loss": 0.7567,
+      "step": 12963
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0596119352988718,
+      "learning_rate": 7.410696660846761e-07,
+      "loss": 0.6018,
+      "step": 12964
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.944498639280064,
+      "learning_rate": 7.405267525589183e-07,
+      "loss": 0.7227,
+      "step": 12965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7016183504478612,
+      "learning_rate": 7.399840220722127e-07,
+      "loss": 0.7213,
+      "step": 12966
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8598574690679213,
+      "learning_rate": 7.394414746478817e-07,
+      "loss": 0.7046,
+      "step": 12967
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6199715184140737,
+      "learning_rate": 7.388991103092374e-07,
+      "loss": 0.6761,
+      "step": 12968
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6413319196824427,
+      "learning_rate": 7.383569290795911e-07,
+      "loss": 0.5899,
+      "step": 12969
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5869197497049485,
+      "learning_rate": 7.378149309822374e-07,
+      "loss": 0.7683,
+      "step": 12970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8170580928059514,
+      "learning_rate": 7.372731160404672e-07,
+      "loss": 0.7688,
+      "step": 12971
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1591224016539592,
+      "learning_rate": 7.367314842775674e-07,
+      "loss": 0.6774,
+      "step": 12972
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0713308972794844,
+      "learning_rate": 7.361900357168084e-07,
+      "loss": 0.6034,
+      "step": 12973
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7766731374774507,
+      "learning_rate": 7.356487703814602e-07,
+      "loss": 0.8306,
+      "step": 12974
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7438074926463043,
+      "learning_rate": 7.351076882947817e-07,
+      "loss": 0.6635,
+      "step": 12975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7369939201145068,
+      "learning_rate": 7.345667894800246e-07,
+      "loss": 0.7774,
+      "step": 12976
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5827821279500103,
+      "learning_rate": 7.340260739604316e-07,
+      "loss": 0.7038,
+      "step": 12977
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8648598073999734,
+      "learning_rate": 7.334855417592385e-07,
+      "loss": 0.8081,
+      "step": 12978
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5342381737502033,
+      "learning_rate": 7.329451928996745e-07,
+      "loss": 0.6964,
+      "step": 12979
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.241583473280332,
+      "learning_rate": 7.324050274049599e-07,
+      "loss": 0.7048,
+      "step": 12980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9229846111845152,
+      "learning_rate": 7.318650452983039e-07,
+      "loss": 0.6755,
+      "step": 12981
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6300064142411927,
+      "learning_rate": 7.313252466029131e-07,
+      "loss": 0.6618,
+      "step": 12982
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8456517023184167,
+      "learning_rate": 7.307856313419837e-07,
+      "loss": 0.6649,
+      "step": 12983
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9655266161983225,
+      "learning_rate": 7.302461995387033e-07,
+      "loss": 0.8277,
+      "step": 12984
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5431592968089982,
+      "learning_rate": 7.297069512162535e-07,
+      "loss": 0.6194,
+      "step": 12985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8556708961252284,
+      "learning_rate": 7.291678863978052e-07,
+      "loss": 0.6361,
+      "step": 12986
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7468005542905276,
+      "learning_rate": 7.286290051065264e-07,
+      "loss": 0.7793,
+      "step": 12987
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5097915571712794,
+      "learning_rate": 7.280903073655704e-07,
+      "loss": 0.7364,
+      "step": 12988
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4400243949917817,
+      "learning_rate": 7.275517931980886e-07,
+      "loss": 0.6331,
+      "step": 12989
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9571021499985772,
+      "learning_rate": 7.270134626272229e-07,
+      "loss": 0.6908,
+      "step": 12990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8009322229080629,
+      "learning_rate": 7.264753156761028e-07,
+      "loss": 0.6798,
+      "step": 12991
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.622344184232143,
+      "learning_rate": 7.25937352367857e-07,
+      "loss": 0.6673,
+      "step": 12992
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5267850621567438,
+      "learning_rate": 7.253995727256024e-07,
+      "loss": 0.6427,
+      "step": 12993
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1261899103393473,
+      "learning_rate": 7.24861976772448e-07,
+      "loss": 0.6601,
+      "step": 12994
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.249440628566209,
+      "learning_rate": 7.243245645314955e-07,
+      "loss": 0.6469,
+      "step": 12995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2177702099596828,
+      "learning_rate": 7.23787336025838e-07,
+      "loss": 0.8287,
+      "step": 12996
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7164004780073685,
+      "learning_rate": 7.232502912785633e-07,
+      "loss": 0.666,
+      "step": 12997
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7674102548222246,
+      "learning_rate": 7.227134303127487e-07,
+      "loss": 0.723,
+      "step": 12998
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6776342869172938,
+      "learning_rate": 7.22176753151464e-07,
+      "loss": 0.7583,
+      "step": 12999
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.226971807624395,
+      "learning_rate": 7.216402598177713e-07,
+      "loss": 0.6801,
+      "step": 13000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7417257134948203,
+      "learning_rate": 7.21103950334725e-07,
+      "loss": 0.7031,
+      "step": 13001
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3962295223935042,
+      "learning_rate": 7.205678247253717e-07,
+      "loss": 0.7049,
+      "step": 13002
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6610925114529036,
+      "learning_rate": 7.200318830127495e-07,
+      "loss": 0.6562,
+      "step": 13003
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3312421994373274,
+      "learning_rate": 7.194961252198885e-07,
+      "loss": 0.6402,
+      "step": 13004
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7484450262221496,
+      "learning_rate": 7.189605513698139e-07,
+      "loss": 0.7463,
+      "step": 13005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4688344992062476,
+      "learning_rate": 7.184251614855369e-07,
+      "loss": 0.7331,
+      "step": 13006
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6777231035798412,
+      "learning_rate": 7.178899555900675e-07,
+      "loss": 0.5911,
+      "step": 13007
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5387573955715967,
+      "learning_rate": 7.173549337064035e-07,
+      "loss": 0.703,
+      "step": 13008
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6874530717076,
+      "learning_rate": 7.168200958575361e-07,
+      "loss": 0.7042,
+      "step": 13009
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7162602768691584,
+      "learning_rate": 7.162854420664478e-07,
+      "loss": 0.6704,
+      "step": 13010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.78806898355492,
+      "learning_rate": 7.157509723561134e-07,
+      "loss": 0.6646,
+      "step": 13011
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0940301386576639,
+      "learning_rate": 7.152166867495042e-07,
+      "loss": 0.6133,
+      "step": 13012
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.937925263575989,
+      "learning_rate": 7.146825852695749e-07,
+      "loss": 0.5867,
+      "step": 13013
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8337197074167124,
+      "learning_rate": 7.141486679392778e-07,
+      "loss": 0.6733,
+      "step": 13014
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.103364766090575,
+      "learning_rate": 7.136149347815591e-07,
+      "loss": 0.7174,
+      "step": 13015
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1662403089455333,
+      "learning_rate": 7.130813858193525e-07,
+      "loss": 0.5537,
+      "step": 13016
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7166544717224919,
+      "learning_rate": 7.125480210755858e-07,
+      "loss": 0.7566,
+      "step": 13017
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8046055477044174,
+      "learning_rate": 7.120148405731803e-07,
+      "loss": 0.7147,
+      "step": 13018
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.648309097369082,
+      "learning_rate": 7.114818443350463e-07,
+      "loss": 0.6257,
+      "step": 13019
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5609439898556436,
+      "learning_rate": 7.109490323840884e-07,
+      "loss": 0.6361,
+      "step": 13020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0816572398285802,
+      "learning_rate": 7.104164047432022e-07,
+      "loss": 0.6958,
+      "step": 13021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8914516055331403,
+      "learning_rate": 7.098839614352775e-07,
+      "loss": 0.6645,
+      "step": 13022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8670710464856584,
+      "learning_rate": 7.093517024831941e-07,
+      "loss": 0.6918,
+      "step": 13023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0321837418835345,
+      "learning_rate": 7.088196279098225e-07,
+      "loss": 0.6102,
+      "step": 13024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6692778022612265,
+      "learning_rate": 7.082877377380293e-07,
+      "loss": 0.8322,
+      "step": 13025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9618032799731433,
+      "learning_rate": 7.077560319906696e-07,
+      "loss": 0.7514,
+      "step": 13026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8011355324879812,
+      "learning_rate": 7.072245106905928e-07,
+      "loss": 0.7888,
+      "step": 13027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6921491774691455,
+      "learning_rate": 7.066931738606397e-07,
+      "loss": 0.8419,
+      "step": 13028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7403440248674151,
+      "learning_rate": 7.061620215236415e-07,
+      "loss": 0.7353,
+      "step": 13029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.948192154707534,
+      "learning_rate": 7.056310537024257e-07,
+      "loss": 0.7416,
+      "step": 13030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.987227911616414,
+      "learning_rate": 7.051002704198074e-07,
+      "loss": 0.5942,
+      "step": 13031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9036718983012888,
+      "learning_rate": 7.045696716985939e-07,
+      "loss": 0.7559,
+      "step": 13032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7118890248592702,
+      "learning_rate": 7.040392575615896e-07,
+      "loss": 0.7284,
+      "step": 13033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6048024524547677,
+      "learning_rate": 7.035090280315854e-07,
+      "loss": 0.6946,
+      "step": 13034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5842254464109957,
+      "learning_rate": 7.029789831313677e-07,
+      "loss": 0.7032,
+      "step": 13035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6707306240964572,
+      "learning_rate": 7.024491228837127e-07,
+      "loss": 0.7404,
+      "step": 13036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9600845749255398,
+      "learning_rate": 7.0191944731139e-07,
+      "loss": 0.6408,
+      "step": 13037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6007717865982711,
+      "learning_rate": 7.013899564371613e-07,
+      "loss": 0.6786,
+      "step": 13038
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7806638995434412,
+      "learning_rate": 7.008606502837784e-07,
+      "loss": 0.7738,
+      "step": 13039
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7540272397717294,
+      "learning_rate": 7.003315288739893e-07,
+      "loss": 0.5926,
+      "step": 13040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.91249161109636,
+      "learning_rate": 6.998025922305313e-07,
+      "loss": 0.7551,
+      "step": 13041
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.709528503208186,
+      "learning_rate": 6.992738403761307e-07,
+      "loss": 0.6658,
+      "step": 13042
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4828250404246832,
+      "learning_rate": 6.987452733335126e-07,
+      "loss": 0.6382,
+      "step": 13043
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7187857602580536,
+      "learning_rate": 6.982168911253895e-07,
+      "loss": 0.6935,
+      "step": 13044
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8551124685599547,
+      "learning_rate": 6.97688693774467e-07,
+      "loss": 0.7791,
+      "step": 13045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.37741493675225,
+      "learning_rate": 6.971606813034432e-07,
+      "loss": 0.7106,
+      "step": 13046
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7646179321290736,
+      "learning_rate": 6.96632853735007e-07,
+      "loss": 0.712,
+      "step": 13047
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.400986464822154,
+      "learning_rate": 6.961052110918432e-07,
+      "loss": 0.6424,
+      "step": 13048
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8400390795663464,
+      "learning_rate": 6.955777533966212e-07,
+      "loss": 0.6917,
+      "step": 13049
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1152354230746848,
+      "learning_rate": 6.950504806720115e-07,
+      "loss": 0.7078,
+      "step": 13050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3685389678743811,
+      "learning_rate": 6.945233929406697e-07,
+      "loss": 0.6124,
+      "step": 13051
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0139014320423003,
+      "learning_rate": 6.939964902252466e-07,
+      "loss": 0.9096,
+      "step": 13052
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6316677095945036,
+      "learning_rate": 6.934697725483846e-07,
+      "loss": 0.7845,
+      "step": 13053
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9835768989159968,
+      "learning_rate": 6.929432399327174e-07,
+      "loss": 0.6317,
+      "step": 13054
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9985275726472866,
+      "learning_rate": 6.924168924008712e-07,
+      "loss": 0.6433,
+      "step": 13055
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5752484560918432,
+      "learning_rate": 6.918907299754651e-07,
+      "loss": 0.7657,
+      "step": 13056
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8700294631741021,
+      "learning_rate": 6.913647526791078e-07,
+      "loss": 0.6532,
+      "step": 13057
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9956987302261016,
+      "learning_rate": 6.90838960534404e-07,
+      "loss": 0.7937,
+      "step": 13058
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5693758264187425,
+      "learning_rate": 6.903133535639467e-07,
+      "loss": 0.7674,
+      "step": 13059
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.764205952808154,
+      "learning_rate": 6.897879317903233e-07,
+      "loss": 0.7549,
+      "step": 13060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5540275489743878,
+      "learning_rate": 6.89262695236112e-07,
+      "loss": 0.6594,
+      "step": 13061
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4712897203212556,
+      "learning_rate": 6.887376439238813e-07,
+      "loss": 0.7103,
+      "step": 13062
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6325613126782894,
+      "learning_rate": 6.882127778761987e-07,
+      "loss": 0.6411,
+      "step": 13063
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6193753325874216,
+      "learning_rate": 6.876880971156147e-07,
+      "loss": 0.5481,
+      "step": 13064
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.123438163867768,
+      "learning_rate": 6.871636016646754e-07,
+      "loss": 0.6618,
+      "step": 13065
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3417077648981688,
+      "learning_rate": 6.866392915459241e-07,
+      "loss": 0.6514,
+      "step": 13066
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8136948480091586,
+      "learning_rate": 6.861151667818866e-07,
+      "loss": 0.8044,
+      "step": 13067
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.734480538832143,
+      "learning_rate": 6.855912273950893e-07,
+      "loss": 0.719,
+      "step": 13068
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8728363015934892,
+      "learning_rate": 6.850674734080454e-07,
+      "loss": 0.6559,
+      "step": 13069
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6494307960762038,
+      "learning_rate": 6.845439048432617e-07,
+      "loss": 0.7433,
+      "step": 13070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9116160501733395,
+      "learning_rate": 6.840205217232376e-07,
+      "loss": 0.7491,
+      "step": 13071
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.676907392825834,
+      "learning_rate": 6.834973240704629e-07,
+      "loss": 0.6648,
+      "step": 13072
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.140542350516774,
+      "learning_rate": 6.829743119074239e-07,
+      "loss": 0.7493,
+      "step": 13073
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6107645444928589,
+      "learning_rate": 6.824514852565922e-07,
+      "loss": 0.6258,
+      "step": 13074
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.207104270782199,
+      "learning_rate": 6.819288441404343e-07,
+      "loss": 0.837,
+      "step": 13075
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6446336651789089,
+      "learning_rate": 6.814063885814127e-07,
+      "loss": 0.6426,
+      "step": 13076
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.93036299810974,
+      "learning_rate": 6.80884118601976e-07,
+      "loss": 0.7353,
+      "step": 13077
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0510469827079105,
+      "learning_rate": 6.803620342245676e-07,
+      "loss": 0.5733,
+      "step": 13078
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6159579803763517,
+      "learning_rate": 6.798401354716233e-07,
+      "loss": 0.7121,
+      "step": 13079
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8418446425373594,
+      "learning_rate": 6.793184223655691e-07,
+      "loss": 0.6522,
+      "step": 13080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.452448726723077,
+      "learning_rate": 6.78796894928827e-07,
+      "loss": 0.6169,
+      "step": 13081
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7703577244766482,
+      "learning_rate": 6.782755531838047e-07,
+      "loss": 0.7556,
+      "step": 13082
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5253492233189865,
+      "learning_rate": 6.77754397152906e-07,
+      "loss": 0.7303,
+      "step": 13083
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7410773715386185,
+      "learning_rate": 6.772334268585296e-07,
+      "loss": 0.684,
+      "step": 13084
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7899036449600807,
+      "learning_rate": 6.767126423230574e-07,
+      "loss": 0.6416,
+      "step": 13085
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7674104730940556,
+      "learning_rate": 6.76192043568873e-07,
+      "loss": 0.7319,
+      "step": 13086
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.677160488547989,
+      "learning_rate": 6.756716306183459e-07,
+      "loss": 0.6812,
+      "step": 13087
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7462107604622739,
+      "learning_rate": 6.751514034938395e-07,
+      "loss": 0.6409,
+      "step": 13088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8580732850573094,
+      "learning_rate": 6.746313622177097e-07,
+      "loss": 0.6832,
+      "step": 13089
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8932307714023149,
+      "learning_rate": 6.741115068123017e-07,
+      "loss": 0.6921,
+      "step": 13090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6420105164757393,
+      "learning_rate": 6.73591837299959e-07,
+      "loss": 0.6753,
+      "step": 13091
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7344293982506869,
+      "learning_rate": 6.730723537030092e-07,
+      "loss": 0.653,
+      "step": 13092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6529262200253587,
+      "learning_rate": 6.725530560437759e-07,
+      "loss": 0.6683,
+      "step": 13093
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5595196864409289,
+      "learning_rate": 6.720339443445772e-07,
+      "loss": 0.7861,
+      "step": 13094
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.90543431743506,
+      "learning_rate": 6.71515018627718e-07,
+      "loss": 0.681,
+      "step": 13095
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0718603109502802,
+      "learning_rate": 6.709962789154989e-07,
+      "loss": 0.6902,
+      "step": 13096
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6989541964039545,
+      "learning_rate": 6.704777252302108e-07,
+      "loss": 0.6605,
+      "step": 13097
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7027994809849816,
+      "learning_rate": 6.699593575941365e-07,
+      "loss": 0.754,
+      "step": 13098
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.678547304071599,
+      "learning_rate": 6.694411760295538e-07,
+      "loss": 0.7523,
+      "step": 13099
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9318537534142384,
+      "learning_rate": 6.689231805587265e-07,
+      "loss": 0.8125,
+      "step": 13100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5822511521840157,
+      "learning_rate": 6.684053712039168e-07,
+      "loss": 0.6198,
+      "step": 13101
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7042064114062454,
+      "learning_rate": 6.678877479873769e-07,
+      "loss": 0.7114,
+      "step": 13102
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.641801261578784,
+      "learning_rate": 6.673703109313456e-07,
+      "loss": 0.7442,
+      "step": 13103
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2557303395624084,
+      "learning_rate": 6.66853060058063e-07,
+      "loss": 0.7071,
+      "step": 13104
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7067203697435713,
+      "learning_rate": 6.663359953897547e-07,
+      "loss": 0.6271,
+      "step": 13105
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.718566762202855,
+      "learning_rate": 6.658191169486405e-07,
+      "loss": 0.6938,
+      "step": 13106
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6578593220294704,
+      "learning_rate": 6.653024247569312e-07,
+      "loss": 0.6179,
+      "step": 13107
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0331475791338525,
+      "learning_rate": 6.647859188368294e-07,
+      "loss": 0.7172,
+      "step": 13108
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.743154418057879,
+      "learning_rate": 6.642695992105347e-07,
+      "loss": 0.7974,
+      "step": 13109
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5779606536095987,
+      "learning_rate": 6.637534659002287e-07,
+      "loss": 0.673,
+      "step": 13110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.657035003003982,
+      "learning_rate": 6.632375189280948e-07,
+      "loss": 0.6056,
+      "step": 13111
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.938412400763961,
+      "learning_rate": 6.627217583163031e-07,
+      "loss": 0.6747,
+      "step": 13112
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8288279183430785,
+      "learning_rate": 6.622061840870169e-07,
+      "loss": 0.6716,
+      "step": 13113
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7464837913733766,
+      "learning_rate": 6.61690796262392e-07,
+      "loss": 0.7104,
+      "step": 13114
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1382347002470117,
+      "learning_rate": 6.61175594864576e-07,
+      "loss": 0.6627,
+      "step": 13115
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8722347244149202,
+      "learning_rate": 6.606605799157062e-07,
+      "loss": 0.8037,
+      "step": 13116
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.230363274369291,
+      "learning_rate": 6.601457514379173e-07,
+      "loss": 0.6633,
+      "step": 13117
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4806876579260084,
+      "learning_rate": 6.596311094533292e-07,
+      "loss": 0.6332,
+      "step": 13118
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7390666061524707,
+      "learning_rate": 6.591166539840599e-07,
+      "loss": 0.6857,
+      "step": 13119
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.424538329303785,
+      "learning_rate": 6.586023850522156e-07,
+      "loss": 0.6572,
+      "step": 13120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.149096751843857,
+      "learning_rate": 6.580883026798957e-07,
+      "loss": 0.5496,
+      "step": 13121
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0537213657800961,
+      "learning_rate": 6.575744068891909e-07,
+      "loss": 0.7023,
+      "step": 13122
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.321109671414076,
+      "learning_rate": 6.570606977021842e-07,
+      "loss": 0.7719,
+      "step": 13123
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5366479451649395,
+      "learning_rate": 6.565471751409541e-07,
+      "loss": 0.7362,
+      "step": 13124
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.751340520312619,
+      "learning_rate": 6.56033839227564e-07,
+      "loss": 0.7378,
+      "step": 13125
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.021394641975625,
+      "learning_rate": 6.555206899840727e-07,
+      "loss": 0.6761,
+      "step": 13126
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8589857253926516,
+      "learning_rate": 6.550077274325345e-07,
+      "loss": 0.7302,
+      "step": 13127
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8224220592322595,
+      "learning_rate": 6.544949515949911e-07,
+      "loss": 0.7141,
+      "step": 13128
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7462573070769911,
+      "learning_rate": 6.539823624934777e-07,
+      "loss": 0.7024,
+      "step": 13129
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6888750167823177,
+      "learning_rate": 6.53469960150021e-07,
+      "loss": 0.7129,
+      "step": 13130
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7158636666433882,
+      "learning_rate": 6.529577445866409e-07,
+      "loss": 0.7985,
+      "step": 13131
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1886053052325964,
+      "learning_rate": 6.524457158253472e-07,
+      "loss": 0.6197,
+      "step": 13132
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.821482883880476,
+      "learning_rate": 6.519338738881442e-07,
+      "loss": 0.723,
+      "step": 13133
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.83307630343836,
+      "learning_rate": 6.514222187970248e-07,
+      "loss": 0.7166,
+      "step": 13134
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.731475836044587,
+      "learning_rate": 6.509107505739798e-07,
+      "loss": 0.7508,
+      "step": 13135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6512637439781024,
+      "learning_rate": 6.503994692409838e-07,
+      "loss": 0.7339,
+      "step": 13136
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5476442126242913,
+      "learning_rate": 6.498883748200102e-07,
+      "loss": 0.6723,
+      "step": 13137
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6045861769380132,
+      "learning_rate": 6.493774673330217e-07,
+      "loss": 0.725,
+      "step": 13138
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4976720989389285,
+      "learning_rate": 6.488667468019727e-07,
+      "loss": 0.7513,
+      "step": 13139
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9315390207750118,
+      "learning_rate": 6.483562132488103e-07,
+      "loss": 0.68,
+      "step": 13140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8527185718975483,
+      "learning_rate": 6.478458666954718e-07,
+      "loss": 0.7442,
+      "step": 13141
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7244411770649342,
+      "learning_rate": 6.473357071638914e-07,
+      "loss": 0.771,
+      "step": 13142
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.228635823188803,
+      "learning_rate": 6.468257346759888e-07,
+      "loss": 0.6754,
+      "step": 13143
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6968100728467097,
+      "learning_rate": 6.46315949253678e-07,
+      "loss": 0.6152,
+      "step": 13144
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7885318616734733,
+      "learning_rate": 6.458063509188678e-07,
+      "loss": 0.7809,
+      "step": 13145
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1367458960955825,
+      "learning_rate": 6.452969396934567e-07,
+      "loss": 0.7767,
+      "step": 13146
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2084919664760068,
+      "learning_rate": 6.447877155993343e-07,
+      "loss": 0.5999,
+      "step": 13147
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5873831481119263,
+      "learning_rate": 6.442786786583832e-07,
+      "loss": 0.7347,
+      "step": 13148
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1175014612866194,
+      "learning_rate": 6.437698288924777e-07,
+      "loss": 0.6289,
+      "step": 13149
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.91841188239984,
+      "learning_rate": 6.432611663234851e-07,
+      "loss": 0.8348,
+      "step": 13150
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.735946968617579,
+      "learning_rate": 6.427526909732617e-07,
+      "loss": 0.7534,
+      "step": 13151
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5089940182741077,
+      "learning_rate": 6.422444028636604e-07,
+      "loss": 0.7167,
+      "step": 13152
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7101866616246242,
+      "learning_rate": 6.417363020165235e-07,
+      "loss": 0.6956,
+      "step": 13153
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.710485902739192,
+      "learning_rate": 6.412283884536818e-07,
+      "loss": 0.714,
+      "step": 13154
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1774958671792832,
+      "learning_rate": 6.407206621969648e-07,
+      "loss": 0.6944,
+      "step": 13155
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2455363887365025,
+      "learning_rate": 6.402131232681897e-07,
+      "loss": 0.5213,
+      "step": 13156
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6623881809658845,
+      "learning_rate": 6.397057716891669e-07,
+      "loss": 0.6817,
+      "step": 13157
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.500781349721985,
+      "learning_rate": 6.391986074816975e-07,
+      "loss": 0.5951,
+      "step": 13158
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8722883379058803,
+      "learning_rate": 6.38691630667575e-07,
+      "loss": 0.7241,
+      "step": 13159
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.66522032431491,
+      "learning_rate": 6.381848412685882e-07,
+      "loss": 0.6598,
+      "step": 13160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.779557440469753,
+      "learning_rate": 6.376782393065111e-07,
+      "loss": 0.7057,
+      "step": 13161
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9100785322238591,
+      "learning_rate": 6.371718248031167e-07,
+      "loss": 0.7555,
+      "step": 13162
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6942403893944757,
+      "learning_rate": 6.36665597780165e-07,
+      "loss": 0.6404,
+      "step": 13163
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.879441320444088,
+      "learning_rate": 6.36159558259411e-07,
+      "loss": 0.727,
+      "step": 13164
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5448052230822111,
+      "learning_rate": 6.35653706262599e-07,
+      "loss": 0.5888,
+      "step": 13165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4389058840542563,
+      "learning_rate": 6.351480418114675e-07,
+      "loss": 0.6629,
+      "step": 13166
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5037385530754819,
+      "learning_rate": 6.346425649277454e-07,
+      "loss": 0.653,
+      "step": 13167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.044366926093898,
+      "learning_rate": 6.341372756331543e-07,
+      "loss": 0.6623,
+      "step": 13168
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2871562047725407,
+      "learning_rate": 6.336321739494072e-07,
+      "loss": 0.7648,
+      "step": 13169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.914966384898911,
+      "learning_rate": 6.331272598982108e-07,
+      "loss": 0.7586,
+      "step": 13170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7359739636429785,
+      "learning_rate": 6.326225335012615e-07,
+      "loss": 0.6827,
+      "step": 13171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8205160028741585,
+      "learning_rate": 6.321179947802492e-07,
+      "loss": 0.6226,
+      "step": 13172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6375533643902511,
+      "learning_rate": 6.316136437568538e-07,
+      "loss": 0.6803,
+      "step": 13173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7847945166600436,
+      "learning_rate": 6.31109480452749e-07,
+      "loss": 0.59,
+      "step": 13174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9913157491123383,
+      "learning_rate": 6.306055048896004e-07,
+      "loss": 0.6052,
+      "step": 13175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6544258255129196,
+      "learning_rate": 6.301017170890639e-07,
+      "loss": 0.6821,
+      "step": 13176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.678505919081579,
+      "learning_rate": 6.295981170727883e-07,
+      "loss": 0.7843,
+      "step": 13177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6870949859498412,
+      "learning_rate": 6.290947048624169e-07,
+      "loss": 0.8084,
+      "step": 13178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8457364665263483,
+      "learning_rate": 6.285914804795784e-07,
+      "loss": 0.7022,
+      "step": 13179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6575014764030056,
+      "learning_rate": 6.280884439459006e-07,
+      "loss": 0.741,
+      "step": 13180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7250615721504883,
+      "learning_rate": 6.275855952829995e-07,
+      "loss": 0.7647,
+      "step": 13181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7488022464745523,
+      "learning_rate": 6.270829345124829e-07,
+      "loss": 0.7271,
+      "step": 13182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.867517296164495,
+      "learning_rate": 6.265804616559518e-07,
+      "loss": 0.5945,
+      "step": 13183
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5338914871365543,
+      "learning_rate": 6.260781767349983e-07,
+      "loss": 0.5375,
+      "step": 13184
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6498245268078655,
+      "learning_rate": 6.255760797712068e-07,
+      "loss": 0.7744,
+      "step": 13185
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4955079273926872,
+      "learning_rate": 6.250741707861535e-07,
+      "loss": 0.6362,
+      "step": 13186
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7656683157832584,
+      "learning_rate": 6.245724498014055e-07,
+      "loss": 0.7463,
+      "step": 13187
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2539842198715703,
+      "learning_rate": 6.240709168385251e-07,
+      "loss": 0.6739,
+      "step": 13188
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7816078864476625,
+      "learning_rate": 6.235695719190632e-07,
+      "loss": 0.7395,
+      "step": 13189
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8258044131342608,
+      "learning_rate": 6.230684150645633e-07,
+      "loss": 0.8154,
+      "step": 13190
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8635863747162538,
+      "learning_rate": 6.22567446296562e-07,
+      "loss": 0.7318,
+      "step": 13191
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.781769206701591,
+      "learning_rate": 6.220666656365865e-07,
+      "loss": 0.699,
+      "step": 13192
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2172376672393266,
+      "learning_rate": 6.215660731061563e-07,
+      "loss": 0.6956,
+      "step": 13193
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7494608002388583,
+      "learning_rate": 6.210656687267835e-07,
+      "loss": 0.6577,
+      "step": 13194
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0733821943164632,
+      "learning_rate": 6.2056545251997e-07,
+      "loss": 0.6884,
+      "step": 13195
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6070375909591441,
+      "learning_rate": 6.200654245072146e-07,
+      "loss": 0.7376,
+      "step": 13196
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5476308931857643,
+      "learning_rate": 6.195655847100008e-07,
+      "loss": 0.6957,
+      "step": 13197
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8660136511192706,
+      "learning_rate": 6.1906593314981e-07,
+      "loss": 0.7696,
+      "step": 13198
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.383080313883777,
+      "learning_rate": 6.185664698481137e-07,
+      "loss": 0.6085,
+      "step": 13199
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5962624466804345,
+      "learning_rate": 6.180671948263739e-07,
+      "loss": 0.6909,
+      "step": 13200
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8738299807571985,
+      "learning_rate": 6.175681081060453e-07,
+      "loss": 0.7244,
+      "step": 13201
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9401769815939305,
+      "learning_rate": 6.170692097085751e-07,
+      "loss": 0.7584,
+      "step": 13202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9499856054435545,
+      "learning_rate": 6.165704996554039e-07,
+      "loss": 0.7152,
+      "step": 13203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4221369134587132,
+      "learning_rate": 6.160719779679597e-07,
+      "loss": 0.5717,
+      "step": 13204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8513865884292802,
+      "learning_rate": 6.155736446676652e-07,
+      "loss": 0.7416,
+      "step": 13205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1974034750114377,
+      "learning_rate": 6.150754997759367e-07,
+      "loss": 0.7522,
+      "step": 13206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6478613232687824,
+      "learning_rate": 6.145775433141798e-07,
+      "loss": 0.6339,
+      "step": 13207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.657570826316556,
+      "learning_rate": 6.140797753037925e-07,
+      "loss": 0.6305,
+      "step": 13208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9223503816288723,
+      "learning_rate": 6.135821957661658e-07,
+      "loss": 0.7804,
+      "step": 13209
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4844310552523958,
+      "learning_rate": 6.130848047226806e-07,
+      "loss": 0.6198,
+      "step": 13210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8190981539230753,
+      "learning_rate": 6.125876021947119e-07,
+      "loss": 0.6228,
+      "step": 13211
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0505683747388717,
+      "learning_rate": 6.120905882036244e-07,
+      "loss": 0.5708,
+      "step": 13212
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9669603359404144,
+      "learning_rate": 6.115937627707775e-07,
+      "loss": 0.7178,
+      "step": 13213
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4732981170727093,
+      "learning_rate": 6.110971259175208e-07,
+      "loss": 0.5616,
+      "step": 13214
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7809089606001316,
+      "learning_rate": 6.106006776651934e-07,
+      "loss": 0.6865,
+      "step": 13215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.004981451892176,
+      "learning_rate": 6.101044180351318e-07,
+      "loss": 0.5912,
+      "step": 13216
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2010425764019492,
+      "learning_rate": 6.096083470486602e-07,
+      "loss": 0.588,
+      "step": 13217
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9283248123349082,
+      "learning_rate": 6.091124647270958e-07,
+      "loss": 0.7708,
+      "step": 13218
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6032013621014671,
+      "learning_rate": 6.086167710917479e-07,
+      "loss": 0.6026,
+      "step": 13219
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5861321738048362,
+      "learning_rate": 6.081212661639163e-07,
+      "loss": 0.6845,
+      "step": 13220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.296615836444662,
+      "learning_rate": 6.076259499648968e-07,
+      "loss": 0.6932,
+      "step": 13221
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6532576976626057,
+      "learning_rate": 6.071308225159712e-07,
+      "loss": 0.5744,
+      "step": 13222
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8100679377904119,
+      "learning_rate": 6.066358838384184e-07,
+      "loss": 0.6674,
+      "step": 13223
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9844761572509184,
+      "learning_rate": 6.061411339535062e-07,
+      "loss": 0.6694,
+      "step": 13224
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8776383643216337,
+      "learning_rate": 6.056465728824951e-07,
+      "loss": 0.6586,
+      "step": 13225
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9987745885757866,
+      "learning_rate": 6.051522006466376e-07,
+      "loss": 0.7324,
+      "step": 13226
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5802295545840672,
+      "learning_rate": 6.046580172671773e-07,
+      "loss": 0.6447,
+      "step": 13227
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8217983534061712,
+      "learning_rate": 6.041640227653517e-07,
+      "loss": 0.7462,
+      "step": 13228
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.580977013757972,
+      "learning_rate": 6.036702171623876e-07,
+      "loss": 0.7858,
+      "step": 13229
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8210349008168116,
+      "learning_rate": 6.031766004795047e-07,
+      "loss": 0.6515,
+      "step": 13230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7270391095247346,
+      "learning_rate": 6.026831727379168e-07,
+      "loss": 0.7564,
+      "step": 13231
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7337453889839032,
+      "learning_rate": 6.021899339588266e-07,
+      "loss": 0.7845,
+      "step": 13232
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.583250753226271,
+      "learning_rate": 6.016968841634279e-07,
+      "loss": 0.5293,
+      "step": 13233
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8100910710589955,
+      "learning_rate": 6.012040233729105e-07,
+      "loss": 0.6651,
+      "step": 13234
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6300960230302908,
+      "learning_rate": 6.007113516084528e-07,
+      "loss": 0.7277,
+      "step": 13235
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9179244146227228,
+      "learning_rate": 6.002188688912264e-07,
+      "loss": 0.7076,
+      "step": 13236
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7936441526628482,
+      "learning_rate": 5.997265752423936e-07,
+      "loss": 0.7831,
+      "step": 13237
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.124047374501943,
+      "learning_rate": 5.992344706831088e-07,
+      "loss": 0.5659,
+      "step": 13238
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4608593655787026,
+      "learning_rate": 5.987425552345222e-07,
+      "loss": 0.7434,
+      "step": 13239
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9007363140909581,
+      "learning_rate": 5.98250828917768e-07,
+      "loss": 0.8719,
+      "step": 13240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.540654956765076,
+      "learning_rate": 5.977592917539799e-07,
+      "loss": 0.6781,
+      "step": 13241
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4194262134226707,
+      "learning_rate": 5.972679437642793e-07,
+      "loss": 0.8111,
+      "step": 13242
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.470070687702503,
+      "learning_rate": 5.967767849697803e-07,
+      "loss": 0.7742,
+      "step": 13243
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1513381089318035,
+      "learning_rate": 5.962858153915896e-07,
+      "loss": 0.6668,
+      "step": 13244
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.261059829991794,
+      "learning_rate": 5.957950350508052e-07,
+      "loss": 0.6864,
+      "step": 13245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7436997928593345,
+      "learning_rate": 5.953044439685157e-07,
+      "loss": 0.6269,
+      "step": 13246
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.771497005676223,
+      "learning_rate": 5.948140421658061e-07,
+      "loss": 0.6993,
+      "step": 13247
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1482309313826757,
+      "learning_rate": 5.943238296637455e-07,
+      "loss": 0.7214,
+      "step": 13248
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6380147380511725,
+      "learning_rate": 5.938338064834037e-07,
+      "loss": 0.7527,
+      "step": 13249
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7218963667659215,
+      "learning_rate": 5.933439726458357e-07,
+      "loss": 0.6246,
+      "step": 13250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3777333702884151,
+      "learning_rate": 5.928543281720917e-07,
+      "loss": 0.6196,
+      "step": 13251
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7058510548779402,
+      "learning_rate": 5.923648730832121e-07,
+      "loss": 0.5661,
+      "step": 13252
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7997075637398996,
+      "learning_rate": 5.918756074002291e-07,
+      "loss": 0.6524,
+      "step": 13253
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6673470830515946,
+      "learning_rate": 5.913865311441714e-07,
+      "loss": 0.7481,
+      "step": 13254
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5970058593900232,
+      "learning_rate": 5.908976443360515e-07,
+      "loss": 0.7454,
+      "step": 13255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5964536557549094,
+      "learning_rate": 5.904089469968782e-07,
+      "loss": 0.7,
+      "step": 13256
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.791552272402564,
+      "learning_rate": 5.899204391476554e-07,
+      "loss": 0.6345,
+      "step": 13257
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9926483764657776,
+      "learning_rate": 5.894321208093712e-07,
+      "loss": 0.7878,
+      "step": 13258
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1432514269564105,
+      "learning_rate": 5.889439920030127e-07,
+      "loss": 0.7067,
+      "step": 13259
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6612085925520368,
+      "learning_rate": 5.884560527495542e-07,
+      "loss": 0.7071,
+      "step": 13260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0482248502494724,
+      "learning_rate": 5.879683030699645e-07,
+      "loss": 0.76,
+      "step": 13261
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2196259809276944,
+      "learning_rate": 5.874807429852031e-07,
+      "loss": 0.8015,
+      "step": 13262
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6047487767928261,
+      "learning_rate": 5.869933725162203e-07,
+      "loss": 0.853,
+      "step": 13263
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9739646784902438,
+      "learning_rate": 5.865061916839615e-07,
+      "loss": 0.904,
+      "step": 13264
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.773381832587706,
+      "learning_rate": 5.860192005093624e-07,
+      "loss": 0.678,
+      "step": 13265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7022061987779618,
+      "learning_rate": 5.855323990133465e-07,
+      "loss": 0.6949,
+      "step": 13266
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.738720335535122,
+      "learning_rate": 5.850457872168358e-07,
+      "loss": 0.6389,
+      "step": 13267
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6320650129954783,
+      "learning_rate": 5.845593651407405e-07,
+      "loss": 0.5666,
+      "step": 13268
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7186556766863117,
+      "learning_rate": 5.840731328059629e-07,
+      "loss": 0.6465,
+      "step": 13269
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.611695810835993,
+      "learning_rate": 5.835870902333974e-07,
+      "loss": 0.6594,
+      "step": 13270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5011944270759847,
+      "learning_rate": 5.831012374439293e-07,
+      "loss": 0.6141,
+      "step": 13271
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5207890083197169,
+      "learning_rate": 5.826155744584405e-07,
+      "loss": 0.608,
+      "step": 13272
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6923365843774365,
+      "learning_rate": 5.821301012977965e-07,
+      "loss": 0.6855,
+      "step": 13273
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0338321093434817,
+      "learning_rate": 5.816448179828616e-07,
+      "loss": 0.7102,
+      "step": 13274
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7755970817504063,
+      "learning_rate": 5.811597245344908e-07,
+      "loss": 0.7768,
+      "step": 13275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.13446525585062,
+      "learning_rate": 5.806748209735252e-07,
+      "loss": 0.7427,
+      "step": 13276
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5839187736591551,
+      "learning_rate": 5.801901073208066e-07,
+      "loss": 0.6583,
+      "step": 13277
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5903455610101056,
+      "learning_rate": 5.797055835971621e-07,
+      "loss": 0.6356,
+      "step": 13278
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6795570572037077,
+      "learning_rate": 5.792212498234134e-07,
+      "loss": 0.6693,
+      "step": 13279
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6521193932789975,
+      "learning_rate": 5.787371060203728e-07,
+      "loss": 0.7028,
+      "step": 13280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8116477801743236,
+      "learning_rate": 5.782531522088452e-07,
+      "loss": 0.8042,
+      "step": 13281
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.121529600826972,
+      "learning_rate": 5.777693884096275e-07,
+      "loss": 0.6289,
+      "step": 13282
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7839733739776171,
+      "learning_rate": 5.772858146435095e-07,
+      "loss": 0.8087,
+      "step": 13283
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.849152512917098,
+      "learning_rate": 5.768024309312681e-07,
+      "loss": 0.6564,
+      "step": 13284
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6268580355505589,
+      "learning_rate": 5.763192372936782e-07,
+      "loss": 0.7455,
+      "step": 13285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6713078858106523,
+      "learning_rate": 5.758362337515028e-07,
+      "loss": 0.67,
+      "step": 13286
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.623544295790033,
+      "learning_rate": 5.753534203254973e-07,
+      "loss": 0.7898,
+      "step": 13287
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6491541556246798,
+      "learning_rate": 5.748707970364092e-07,
+      "loss": 0.7145,
+      "step": 13288
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6858715554991222,
+      "learning_rate": 5.74388363904978e-07,
+      "loss": 0.6512,
+      "step": 13289
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.96206162935692,
+      "learning_rate": 5.73906120951937e-07,
+      "loss": 0.7736,
+      "step": 13290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4813791083885206,
+      "learning_rate": 5.734240681980052e-07,
+      "loss": 0.6626,
+      "step": 13291
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6221978350252368,
+      "learning_rate": 5.729422056639006e-07,
+      "loss": 0.6891,
+      "step": 13292
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.892384595343566,
+      "learning_rate": 5.724605333703303e-07,
+      "loss": 0.82,
+      "step": 13293
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.113752537186627,
+      "learning_rate": 5.719790513379891e-07,
+      "loss": 0.6354,
+      "step": 13294
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6374237285226416,
+      "learning_rate": 5.714977595875715e-07,
+      "loss": 0.7754,
+      "step": 13295
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.657995043714286,
+      "learning_rate": 5.710166581397575e-07,
+      "loss": 0.6701,
+      "step": 13296
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.016805055950283,
+      "learning_rate": 5.705357470152218e-07,
+      "loss": 0.7141,
+      "step": 13297
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4079371271963508,
+      "learning_rate": 5.700550262346294e-07,
+      "loss": 0.6104,
+      "step": 13298
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6939854054504115,
+      "learning_rate": 5.695744958186383e-07,
+      "loss": 0.7655,
+      "step": 13299
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8554929286907966,
+      "learning_rate": 5.690941557878988e-07,
+      "loss": 0.7324,
+      "step": 13300
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.488845700848539,
+      "learning_rate": 5.686140061630519e-07,
+      "loss": 0.7447,
+      "step": 13301
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7384116621417822,
+      "learning_rate": 5.6813404696473e-07,
+      "loss": 0.7274,
+      "step": 13302
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.727157687397742,
+      "learning_rate": 5.676542782135585e-07,
+      "loss": 0.6935,
+      "step": 13303
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8568126854453402,
+      "learning_rate": 5.671746999301542e-07,
+      "loss": 0.7213,
+      "step": 13304
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6074275513247627,
+      "learning_rate": 5.666953121351254e-07,
+      "loss": 0.6213,
+      "step": 13305
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6767095402975383,
+      "learning_rate": 5.662161148490725e-07,
+      "loss": 0.751,
+      "step": 13306
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0299376396064455,
+      "learning_rate": 5.657371080925866e-07,
+      "loss": 0.6812,
+      "step": 13307
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.173710668379173,
+      "learning_rate": 5.652582918862548e-07,
+      "loss": 0.7577,
+      "step": 13308
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.772340653607637,
+      "learning_rate": 5.647796662506493e-07,
+      "loss": 0.6087,
+      "step": 13309
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7459650664546666,
+      "learning_rate": 5.6430123120634e-07,
+      "loss": 0.6696,
+      "step": 13310
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7793503738326335,
+      "learning_rate": 5.638229867738854e-07,
+      "loss": 0.7576,
+      "step": 13311
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.243744078485943,
+      "learning_rate": 5.633449329738372e-07,
+      "loss": 0.6784,
+      "step": 13312
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6978294376503478,
+      "learning_rate": 5.628670698267386e-07,
+      "loss": 0.7938,
+      "step": 13313
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.549878188236692,
+      "learning_rate": 5.623893973531225e-07,
+      "loss": 0.6526,
+      "step": 13314
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7842981521708643,
+      "learning_rate": 5.619119155735192e-07,
+      "loss": 0.7618,
+      "step": 13315
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8698054628288596,
+      "learning_rate": 5.61434624508444e-07,
+      "loss": 0.7267,
+      "step": 13316
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9592533965936021,
+      "learning_rate": 5.609575241784071e-07,
+      "loss": 0.6039,
+      "step": 13317
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.609747514870844,
+      "learning_rate": 5.604806146039126e-07,
+      "loss": 0.7223,
+      "step": 13318
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.858005689731561,
+      "learning_rate": 5.600038958054538e-07,
+      "loss": 0.719,
+      "step": 13319
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6199581882028697,
+      "learning_rate": 5.595273678035156e-07,
+      "loss": 0.7714,
+      "step": 13320
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6322627865932218,
+      "learning_rate": 5.590510306185765e-07,
+      "loss": 0.6894,
+      "step": 13321
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5508644835706906,
+      "learning_rate": 5.585748842711047e-07,
+      "loss": 0.7935,
+      "step": 13322
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7003623056200285,
+      "learning_rate": 5.580989287815619e-07,
+      "loss": 0.795,
+      "step": 13323
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.864005986271867,
+      "learning_rate": 5.576231641703994e-07,
+      "loss": 0.7629,
+      "step": 13324
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7250765601169893,
+      "learning_rate": 5.571475904580648e-07,
+      "loss": 0.7184,
+      "step": 13325
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0923611576841918,
+      "learning_rate": 5.566722076649933e-07,
+      "loss": 0.6317,
+      "step": 13326
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7763239379509042,
+      "learning_rate": 5.561970158116114e-07,
+      "loss": 0.7804,
+      "step": 13327
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1461088747471955,
+      "learning_rate": 5.557220149183412e-07,
+      "loss": 0.635,
+      "step": 13328
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5697753582806857,
+      "learning_rate": 5.552472050055946e-07,
+      "loss": 0.6434,
+      "step": 13329
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6392781722369765,
+      "learning_rate": 5.547725860937741e-07,
+      "loss": 0.6636,
+      "step": 13330
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.849705992218724,
+      "learning_rate": 5.542981582032759e-07,
+      "loss": 0.7772,
+      "step": 13331
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.557544799540494,
+      "learning_rate": 5.538239213544855e-07,
+      "loss": 0.7206,
+      "step": 13332
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0604968328884095,
+      "learning_rate": 5.533498755677857e-07,
+      "loss": 0.836,
+      "step": 13333
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0639775423165343,
+      "learning_rate": 5.528760208635436e-07,
+      "loss": 0.6962,
+      "step": 13334
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7085717484088394,
+      "learning_rate": 5.524023572621229e-07,
+      "loss": 0.7216,
+      "step": 13335
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7422623736803955,
+      "learning_rate": 5.519288847838783e-07,
+      "loss": 0.7184,
+      "step": 13336
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.827946765987578,
+      "learning_rate": 5.514556034491569e-07,
+      "loss": 0.7685,
+      "step": 13337
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5888344087004744,
+      "learning_rate": 5.509825132782948e-07,
+      "loss": 0.6951,
+      "step": 13338
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.78615620385793,
+      "learning_rate": 5.505096142916233e-07,
+      "loss": 0.7603,
+      "step": 13339
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2706720495648862,
+      "learning_rate": 5.50036906509463e-07,
+      "loss": 0.8216,
+      "step": 13340
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6786096348613921,
+      "learning_rate": 5.495643899521269e-07,
+      "loss": 0.7247,
+      "step": 13341
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5981658588722494,
+      "learning_rate": 5.4909206463992e-07,
+      "loss": 0.7265,
+      "step": 13342
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6763834755962272,
+      "learning_rate": 5.486199305931411e-07,
+      "loss": 0.6832,
+      "step": 13343
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7879785900224539,
+      "learning_rate": 5.481479878320784e-07,
+      "loss": 0.7898,
+      "step": 13344
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6322833517928996,
+      "learning_rate": 5.476762363770088e-07,
+      "loss": 0.6509,
+      "step": 13345
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7095024312864961,
+      "learning_rate": 5.472046762482092e-07,
+      "loss": 0.5893,
+      "step": 13346
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5575792191173512,
+      "learning_rate": 5.467333074659409e-07,
+      "loss": 0.6447,
+      "step": 13347
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0326838161133534,
+      "learning_rate": 5.462621300504606e-07,
+      "loss": 0.7183,
+      "step": 13348
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.719996358220138,
+      "learning_rate": 5.457911440220154e-07,
+      "loss": 0.6887,
+      "step": 13349
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8207722629185858,
+      "learning_rate": 5.453203494008435e-07,
+      "loss": 0.7426,
+      "step": 13350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5209838382754242,
+      "learning_rate": 5.4484974620718e-07,
+      "loss": 0.7097,
+      "step": 13351
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.518963493675451,
+      "learning_rate": 5.443793344612425e-07,
+      "loss": 0.6665,
+      "step": 13352
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7014157318640062,
+      "learning_rate": 5.439091141832498e-07,
+      "loss": 0.7888,
+      "step": 13353
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7741470319402504,
+      "learning_rate": 5.434390853934063e-07,
+      "loss": 0.7609,
+      "step": 13354
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.621192814086676,
+      "learning_rate": 5.429692481119103e-07,
+      "loss": 0.7938,
+      "step": 13355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5752500397439777,
+      "learning_rate": 5.424996023589524e-07,
+      "loss": 0.7764,
+      "step": 13356
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.857832600535346,
+      "learning_rate": 5.420301481547141e-07,
+      "loss": 0.7485,
+      "step": 13357
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7740561953614762,
+      "learning_rate": 5.415608855193688e-07,
+      "loss": 0.7352,
+      "step": 13358
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5867538855324828,
+      "learning_rate": 5.410918144730815e-07,
+      "loss": 0.726,
+      "step": 13359
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.872524924707328,
+      "learning_rate": 5.406229350360087e-07,
+      "loss": 0.6471,
+      "step": 13360
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4380897758600988,
+      "learning_rate": 5.401542472283006e-07,
+      "loss": 0.6423,
+      "step": 13361
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.075371944198962,
+      "learning_rate": 5.39685751070097e-07,
+      "loss": 0.6528,
+      "step": 13362
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6562275352523392,
+      "learning_rate": 5.392174465815308e-07,
+      "loss": 0.6608,
+      "step": 13363
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7334952663255256,
+      "learning_rate": 5.387493337827254e-07,
+      "loss": 0.8285,
+      "step": 13364
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7594957168137415,
+      "learning_rate": 5.382814126937969e-07,
+      "loss": 0.6012,
+      "step": 13365
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7795743467898402,
+      "learning_rate": 5.378136833348529e-07,
+      "loss": 0.7464,
+      "step": 13366
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2113162745659614,
+      "learning_rate": 5.373461457259926e-07,
+      "loss": 0.6105,
+      "step": 13367
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5511086081218275,
+      "learning_rate": 5.368787998873065e-07,
+      "loss": 0.6415,
+      "step": 13368
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7198140930306882,
+      "learning_rate": 5.364116458388802e-07,
+      "loss": 0.7031,
+      "step": 13369
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6834059385707345,
+      "learning_rate": 5.359446836007842e-07,
+      "loss": 0.7577,
+      "step": 13370
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6859866740390719,
+      "learning_rate": 5.354779131930882e-07,
+      "loss": 0.719,
+      "step": 13371
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.558795134786914,
+      "learning_rate": 5.350113346358493e-07,
+      "loss": 0.622,
+      "step": 13372
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.621520179970155,
+      "learning_rate": 5.345449479491177e-07,
+      "loss": 0.7599,
+      "step": 13373
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8112734245268198,
+      "learning_rate": 5.340787531529346e-07,
+      "loss": 0.7058,
+      "step": 13374
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.714260264962985,
+      "learning_rate": 5.336127502673328e-07,
+      "loss": 0.699,
+      "step": 13375
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.690548474267219,
+      "learning_rate": 5.331469393123401e-07,
+      "loss": 0.6334,
+      "step": 13376
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8605633498900698,
+      "learning_rate": 5.326813203079706e-07,
+      "loss": 0.8327,
+      "step": 13377
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.814973913107471,
+      "learning_rate": 5.322158932742328e-07,
+      "loss": 0.6975,
+      "step": 13378
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7148489824650133,
+      "learning_rate": 5.3175065823113e-07,
+      "loss": 0.7047,
+      "step": 13379
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6003520171356653,
+      "learning_rate": 5.312856151986523e-07,
+      "loss": 0.6618,
+      "step": 13380
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5711810587841333,
+      "learning_rate": 5.308207641967839e-07,
+      "loss": 0.6569,
+      "step": 13381
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9718896618693105,
+      "learning_rate": 5.303561052455009e-07,
+      "loss": 0.7388,
+      "step": 13382
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.70640400123013,
+      "learning_rate": 5.298916383647695e-07,
+      "loss": 0.6308,
+      "step": 13383
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7391283904165198,
+      "learning_rate": 5.294273635745517e-07,
+      "loss": 0.6968,
+      "step": 13384
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6102115812829214,
+      "learning_rate": 5.289632808947959e-07,
+      "loss": 0.6764,
+      "step": 13385
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6697552254313064,
+      "learning_rate": 5.284993903454444e-07,
+      "loss": 0.6665,
+      "step": 13386
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.574236027401285,
+      "learning_rate": 5.280356919464341e-07,
+      "loss": 0.7493,
+      "step": 13387
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6753437372505666,
+      "learning_rate": 5.275721857176885e-07,
+      "loss": 0.7603,
+      "step": 13388
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6514819032396046,
+      "learning_rate": 5.271088716791273e-07,
+      "loss": 0.7091,
+      "step": 13389
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7593289735516449,
+      "learning_rate": 5.266457498506594e-07,
+      "loss": 0.6906,
+      "step": 13390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0193980560109765,
+      "learning_rate": 5.261828202521868e-07,
+      "loss": 0.6845,
+      "step": 13391
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8141015331548405,
+      "learning_rate": 5.257200829036018e-07,
+      "loss": 0.7675,
+      "step": 13392
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6849454107734534,
+      "learning_rate": 5.252575378247887e-07,
+      "loss": 0.6679,
+      "step": 13393
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.730883227516147,
+      "learning_rate": 5.24795185035627e-07,
+      "loss": 0.6405,
+      "step": 13394
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5238771230841124,
+      "learning_rate": 5.243330245559819e-07,
+      "loss": 0.7393,
+      "step": 13395
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9356804362642768,
+      "learning_rate": 5.238710564057137e-07,
+      "loss": 0.8045,
+      "step": 13396
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.181663130756189,
+      "learning_rate": 5.234092806046759e-07,
+      "loss": 0.5861,
+      "step": 13397
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.521312305316203,
+      "learning_rate": 5.229476971727115e-07,
+      "loss": 0.8253,
+      "step": 13398
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7159545815509305,
+      "learning_rate": 5.224863061296553e-07,
+      "loss": 0.6415,
+      "step": 13399
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9396715428476081,
+      "learning_rate": 5.220251074953348e-07,
+      "loss": 0.7342,
+      "step": 13400
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8015659108527453,
+      "learning_rate": 5.215641012895667e-07,
+      "loss": 0.6979,
+      "step": 13401
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6099814302208215,
+      "learning_rate": 5.211032875321659e-07,
+      "loss": 0.6848,
+      "step": 13402
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.64989212198108,
+      "learning_rate": 5.206426662429298e-07,
+      "loss": 0.7904,
+      "step": 13403
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8106937000575294,
+      "learning_rate": 5.201822374416549e-07,
+      "loss": 0.6984,
+      "step": 13404
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3133281611023784,
+      "learning_rate": 5.197220011481274e-07,
+      "loss": 0.6801,
+      "step": 13405
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7439939391410106,
+      "learning_rate": 5.192619573821223e-07,
+      "loss": 0.6691,
+      "step": 13406
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7225723659337595,
+      "learning_rate": 5.188021061634102e-07,
+      "loss": 0.8359,
+      "step": 13407
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4768288182625693,
+      "learning_rate": 5.183424475117521e-07,
+      "loss": 0.6096,
+      "step": 13408
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.175071514073422,
+      "learning_rate": 5.178829814469006e-07,
+      "loss": 0.6874,
+      "step": 13409
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7901769526698519,
+      "learning_rate": 5.174237079885991e-07,
+      "loss": 0.7717,
+      "step": 13410
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7542264244921633,
+      "learning_rate": 5.169646271565831e-07,
+      "loss": 0.6621,
+      "step": 13411
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9323815820394403,
+      "learning_rate": 5.165057389705835e-07,
+      "loss": 0.7089,
+      "step": 13412
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8928564601174487,
+      "learning_rate": 5.160470434503156e-07,
+      "loss": 0.7236,
+      "step": 13413
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6711776110804666,
+      "learning_rate": 5.155885406154937e-07,
+      "loss": 0.7688,
+      "step": 13414
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7290006553778117,
+      "learning_rate": 5.151302304858192e-07,
+      "loss": 0.6471,
+      "step": 13415
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7687610842342223,
+      "learning_rate": 5.146721130809873e-07,
+      "loss": 0.752,
+      "step": 13416
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.689973339178207,
+      "learning_rate": 5.142141884206842e-07,
+      "loss": 0.7305,
+      "step": 13417
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0514944235260506,
+      "learning_rate": 5.137564565245873e-07,
+      "loss": 0.6707,
+      "step": 13418
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6248816937958892,
+      "learning_rate": 5.132989174123659e-07,
+      "loss": 0.6759,
+      "step": 13419
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.785422296628784,
+      "learning_rate": 5.128415711036849e-07,
+      "loss": 0.8129,
+      "step": 13420
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5460445464072843,
+      "learning_rate": 5.12384417618193e-07,
+      "loss": 0.6696,
+      "step": 13421
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.949178999532301,
+      "learning_rate": 5.119274569755378e-07,
+      "loss": 0.8406,
+      "step": 13422
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7109195461022852,
+      "learning_rate": 5.114706891953552e-07,
+      "loss": 0.769,
+      "step": 13423
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.239715879611615,
+      "learning_rate": 5.110141142972735e-07,
+      "loss": 0.8146,
+      "step": 13424
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4942177792377158,
+      "learning_rate": 5.105577323009131e-07,
+      "loss": 0.6175,
+      "step": 13425
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1631141848349578,
+      "learning_rate": 5.101015432258843e-07,
+      "loss": 0.567,
+      "step": 13426
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1587840784015357,
+      "learning_rate": 5.096455470917943e-07,
+      "loss": 0.6913,
+      "step": 13427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5225367258288791,
+      "learning_rate": 5.091897439182342e-07,
+      "loss": 0.6657,
+      "step": 13428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7441424505011716,
+      "learning_rate": 5.087341337247914e-07,
+      "loss": 0.7355,
+      "step": 13429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.511336183181765,
+      "learning_rate": 5.082787165310471e-07,
+      "loss": 0.6148,
+      "step": 13430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0593273203448876,
+      "learning_rate": 5.078234923565684e-07,
+      "loss": 0.5819,
+      "step": 13431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7667470875108824,
+      "learning_rate": 5.073684612209201e-07,
+      "loss": 0.792,
+      "step": 13432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7100299205776575,
+      "learning_rate": 5.069136231436539e-07,
+      "loss": 0.7044,
+      "step": 13433
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6520603292906428,
+      "learning_rate": 5.064589781443163e-07,
+      "loss": 0.7523,
+      "step": 13434
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7670851179251197,
+      "learning_rate": 5.060045262424446e-07,
+      "loss": 0.7201,
+      "step": 13435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7369003911215974,
+      "learning_rate": 5.055502674575663e-07,
+      "loss": 0.6742,
+      "step": 13436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0736531819455442,
+      "learning_rate": 5.05096201809202e-07,
+      "loss": 0.6918,
+      "step": 13437
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7592116642145104,
+      "learning_rate": 5.046423293168667e-07,
+      "loss": 0.7293,
+      "step": 13438
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9249437780908953,
+      "learning_rate": 5.041886500000603e-07,
+      "loss": 0.645,
+      "step": 13439
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0618774686336256,
+      "learning_rate": 5.037351638782812e-07,
+      "loss": 0.6788,
+      "step": 13440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5886247155355613,
+      "learning_rate": 5.032818709710152e-07,
+      "loss": 0.678,
+      "step": 13441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6768902358572564,
+      "learning_rate": 5.028287712977425e-07,
+      "loss": 0.6717,
+      "step": 13442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5710276993197958,
+      "learning_rate": 5.023758648779336e-07,
+      "loss": 0.657,
+      "step": 13443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6771460866417478,
+      "learning_rate": 5.019231517310491e-07,
+      "loss": 0.6971,
+      "step": 13444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.057708141926013,
+      "learning_rate": 5.014706318765466e-07,
+      "loss": 0.6254,
+      "step": 13445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0455709157847448,
+      "learning_rate": 5.010183053338685e-07,
+      "loss": 0.7167,
+      "step": 13446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0691669510273547,
+      "learning_rate": 5.00566172122453e-07,
+      "loss": 0.7567,
+      "step": 13447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8533002104693503,
+      "learning_rate": 5.001142322617314e-07,
+      "loss": 0.7066,
+      "step": 13448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6937196509307348,
+      "learning_rate": 4.996624857711219e-07,
+      "loss": 0.7215,
+      "step": 13449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1849859618208565,
+      "learning_rate": 4.992109326700384e-07,
+      "loss": 0.7338,
+      "step": 13450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8078899330774905,
+      "learning_rate": 4.987595729778849e-07,
+      "loss": 0.6267,
+      "step": 13451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7132674260407414,
+      "learning_rate": 4.983084067140575e-07,
+      "loss": 0.8437,
+      "step": 13452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8530002204667047,
+      "learning_rate": 4.978574338979442e-07,
+      "loss": 0.7467,
+      "step": 13453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.716560526141362,
+      "learning_rate": 4.97406654548922e-07,
+      "loss": 0.681,
+      "step": 13454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1275799408645828,
+      "learning_rate": 4.96956068686365e-07,
+      "loss": 0.6314,
+      "step": 13455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.485714872318826,
+      "learning_rate": 4.965056763296356e-07,
+      "loss": 0.7096,
+      "step": 13456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7097132316265895,
+      "learning_rate": 4.960554774980853e-07,
+      "loss": 0.6841,
+      "step": 13457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1653982009728105,
+      "learning_rate": 4.956054722110626e-07,
+      "loss": 0.7664,
+      "step": 13458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7447657239710908,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.7527,
+      "step": 13459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.172010682423612,
+      "learning_rate": 4.94706042347941e-07,
+      "loss": 0.6742,
+      "step": 13460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6743792819801313,
+      "learning_rate": 4.942566178104924e-07,
+      "loss": 0.7608,
+      "step": 13461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1765149229094414,
+      "learning_rate": 4.938073868948707e-07,
+      "loss": 0.7707,
+      "step": 13462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9139348171289234,
+      "learning_rate": 4.933583496203831e-07,
+      "loss": 0.6445,
+      "step": 13463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7074361939731286,
+      "learning_rate": 4.929095060063227e-07,
+      "loss": 0.6937,
+      "step": 13464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.930714470299224,
+      "learning_rate": 4.92460856071979e-07,
+      "loss": 0.7789,
+      "step": 13465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6684407823994358,
+      "learning_rate": 4.920123998366327e-07,
+      "loss": 0.5837,
+      "step": 13466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.037285011081513,
+      "learning_rate": 4.915641373195512e-07,
+      "loss": 0.7286,
+      "step": 13467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.679292856519002,
+      "learning_rate": 4.911160685400008e-07,
+      "loss": 0.7165,
+      "step": 13468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.364750673598308,
+      "learning_rate": 4.906681935172342e-07,
+      "loss": 0.7534,
+      "step": 13469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6806008380417587,
+      "learning_rate": 4.902205122704983e-07,
+      "loss": 0.5303,
+      "step": 13470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7795772242843795,
+      "learning_rate": 4.897730248190313e-07,
+      "loss": 0.5943,
+      "step": 13471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5220831865757118,
+      "learning_rate": 4.893257311820609e-07,
+      "loss": 0.7366,
+      "step": 13472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.23289556804765,
+      "learning_rate": 4.888786313788108e-07,
+      "loss": 0.7366,
+      "step": 13473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7219811309530804,
+      "learning_rate": 4.88431725428492e-07,
+      "loss": 0.6717,
+      "step": 13474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8236488811550042,
+      "learning_rate": 4.879850133503106e-07,
+      "loss": 0.7019,
+      "step": 13475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6622277027564967,
+      "learning_rate": 4.875384951634616e-07,
+      "loss": 0.7037,
+      "step": 13476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7859335734695523,
+      "learning_rate": 4.870921708871334e-07,
+      "loss": 0.6972,
+      "step": 13477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.394147010594922,
+      "learning_rate": 4.866460405405054e-07,
+      "loss": 0.7237,
+      "step": 13478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6319569349467924,
+      "learning_rate": 4.862001041427488e-07,
+      "loss": 0.696,
+      "step": 13479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6586017610875714,
+      "learning_rate": 4.857543617130256e-07,
+      "loss": 0.7982,
+      "step": 13480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7368725719333455,
+      "learning_rate": 4.85308813270493e-07,
+      "loss": 0.738,
+      "step": 13481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7942605754657817,
+      "learning_rate": 4.848634588342932e-07,
+      "loss": 0.6718,
+      "step": 13482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8628095296795715,
+      "learning_rate": 4.844182984235679e-07,
+      "loss": 0.8014,
+      "step": 13483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6392872306327866,
+      "learning_rate": 4.839733320574457e-07,
+      "loss": 0.7456,
+      "step": 13484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9909362969486886,
+      "learning_rate": 4.835285597550454e-07,
+      "loss": 0.6648,
+      "step": 13485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1682571714556,
+      "learning_rate": 4.830839815354827e-07,
+      "loss": 0.7125,
+      "step": 13486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.822025916314075,
+      "learning_rate": 4.826395974178605e-07,
+      "loss": 0.6694,
+      "step": 13487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.21473497387942,
+      "learning_rate": 4.821954074212759e-07,
+      "loss": 0.5604,
+      "step": 13488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6293283162555363,
+      "learning_rate": 4.817514115648164e-07,
+      "loss": 0.7435,
+      "step": 13489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.039034775037652,
+      "learning_rate": 4.813076098675601e-07,
+      "loss": 0.6111,
+      "step": 13490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6946835288120434,
+      "learning_rate": 4.808640023485805e-07,
+      "loss": 0.7255,
+      "step": 13491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9808147122850481,
+      "learning_rate": 4.804205890269398e-07,
+      "loss": 0.6169,
+      "step": 13492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5741356497635706,
+      "learning_rate": 4.799773699216914e-07,
+      "loss": 0.6608,
+      "step": 13493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.57877609545385,
+      "learning_rate": 4.795343450518825e-07,
+      "loss": 0.8141,
+      "step": 13494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1720926471228474,
+      "learning_rate": 4.790915144365504e-07,
+      "loss": 0.7187,
+      "step": 13495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5330008086622673,
+      "learning_rate": 4.786488780947246e-07,
+      "loss": 0.6982,
+      "step": 13496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9111119708609752,
+      "learning_rate": 4.782064360454258e-07,
+      "loss": 0.6397,
+      "step": 13497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8203171894290264,
+      "learning_rate": 4.777641883076661e-07,
+      "loss": 0.7632,
+      "step": 13498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7142484692230908,
+      "learning_rate": 4.773221349004531e-07,
+      "loss": 0.7379,
+      "step": 13499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1479966733617915,
+      "learning_rate": 4.768802758427782e-07,
+      "loss": 0.6223,
+      "step": 13500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5949413729280044,
+      "learning_rate": 4.7643861115363234e-07,
+      "loss": 0.7136,
+      "step": 13501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8488738294833913,
+      "learning_rate": 4.759971408519937e-07,
+      "loss": 0.7427,
+      "step": 13502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9666569564499359,
+      "learning_rate": 4.755558649568337e-07,
+      "loss": 0.781,
+      "step": 13503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0046710427072703,
+      "learning_rate": 4.7511478348711447e-07,
+      "loss": 0.6343,
+      "step": 13504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9534687737911502,
+      "learning_rate": 4.7467389646178976e-07,
+      "loss": 0.7935,
+      "step": 13505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1094105716172964,
+      "learning_rate": 4.742332038998082e-07,
+      "loss": 0.6341,
+      "step": 13506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1637761071587005,
+      "learning_rate": 4.7379270582010416e-07,
+      "loss": 0.6372,
+      "step": 13507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8791004791387422,
+      "learning_rate": 4.7335240224160695e-07,
+      "loss": 0.6114,
+      "step": 13508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0824108022241523,
+      "learning_rate": 4.729122931832392e-07,
+      "loss": 0.6848,
+      "step": 13509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.067238153865366,
+      "learning_rate": 4.7247237866391236e-07,
+      "loss": 0.7865,
+      "step": 13510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1674092464819066,
+      "learning_rate": 4.7203265870253144e-07,
+      "loss": 0.5895,
+      "step": 13511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7431387341728437,
+      "learning_rate": 4.715931333179913e-07,
+      "loss": 0.6926,
+      "step": 13512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.474273886215764,
+      "learning_rate": 4.7115380252917897e-07,
+      "loss": 0.7264,
+      "step": 13513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.746029263316338,
+      "learning_rate": 4.707146663549744e-07,
+      "loss": 0.7458,
+      "step": 13514
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.764435105457451,
+      "learning_rate": 4.7027572481424753e-07,
+      "loss": 0.6742,
+      "step": 13515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1694510785914554,
+      "learning_rate": 4.6983697792586103e-07,
+      "loss": 0.6927,
+      "step": 13516
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6261621835236344,
+      "learning_rate": 4.6939842570867034e-07,
+      "loss": 0.8001,
+      "step": 13517
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.799426329034035,
+      "learning_rate": 4.6896006818151706e-07,
+      "loss": 0.8638,
+      "step": 13518
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8972733102883406,
+      "learning_rate": 4.685219053632423e-07,
+      "loss": 0.6816,
+      "step": 13519
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6077193115240864,
+      "learning_rate": 4.6808393727267307e-07,
+      "loss": 0.7998,
+      "step": 13520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6410015782317786,
+      "learning_rate": 4.676461639286306e-07,
+      "loss": 0.699,
+      "step": 13521
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7232156473272364,
+      "learning_rate": 4.6720858534992584e-07,
+      "loss": 0.7198,
+      "step": 13522
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5592471698296984,
+      "learning_rate": 4.667712015553627e-07,
+      "loss": 0.6256,
+      "step": 13523
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8124427316487177,
+      "learning_rate": 4.663340125637389e-07,
+      "loss": 0.7575,
+      "step": 13524
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7014935974377081,
+      "learning_rate": 4.658970183938383e-07,
+      "loss": 0.7192,
+      "step": 13525
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.472059564683894,
+      "learning_rate": 4.654602190644408e-07,
+      "loss": 0.799,
+      "step": 13526
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7770537606825503,
+      "learning_rate": 4.650236145943171e-07,
+      "loss": 0.7377,
+      "step": 13527
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4255462962218024,
+      "learning_rate": 4.6458720500222873e-07,
+      "loss": 0.8075,
+      "step": 13528
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.149960109417062,
+      "learning_rate": 4.6415099030692914e-07,
+      "loss": 0.6656,
+      "step": 13529
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7875212609958941,
+      "learning_rate": 4.6371497052716265e-07,
+      "loss": 0.692,
+      "step": 13530
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.041051593244562,
+      "learning_rate": 4.6327914568166763e-07,
+      "loss": 0.786,
+      "step": 13531
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.782875739761756,
+      "learning_rate": 4.628435157891709e-07,
+      "loss": 0.8778,
+      "step": 13532
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6188930887790882,
+      "learning_rate": 4.624080808683923e-07,
+      "loss": 0.6686,
+      "step": 13533
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8118576319380248,
+      "learning_rate": 4.619728409380453e-07,
+      "loss": 0.81,
+      "step": 13534
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2813292318500626,
+      "learning_rate": 4.615377960168321e-07,
+      "loss": 0.724,
+      "step": 13535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5497355113484608,
+      "learning_rate": 4.611029461234473e-07,
+      "loss": 0.5856,
+      "step": 13536
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4798589030613616,
+      "learning_rate": 4.6066829127657807e-07,
+      "loss": 0.7415,
+      "step": 13537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.70216806048153,
+      "learning_rate": 4.6023383149490066e-07,
+      "loss": 0.6528,
+      "step": 13538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6685785651091458,
+      "learning_rate": 4.597995667970878e-07,
+      "loss": 0.8599,
+      "step": 13539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8945912843748385,
+      "learning_rate": 4.5936549720179855e-07,
+      "loss": 0.8255,
+      "step": 13540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.626125740607885,
+      "learning_rate": 4.589316227276852e-07,
+      "loss": 0.6988,
+      "step": 13541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5910717248178512,
+      "learning_rate": 4.584979433933956e-07,
+      "loss": 0.8767,
+      "step": 13542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2144508705270052,
+      "learning_rate": 4.5806445921756146e-07,
+      "loss": 0.756,
+      "step": 13543
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8064693982024298,
+      "learning_rate": 4.5763117021881467e-07,
+      "loss": 0.7232,
+      "step": 13544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4800140466445837,
+      "learning_rate": 4.571980764157724e-07,
+      "loss": 0.6522,
+      "step": 13545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5465942122507605,
+      "learning_rate": 4.567651778270465e-07,
+      "loss": 0.6213,
+      "step": 13546
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8816759884147936,
+      "learning_rate": 4.563324744712389e-07,
+      "loss": 0.7402,
+      "step": 13547
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0138427519139204,
+      "learning_rate": 4.5589996636694387e-07,
+      "loss": 0.723,
+      "step": 13548
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6519538303813617,
+      "learning_rate": 4.5546765353274846e-07,
+      "loss": 0.6716,
+      "step": 13549
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.622243509492607,
+      "learning_rate": 4.5503553598722826e-07,
+      "loss": 0.6872,
+      "step": 13550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8991283557257523,
+      "learning_rate": 4.546036137489529e-07,
+      "loss": 0.7574,
+      "step": 13551
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8422614640263966,
+      "learning_rate": 4.5417188683648417e-07,
+      "loss": 0.7372,
+      "step": 13552
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.454550786543139,
+      "learning_rate": 4.5374035526837334e-07,
+      "loss": 0.6799,
+      "step": 13553
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.720892801003137,
+      "learning_rate": 4.5330901906316506e-07,
+      "loss": 0.6507,
+      "step": 13554
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9998224764664954,
+      "learning_rate": 4.528778782393939e-07,
+      "loss": 0.5499,
+      "step": 13555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7239042317280826,
+      "learning_rate": 4.5244693281558615e-07,
+      "loss": 0.7788,
+      "step": 13556
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8610065986163495,
+      "learning_rate": 4.5201618281026374e-07,
+      "loss": 0.7134,
+      "step": 13557
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7656795704453727,
+      "learning_rate": 4.5158562824193395e-07,
+      "loss": 0.6335,
+      "step": 13558
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4516260475278973,
+      "learning_rate": 4.511552691290988e-07,
+      "loss": 0.6115,
+      "step": 13559
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6772056982131607,
+      "learning_rate": 4.507251054902539e-07,
+      "loss": 0.7124,
+      "step": 13560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8358463415337605,
+      "learning_rate": 4.5029513734388187e-07,
+      "loss": 0.8001,
+      "step": 13561
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9699837297562772,
+      "learning_rate": 4.498653647084611e-07,
+      "loss": 0.7504,
+      "step": 13562
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.675162210277824,
+      "learning_rate": 4.494357876024591e-07,
+      "loss": 0.7309,
+      "step": 13563
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.551182269407261,
+      "learning_rate": 4.490064060443361e-07,
+      "loss": 0.7062,
+      "step": 13564
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9189045231771802,
+      "learning_rate": 4.4857722005254345e-07,
+      "loss": 0.7355,
+      "step": 13565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5559062575321465,
+      "learning_rate": 4.4814822964552363e-07,
+      "loss": 0.7922,
+      "step": 13566
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9134421147919598,
+      "learning_rate": 4.47719434841713e-07,
+      "loss": 0.6095,
+      "step": 13567
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.786293823915481,
+      "learning_rate": 4.4729083565953626e-07,
+      "loss": 0.6867,
+      "step": 13568
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7186806195095379,
+      "learning_rate": 4.468624321174109e-07,
+      "loss": 0.8613,
+      "step": 13569
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.650976614034322,
+      "learning_rate": 4.4643422423374825e-07,
+      "loss": 0.6535,
+      "step": 13570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7336068006555272,
+      "learning_rate": 4.460062120269476e-07,
+      "loss": 0.7561,
+      "step": 13571
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1269173707863493,
+      "learning_rate": 4.4557839551540303e-07,
+      "loss": 0.7971,
+      "step": 13572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6697907139977435,
+      "learning_rate": 4.4515077471749767e-07,
+      "loss": 0.7504,
+      "step": 13573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8093243541780428,
+      "learning_rate": 4.4472334965160736e-07,
+      "loss": 0.6568,
+      "step": 13574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9792419648983987,
+      "learning_rate": 4.442961203361013e-07,
+      "loss": 0.648,
+      "step": 13575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9516557326985458,
+      "learning_rate": 4.438690867893347e-07,
+      "loss": 0.6622,
+      "step": 13576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5548028927724376,
+      "learning_rate": 4.4344224902966194e-07,
+      "loss": 0.5949,
+      "step": 13577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.729239665119086,
+      "learning_rate": 4.430156070754249e-07,
+      "loss": 0.6906,
+      "step": 13578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.584610404993395,
+      "learning_rate": 4.4258916094495394e-07,
+      "loss": 0.6751,
+      "step": 13579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7683063625542585,
+      "learning_rate": 4.421629106565778e-07,
+      "loss": 0.7485,
+      "step": 13580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6600215251112382,
+      "learning_rate": 4.417368562286117e-07,
+      "loss": 0.6484,
+      "step": 13581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8572101735328654,
+      "learning_rate": 4.4131099767936504e-07,
+      "loss": 0.6665,
+      "step": 13582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5763632306141762,
+      "learning_rate": 4.4088533502713803e-07,
+      "loss": 0.665,
+      "step": 13583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4825186885116954,
+      "learning_rate": 4.4045986829022e-07,
+      "loss": 0.6024,
+      "step": 13584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2403461360257833,
+      "learning_rate": 4.400345974868986e-07,
+      "loss": 0.7148,
+      "step": 13585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0822594800822836,
+      "learning_rate": 4.396095226354441e-07,
+      "loss": 0.6169,
+      "step": 13586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7187476998313125,
+      "learning_rate": 4.391846437541258e-07,
+      "loss": 0.6603,
+      "step": 13587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6940518449711495,
+      "learning_rate": 4.3875996086120085e-07,
+      "loss": 0.7534,
+      "step": 13588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9289746403844477,
+      "learning_rate": 4.38335473974919e-07,
+      "loss": 0.7738,
+      "step": 13589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7704252188514433,
+      "learning_rate": 4.379111831135213e-07,
+      "loss": 0.764,
+      "step": 13590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6184713584904633,
+      "learning_rate": 4.374870882952398e-07,
+      "loss": 0.6388,
+      "step": 13591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.195812159214379,
+      "learning_rate": 4.370631895382993e-07,
+      "loss": 0.6905,
+      "step": 13592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.77803731322333,
+      "learning_rate": 4.3663948686091694e-07,
+      "loss": 0.7975,
+      "step": 13593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7665179242692546,
+      "learning_rate": 4.362159802812971e-07,
+      "loss": 0.7449,
+      "step": 13594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9133968761784619,
+      "learning_rate": 4.357926698176418e-07,
+      "loss": 0.7733,
+      "step": 13595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5573270762333882,
+      "learning_rate": 4.3536955548814153e-07,
+      "loss": 0.7526,
+      "step": 13596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7878476049054792,
+      "learning_rate": 4.3494663731097564e-07,
+      "loss": 0.776,
+      "step": 13597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9201083561982333,
+      "learning_rate": 4.3452391530432016e-07,
+      "loss": 0.743,
+      "step": 13598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8766423684656988,
+      "learning_rate": 4.341013894863405e-07,
+      "loss": 0.6736,
+      "step": 13599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0251795920301174,
+      "learning_rate": 4.3367905987519275e-07,
+      "loss": 0.7399,
+      "step": 13600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.996432586642696,
+      "learning_rate": 4.332569264890252e-07,
+      "loss": 0.7537,
+      "step": 13601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7676569629723138,
+      "learning_rate": 4.328349893459777e-07,
+      "loss": 0.7679,
+      "step": 13602
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6147389270206687,
+      "learning_rate": 4.3241324846418355e-07,
+      "loss": 0.7293,
+      "step": 13603
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.049705196664608,
+      "learning_rate": 4.3199170386176325e-07,
+      "loss": 0.6997,
+      "step": 13604
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5940983160058202,
+      "learning_rate": 4.31570355556834e-07,
+      "loss": 0.6259,
+      "step": 13605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7450834172285596,
+      "learning_rate": 4.3114920356750076e-07,
+      "loss": 0.5698,
+      "step": 13606
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8736108694893268,
+      "learning_rate": 4.3072824791186187e-07,
+      "loss": 0.833,
+      "step": 13607
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8359417835824876,
+      "learning_rate": 4.3030748860800606e-07,
+      "loss": 0.6824,
+      "step": 13608
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5148559184206052,
+      "learning_rate": 4.2988692567401515e-07,
+      "loss": 0.6866,
+      "step": 13609
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.859487638371951,
+      "learning_rate": 4.294665591279601e-07,
+      "loss": 0.6979,
+      "step": 13610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9339241961142208,
+      "learning_rate": 4.290463889879082e-07,
+      "loss": 0.8111,
+      "step": 13611
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1315572898641688,
+      "learning_rate": 4.286264152719111e-07,
+      "loss": 0.6572,
+      "step": 13612
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1759400062559473,
+      "learning_rate": 4.2820663799801876e-07,
+      "loss": 0.4708,
+      "step": 13613
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5964422879094087,
+      "learning_rate": 4.2778705718426907e-07,
+      "loss": 0.642,
+      "step": 13614
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.660645019235148,
+      "learning_rate": 4.273676728486925e-07,
+      "loss": 0.6599,
+      "step": 13615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5470669601931506,
+      "learning_rate": 4.2694848500931086e-07,
+      "loss": 0.6032,
+      "step": 13616
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1678244160679545,
+      "learning_rate": 4.2652949368413634e-07,
+      "loss": 0.6206,
+      "step": 13617
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3098409950626482,
+      "learning_rate": 4.2611069889117727e-07,
+      "loss": 0.6556,
+      "step": 13618
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7775950760089296,
+      "learning_rate": 4.2569210064842716e-07,
+      "loss": 0.6608,
+      "step": 13619
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.535614495059108,
+      "learning_rate": 4.252736989738737e-07,
+      "loss": 0.7455,
+      "step": 13620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9395508712307061,
+      "learning_rate": 4.2485549388549927e-07,
+      "loss": 0.7581,
+      "step": 13621
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6690431211798815,
+      "learning_rate": 4.244374854012734e-07,
+      "loss": 0.7787,
+      "step": 13622
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6627051326696427,
+      "learning_rate": 4.2401967353915885e-07,
+      "loss": 0.6761,
+      "step": 13623
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7661319444928991,
+      "learning_rate": 4.236020583171108e-07,
+      "loss": 0.7315,
+      "step": 13624
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9505166691199822,
+      "learning_rate": 4.2318463975307423e-07,
+      "loss": 0.7453,
+      "step": 13625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.798192966492613,
+      "learning_rate": 4.227674178649865e-07,
+      "loss": 0.73,
+      "step": 13626
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8753424793806421,
+      "learning_rate": 4.2235039267077615e-07,
+      "loss": 0.6513,
+      "step": 13627
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6775641284149914,
+      "learning_rate": 4.219335641883654e-07,
+      "loss": 0.6728,
+      "step": 13628
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.629754545891886,
+      "learning_rate": 4.215169324356666e-07,
+      "loss": 0.7001,
+      "step": 13629
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9043924933094982,
+      "learning_rate": 4.2110049743057933e-07,
+      "loss": 0.8058,
+      "step": 13630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1433241042956834,
+      "learning_rate": 4.206842591910032e-07,
+      "loss": 0.6755,
+      "step": 13631
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.817920087260365,
+      "learning_rate": 4.202682177348227e-07,
+      "loss": 0.6656,
+      "step": 13632
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.899680950589142,
+      "learning_rate": 4.198523730799164e-07,
+      "loss": 0.7332,
+      "step": 13633
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0175347681024216,
+      "learning_rate": 4.194367252441545e-07,
+      "loss": 0.7632,
+      "step": 13634
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7998186385447184,
+      "learning_rate": 4.1902127424539695e-07,
+      "loss": 0.6485,
+      "step": 13635
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8740744054569731,
+      "learning_rate": 4.186060201014991e-07,
+      "loss": 0.8504,
+      "step": 13636
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6103501521307844,
+      "learning_rate": 4.181909628303027e-07,
+      "loss": 0.633,
+      "step": 13637
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.74284176645264,
+      "learning_rate": 4.177761024496452e-07,
+      "loss": 0.749,
+      "step": 13638
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2117770022481387,
+      "learning_rate": 4.1736143897735394e-07,
+      "loss": 0.7178,
+      "step": 13639
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7626311195127577,
+      "learning_rate": 4.1694697243124804e-07,
+      "loss": 0.7246,
+      "step": 13640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9504066677709098,
+      "learning_rate": 4.1653270282913715e-07,
+      "loss": 0.6311,
+      "step": 13641
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5423900688838148,
+      "learning_rate": 4.1611863018882424e-07,
+      "loss": 0.7544,
+      "step": 13642
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.802938405590275,
+      "learning_rate": 4.157047545281029e-07,
+      "loss": 0.6365,
+      "step": 13643
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9478640267240546,
+      "learning_rate": 4.152910758647577e-07,
+      "loss": 0.7833,
+      "step": 13644
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6863697080971436,
+      "learning_rate": 4.14877594216565e-07,
+      "loss": 0.6536,
+      "step": 13645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6285999730442626,
+      "learning_rate": 4.14464309601294e-07,
+      "loss": 0.68,
+      "step": 13646
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6585847194710206,
+      "learning_rate": 4.14051222036706e-07,
+      "loss": 0.7249,
+      "step": 13647
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8162828797074098,
+      "learning_rate": 4.1363833154054786e-07,
+      "loss": 0.6767,
+      "step": 13648
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.234164389824897,
+      "learning_rate": 4.1322563813056606e-07,
+      "loss": 0.5276,
+      "step": 13649
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.736816730863239,
+      "learning_rate": 4.1281314182449405e-07,
+      "loss": 0.6689,
+      "step": 13650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8423449211703595,
+      "learning_rate": 4.1240084264005774e-07,
+      "loss": 0.8521,
+      "step": 13651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.76917859111278,
+      "learning_rate": 4.1198874059497405e-07,
+      "loss": 0.729,
+      "step": 13652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7906956706311514,
+      "learning_rate": 4.1157683570695105e-07,
+      "loss": 0.7703,
+      "step": 13653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8420401972105058,
+      "learning_rate": 4.111651279936929e-07,
+      "loss": 0.7385,
+      "step": 13654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7209915140505627,
+      "learning_rate": 4.1075361747288657e-07,
+      "loss": 0.877,
+      "step": 13655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7740992204246007,
+      "learning_rate": 4.103423041622195e-07,
+      "loss": 0.7507,
+      "step": 13656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.855565627067945,
+      "learning_rate": 4.099311880793655e-07,
+      "loss": 0.7741,
+      "step": 13657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5993605566884814,
+      "learning_rate": 4.0952026924199027e-07,
+      "loss": 0.6805,
+      "step": 13658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6147744364556627,
+      "learning_rate": 4.091095476677531e-07,
+      "loss": 0.6372,
+      "step": 13659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7095474013549437,
+      "learning_rate": 4.0869902337430323e-07,
+      "loss": 0.7358,
+      "step": 13660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8332476120271213,
+      "learning_rate": 4.082886963792815e-07,
+      "loss": 0.7796,
+      "step": 13661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9276551154958697,
+      "learning_rate": 4.0787856670032044e-07,
+      "loss": 0.7849,
+      "step": 13662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.690709198413627,
+      "learning_rate": 4.0746863435504434e-07,
+      "loss": 0.7328,
+      "step": 13663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7350246538485625,
+      "learning_rate": 4.070588993610697e-07,
+      "loss": 0.686,
+      "step": 13664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4272666247843928,
+      "learning_rate": 4.066493617360029e-07,
+      "loss": 0.6873,
+      "step": 13665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7195526684897184,
+      "learning_rate": 4.062400214974432e-07,
+      "loss": 0.7441,
+      "step": 13666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5726357855160988,
+      "learning_rate": 4.0583087866298056e-07,
+      "loss": 0.7568,
+      "step": 13667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7753930089447127,
+      "learning_rate": 4.054219332501974e-07,
+      "loss": 0.8038,
+      "step": 13668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9896188827306274,
+      "learning_rate": 4.050131852766659e-07,
+      "loss": 0.7465,
+      "step": 13669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0168358768556995,
+      "learning_rate": 4.046046347599514e-07,
+      "loss": 0.7706,
+      "step": 13670
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.548671188270457,
+      "learning_rate": 4.0419628171760927e-07,
+      "loss": 0.7187,
+      "step": 13671
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7546645604175477,
+      "learning_rate": 4.0378812616719056e-07,
+      "loss": 0.6715,
+      "step": 13672
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4824448006416742,
+      "learning_rate": 4.0338016812623007e-07,
+      "loss": 0.6915,
+      "step": 13673
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7031766689376904,
+      "learning_rate": 4.029724076122621e-07,
+      "loss": 0.7427,
+      "step": 13674
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8001051983371041,
+      "learning_rate": 4.025648446428082e-07,
+      "loss": 0.6608,
+      "step": 13675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6173315988488932,
+      "learning_rate": 4.021574792353816e-07,
+      "loss": 0.6925,
+      "step": 13676
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7904401606395943,
+      "learning_rate": 4.017503114074883e-07,
+      "loss": 0.8044,
+      "step": 13677
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7267132026339453,
+      "learning_rate": 4.0134334117662375e-07,
+      "loss": 0.5618,
+      "step": 13678
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8803670715530885,
+      "learning_rate": 4.009365685602795e-07,
+      "loss": 0.6948,
+      "step": 13679
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7488546906657052,
+      "learning_rate": 4.005299935759327e-07,
+      "loss": 0.7674,
+      "step": 13680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5979020967333402,
+      "learning_rate": 4.001236162410549e-07,
+      "loss": 0.6176,
+      "step": 13681
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0733238140650507,
+      "learning_rate": 3.997174365731105e-07,
+      "loss": 0.6728,
+      "step": 13682
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9074325515434194,
+      "learning_rate": 3.993114545895532e-07,
+      "loss": 0.7419,
+      "step": 13683
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.273218113852976,
+      "learning_rate": 3.989056703078292e-07,
+      "loss": 0.672,
+      "step": 13684
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0558842610834436,
+      "learning_rate": 3.985000837453756e-07,
+      "loss": 0.7165,
+      "step": 13685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.868333469244614,
+      "learning_rate": 3.9809469491962126e-07,
+      "loss": 0.8349,
+      "step": 13686
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5910075044038607,
+      "learning_rate": 3.976895038479872e-07,
+      "loss": 0.7241,
+      "step": 13687
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0508996637343655,
+      "learning_rate": 3.9728451054788396e-07,
+      "loss": 0.7936,
+      "step": 13688
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4972336766332128,
+      "learning_rate": 3.968797150367171e-07,
+      "loss": 0.6675,
+      "step": 13689
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8415763184155596,
+      "learning_rate": 3.9647511733188147e-07,
+      "loss": 0.8001,
+      "step": 13690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9210785749678247,
+      "learning_rate": 3.960707174507611e-07,
+      "loss": 0.7059,
+      "step": 13691
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6256543252778608,
+      "learning_rate": 3.9566651541073586e-07,
+      "loss": 0.7277,
+      "step": 13692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9227917382374309,
+      "learning_rate": 3.9526251122917526e-07,
+      "loss": 0.6968,
+      "step": 13693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.633017843550958,
+      "learning_rate": 3.948587049234398e-07,
+      "loss": 0.7294,
+      "step": 13694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.792373718843596,
+      "learning_rate": 3.944550965108818e-07,
+      "loss": 0.6844,
+      "step": 13695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.784509428558978,
+      "learning_rate": 3.940516860088445e-07,
+      "loss": 0.7557,
+      "step": 13696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5882930003936508,
+      "learning_rate": 3.936484734346663e-07,
+      "loss": 0.7041,
+      "step": 13697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1512070021932983,
+      "learning_rate": 3.932454588056711e-07,
+      "loss": 0.6555,
+      "step": 13698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.72118886166224,
+      "learning_rate": 3.928426421391773e-07,
+      "loss": 0.7128,
+      "step": 13699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8712786499430867,
+      "learning_rate": 3.924400234524972e-07,
+      "loss": 0.7374,
+      "step": 13700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6304379732403422,
+      "learning_rate": 3.9203760276293024e-07,
+      "loss": 0.7426,
+      "step": 13701
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6804275316763222,
+      "learning_rate": 3.9163538008777035e-07,
+      "loss": 0.7801,
+      "step": 13702
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.270320746131573,
+      "learning_rate": 3.9123335544430153e-07,
+      "loss": 0.682,
+      "step": 13703
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6132128878769294,
+      "learning_rate": 3.9083152884979935e-07,
+      "loss": 0.7345,
+      "step": 13704
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6426852107307968,
+      "learning_rate": 3.9042990032153227e-07,
+      "loss": 0.7337,
+      "step": 13705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8489276440445952,
+      "learning_rate": 3.9002846987675704e-07,
+      "loss": 0.6873,
+      "step": 13706
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7546333949721482,
+      "learning_rate": 3.8962723753272645e-07,
+      "loss": 0.7506,
+      "step": 13707
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9914242189730702,
+      "learning_rate": 3.892262033066818e-07,
+      "loss": 0.7238,
+      "step": 13708
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.539470407580911,
+      "learning_rate": 3.8882536721585486e-07,
+      "loss": 0.7426,
+      "step": 13709
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.618903346476756,
+      "learning_rate": 3.884247292774718e-07,
+      "loss": 0.5779,
+      "step": 13710
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0151814992564243,
+      "learning_rate": 3.8802428950874884e-07,
+      "loss": 0.575,
+      "step": 13711
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8299938818355077,
+      "learning_rate": 3.876240479268939e-07,
+      "loss": 0.6241,
+      "step": 13712
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9730711595626957,
+      "learning_rate": 3.872240045491055e-07,
+      "loss": 0.692,
+      "step": 13713
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5754308461429614,
+      "learning_rate": 3.868241593925742e-07,
+      "loss": 0.7399,
+      "step": 13714
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.794722060536748,
+      "learning_rate": 3.8642451247448477e-07,
+      "loss": 0.8607,
+      "step": 13715
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7234320806962535,
+      "learning_rate": 3.8602506381200666e-07,
+      "loss": 0.637,
+      "step": 13716
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6918063089603046,
+      "learning_rate": 3.8562581342230897e-07,
+      "loss": 0.7532,
+      "step": 13717
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0621033677148333,
+      "learning_rate": 3.8522676132254635e-07,
+      "loss": 0.7467,
+      "step": 13718
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6389354421211837,
+      "learning_rate": 3.848279075298678e-07,
+      "loss": 0.6913,
+      "step": 13719
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0568179706577405,
+      "learning_rate": 3.8442925206141237e-07,
+      "loss": 0.6224,
+      "step": 13720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7267903475155315,
+      "learning_rate": 3.840307949343114e-07,
+      "loss": 0.6985,
+      "step": 13721
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5969953278151003,
+      "learning_rate": 3.8363253616568784e-07,
+      "loss": 0.7228,
+      "step": 13722
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7833643168459947,
+      "learning_rate": 3.8323447577265525e-07,
+      "loss": 0.745,
+      "step": 13723
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8079599434821059,
+      "learning_rate": 3.828366137723183e-07,
+      "loss": 0.7528,
+      "step": 13724
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.655060559605762,
+      "learning_rate": 3.82438950181776e-07,
+      "loss": 0.6544,
+      "step": 13725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.943709211831767,
+      "learning_rate": 3.820414850181153e-07,
+      "loss": 0.6605,
+      "step": 13726
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7423862478782763,
+      "learning_rate": 3.8164421829841756e-07,
+      "loss": 0.8008,
+      "step": 13727
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.663374637894627,
+      "learning_rate": 3.81247150039753e-07,
+      "loss": 0.863,
+      "step": 13728
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8937889870431754,
+      "learning_rate": 3.80850280259184e-07,
+      "loss": 0.7276,
+      "step": 13729
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7516216731461556,
+      "learning_rate": 3.8045360897376814e-07,
+      "loss": 0.6559,
+      "step": 13730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7621293113562313,
+      "learning_rate": 3.800571362005473e-07,
+      "loss": 0.6988,
+      "step": 13731
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8292093970477725,
+      "learning_rate": 3.7966086195656e-07,
+      "loss": 0.7671,
+      "step": 13732
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1893963995197536,
+      "learning_rate": 3.7926478625883725e-07,
+      "loss": 0.7667,
+      "step": 13733
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2310297037754492,
+      "learning_rate": 3.7886890912439633e-07,
+      "loss": 0.6434,
+      "step": 13734
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4965513022759558,
+      "learning_rate": 3.784732305702504e-07,
+      "loss": 0.7715,
+      "step": 13735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8217767977577606,
+      "learning_rate": 3.7807775061340257e-07,
+      "loss": 0.7577,
+      "step": 13736
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8212579694234747,
+      "learning_rate": 3.776824692708475e-07,
+      "loss": 0.8547,
+      "step": 13737
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.005315913030209,
+      "learning_rate": 3.772873865595711e-07,
+      "loss": 0.7607,
+      "step": 13738
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8514230532693319,
+      "learning_rate": 3.768925024965503e-07,
+      "loss": 0.6751,
+      "step": 13739
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5367148343596335,
+      "learning_rate": 3.76497817098756e-07,
+      "loss": 0.5827,
+      "step": 13740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4565878880259926,
+      "learning_rate": 3.761033303831474e-07,
+      "loss": 0.6247,
+      "step": 13741
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5556284984887576,
+      "learning_rate": 3.7570904236667536e-07,
+      "loss": 0.805,
+      "step": 13742
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8631295619710186,
+      "learning_rate": 3.7531495306628584e-07,
+      "loss": 0.8319,
+      "step": 13743
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6609680889953768,
+      "learning_rate": 3.749210624989125e-07,
+      "loss": 0.6969,
+      "step": 13744
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4917358366957265,
+      "learning_rate": 3.7452737068148173e-07,
+      "loss": 0.7439,
+      "step": 13745
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6380128307993522,
+      "learning_rate": 3.741338776309111e-07,
+      "loss": 0.6653,
+      "step": 13746
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.458685559967695,
+      "learning_rate": 3.7374058336410945e-07,
+      "loss": 0.5333,
+      "step": 13747
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7671690496245422,
+      "learning_rate": 3.733474878979798e-07,
+      "loss": 0.6777,
+      "step": 13748
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7954085221459888,
+      "learning_rate": 3.729545912494115e-07,
+      "loss": 0.6683,
+      "step": 13749
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5457914563560644,
+      "learning_rate": 3.7256189343528926e-07,
+      "loss": 0.7331,
+      "step": 13750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9533481274360445,
+      "learning_rate": 3.721693944724897e-07,
+      "loss": 0.7108,
+      "step": 13751
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.722492931596894,
+      "learning_rate": 3.717770943778759e-07,
+      "loss": 0.7539,
+      "step": 13752
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.660135604892124,
+      "learning_rate": 3.7138499316830946e-07,
+      "loss": 0.6901,
+      "step": 13753
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.923702404401195,
+      "learning_rate": 3.7099309086063794e-07,
+      "loss": 0.7531,
+      "step": 13754
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6810411810089851,
+      "learning_rate": 3.706013874717024e-07,
+      "loss": 0.7726,
+      "step": 13755
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.788335306141709,
+      "learning_rate": 3.702098830183359e-07,
+      "loss": 0.8317,
+      "step": 13756
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8850233347627405,
+      "learning_rate": 3.698185775173607e-07,
+      "loss": 0.7269,
+      "step": 13757
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9909172051542147,
+      "learning_rate": 3.694274709855944e-07,
+      "loss": 0.5706,
+      "step": 13758
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.982787920291619,
+      "learning_rate": 3.6903656343984293e-07,
+      "loss": 0.654,
+      "step": 13759
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8767372708450503,
+      "learning_rate": 3.6864585489690297e-07,
+      "loss": 0.6882,
+      "step": 13760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7587565748502256,
+      "learning_rate": 3.6825534537356545e-07,
+      "loss": 0.7419,
+      "step": 13761
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6677785524618667,
+      "learning_rate": 3.678650348866114e-07,
+      "loss": 0.5809,
+      "step": 13762
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6245150534559865,
+      "learning_rate": 3.6747492345281297e-07,
+      "loss": 0.6829,
+      "step": 13763
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0596907114268364,
+      "learning_rate": 3.670850110889346e-07,
+      "loss": 0.6723,
+      "step": 13764
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6367296272589782,
+      "learning_rate": 3.6669529781173053e-07,
+      "loss": 0.8019,
+      "step": 13765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.618245487315309,
+      "learning_rate": 3.6630578363794965e-07,
+      "loss": 0.7174,
+      "step": 13766
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6632936387415096,
+      "learning_rate": 3.659164685843275e-07,
+      "loss": 0.8445,
+      "step": 13767
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7578462195067817,
+      "learning_rate": 3.655273526675962e-07,
+      "loss": 0.6954,
+      "step": 13768
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6789495788151636,
+      "learning_rate": 3.651384359044774e-07,
+      "loss": 0.6497,
+      "step": 13769
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0800377488964394,
+      "learning_rate": 3.647497183116799e-07,
+      "loss": 0.6255,
+      "step": 13770
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.018943154294317,
+      "learning_rate": 3.6436119990591214e-07,
+      "loss": 0.6703,
+      "step": 13771
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6794066006571569,
+      "learning_rate": 3.639728807038667e-07,
+      "loss": 0.6875,
+      "step": 13772
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6649483089610533,
+      "learning_rate": 3.6358476072223205e-07,
+      "loss": 0.707,
+      "step": 13773
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6708847139748308,
+      "learning_rate": 3.631968399776864e-07,
+      "loss": 0.5637,
+      "step": 13774
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6203816857220013,
+      "learning_rate": 3.628091184868976e-07,
+      "loss": 0.7084,
+      "step": 13775
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8446676850190922,
+      "learning_rate": 3.6242159626653004e-07,
+      "loss": 0.6405,
+      "step": 13776
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6904505440414674,
+      "learning_rate": 3.6203427333323495e-07,
+      "loss": 0.7468,
+      "step": 13777
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4674945184696302,
+      "learning_rate": 3.616471497036561e-07,
+      "loss": 0.6967,
+      "step": 13778
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6879198701858924,
+      "learning_rate": 3.6126022539442975e-07,
+      "loss": 0.6765,
+      "step": 13779
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.777730255571441,
+      "learning_rate": 3.6087350042218194e-07,
+      "loss": 0.6408,
+      "step": 13780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8556174641511618,
+      "learning_rate": 3.6048697480353225e-07,
+      "loss": 0.7448,
+      "step": 13781
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.00939742945305,
+      "learning_rate": 3.601006485550895e-07,
+      "loss": 0.7449,
+      "step": 13782
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6203690754490085,
+      "learning_rate": 3.597145216934556e-07,
+      "loss": 0.7095,
+      "step": 13783
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0752490360165465,
+      "learning_rate": 3.593285942352237e-07,
+      "loss": 0.6614,
+      "step": 13784
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.756356201883054,
+      "learning_rate": 3.5894286619697627e-07,
+      "loss": 0.6922,
+      "step": 13785
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.534724184934641,
+      "learning_rate": 3.585573375952911e-07,
+      "loss": 0.6405,
+      "step": 13786
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6973712244879,
+      "learning_rate": 3.581720084467344e-07,
+      "loss": 0.7394,
+      "step": 13787
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8481774780852478,
+      "learning_rate": 3.57786878767864e-07,
+      "loss": 0.7103,
+      "step": 13788
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.919100590741738,
+      "learning_rate": 3.5740194857523e-07,
+      "loss": 0.6089,
+      "step": 13789
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.006280837674837,
+      "learning_rate": 3.570172178853731e-07,
+      "loss": 0.7305,
+      "step": 13790
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.15580283767696,
+      "learning_rate": 3.56632686714829e-07,
+      "loss": 0.7725,
+      "step": 13791
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8394270541245807,
+      "learning_rate": 3.5624835508011824e-07,
+      "loss": 0.7995,
+      "step": 13792
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7077508106900061,
+      "learning_rate": 3.5586422299775714e-07,
+      "loss": 0.706,
+      "step": 13793
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4731548231364004,
+      "learning_rate": 3.554802904842547e-07,
+      "loss": 0.6982,
+      "step": 13794
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4893186513011518,
+      "learning_rate": 3.5509655755610773e-07,
+      "loss": 0.6946,
+      "step": 13795
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.705857375151247,
+      "learning_rate": 3.547130242298064e-07,
+      "loss": 0.7059,
+      "step": 13796
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6069485116108901,
+      "learning_rate": 3.5432969052183186e-07,
+      "loss": 0.6595,
+      "step": 13797
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.56598643530827,
+      "learning_rate": 3.539465564486577e-07,
+      "loss": 0.6753,
+      "step": 13798
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8312929521454753,
+      "learning_rate": 3.5356362202674687e-07,
+      "loss": 0.6308,
+      "step": 13799
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.670035231648277,
+      "learning_rate": 3.5318088727255506e-07,
+      "loss": 0.7262,
+      "step": 13800
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9028436098930999,
+      "learning_rate": 3.527983522025291e-07,
+      "loss": 0.7951,
+      "step": 13801
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9130723727207863,
+      "learning_rate": 3.5241601683310924e-07,
+      "loss": 0.6019,
+      "step": 13802
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8968574073158069,
+      "learning_rate": 3.520338811807222e-07,
+      "loss": 0.5813,
+      "step": 13803
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3122324550822941,
+      "learning_rate": 3.516519452617922e-07,
+      "loss": 0.6756,
+      "step": 13804
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.090691265632236,
+      "learning_rate": 3.5127020909273e-07,
+      "loss": 0.6335,
+      "step": 13805
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0684342161257305,
+      "learning_rate": 3.508886726899402e-07,
+      "loss": 0.5713,
+      "step": 13806
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.768036590973196,
+      "learning_rate": 3.5050733606981855e-07,
+      "loss": 0.7407,
+      "step": 13807
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6093936120133447,
+      "learning_rate": 3.501261992487509e-07,
+      "loss": 0.6377,
+      "step": 13808
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.625665407117105,
+      "learning_rate": 3.4974526224311744e-07,
+      "loss": 0.7521,
+      "step": 13809
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5517353811924275,
+      "learning_rate": 3.493645250692862e-07,
+      "loss": 0.6355,
+      "step": 13810
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8932925530179519,
+      "learning_rate": 3.4898398774361854e-07,
+      "loss": 0.74,
+      "step": 13811
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9372274282030955,
+      "learning_rate": 3.486036502824675e-07,
+      "loss": 0.7103,
+      "step": 13812
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8126963783376033,
+      "learning_rate": 3.4822351270217656e-07,
+      "loss": 0.6699,
+      "step": 13813
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6504745003625028,
+      "learning_rate": 3.478435750190817e-07,
+      "loss": 0.7317,
+      "step": 13814
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5781327741602496,
+      "learning_rate": 3.474638372495098e-07,
+      "loss": 0.6727,
+      "step": 13815
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0287024136326055,
+      "learning_rate": 3.4708429940977785e-07,
+      "loss": 0.6624,
+      "step": 13816
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.762868811329328,
+      "learning_rate": 3.467049615161966e-07,
+      "loss": 0.7421,
+      "step": 13817
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6775859103224096,
+      "learning_rate": 3.463258235850653e-07,
+      "loss": 0.7039,
+      "step": 13818
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8971521274155725,
+      "learning_rate": 3.459468856326792e-07,
+      "loss": 0.7413,
+      "step": 13819
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8631330664273267,
+      "learning_rate": 3.455681476753209e-07,
+      "loss": 0.7262,
+      "step": 13820
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8035885791153898,
+      "learning_rate": 3.451896097292634e-07,
+      "loss": 0.7631,
+      "step": 13821
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.936372288668027,
+      "learning_rate": 3.448112718107766e-07,
+      "loss": 0.7453,
+      "step": 13822
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8547631704274585,
+      "learning_rate": 3.444331339361168e-07,
+      "loss": 0.8005,
+      "step": 13823
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6690488307864433,
+      "learning_rate": 3.4405519612153326e-07,
+      "loss": 0.7405,
+      "step": 13824
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8371735382226777,
+      "learning_rate": 3.4367745838326807e-07,
+      "loss": 0.7216,
+      "step": 13825
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8039972920924834,
+      "learning_rate": 3.432999207375515e-07,
+      "loss": 0.7518,
+      "step": 13826
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7341728641507221,
+      "learning_rate": 3.4292258320061e-07,
+      "loss": 0.7947,
+      "step": 13827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5489433730778754,
+      "learning_rate": 3.425454457886551e-07,
+      "loss": 0.7567,
+      "step": 13828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6102370136318882,
+      "learning_rate": 3.4216850851789663e-07,
+      "loss": 0.7571,
+      "step": 13829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9355041145952565,
+      "learning_rate": 3.4179177140453045e-07,
+      "loss": 0.564,
+      "step": 13830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8749356128790435,
+      "learning_rate": 3.414152344647459e-07,
+      "loss": 0.8362,
+      "step": 13831
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7337604280909296,
+      "learning_rate": 3.410388977147244e-07,
+      "loss": 0.6302,
+      "step": 13832
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9999220273674267,
+      "learning_rate": 3.406627611706376e-07,
+      "loss": 0.6456,
+      "step": 13833
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8506827257914136,
+      "learning_rate": 3.402868248486485e-07,
+      "loss": 0.6859,
+      "step": 13834
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.791747183361858,
+      "learning_rate": 3.3991108876491263e-07,
+      "loss": 0.7229,
+      "step": 13835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.943826265622646,
+      "learning_rate": 3.3953555293557483e-07,
+      "loss": 0.7141,
+      "step": 13836
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.61485412571985,
+      "learning_rate": 3.391602173767744e-07,
+      "loss": 0.5888,
+      "step": 13837
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5872159449629497,
+      "learning_rate": 3.387850821046401e-07,
+      "loss": 0.7537,
+      "step": 13838
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.776967106379513,
+      "learning_rate": 3.3841014713529184e-07,
+      "loss": 0.7669,
+      "step": 13839
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.742240552220232,
+      "learning_rate": 3.380354124848412e-07,
+      "loss": 0.8115,
+      "step": 13840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3211788142573733,
+      "learning_rate": 3.376608781693913e-07,
+      "loss": 0.6727,
+      "step": 13841
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7139840418019912,
+      "learning_rate": 3.372865442050377e-07,
+      "loss": 0.6774,
+      "step": 13842
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0289368837216732,
+      "learning_rate": 3.3691241060786586e-07,
+      "loss": 0.6808,
+      "step": 13843
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4853170303414533,
+      "learning_rate": 3.3653847739395174e-07,
+      "loss": 0.715,
+      "step": 13844
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9211316429625525,
+      "learning_rate": 3.36164744579367e-07,
+      "loss": 0.7769,
+      "step": 13845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6979722470707543,
+      "learning_rate": 3.357912121801682e-07,
+      "loss": 0.679,
+      "step": 13846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4516180728787702,
+      "learning_rate": 3.354178802124103e-07,
+      "loss": 0.7219,
+      "step": 13847
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5691189030556933,
+      "learning_rate": 3.350447486921338e-07,
+      "loss": 0.5841,
+      "step": 13848
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.804672542198326,
+      "learning_rate": 3.346718176353747e-07,
+      "loss": 0.7701,
+      "step": 13849
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.806339197446796,
+      "learning_rate": 3.342990870581575e-07,
+      "loss": 0.6176,
+      "step": 13850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.011725639548621,
+      "learning_rate": 3.339265569764993e-07,
+      "loss": 0.693,
+      "step": 13851
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5692255358643656,
+      "learning_rate": 3.3355422740640905e-07,
+      "loss": 0.8068,
+      "step": 13852
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6124995457117943,
+      "learning_rate": 3.331820983638867e-07,
+      "loss": 0.6275,
+      "step": 13853
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.839314543922145,
+      "learning_rate": 3.3281016986492165e-07,
+      "loss": 0.7065,
+      "step": 13854
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.033076519290409,
+      "learning_rate": 3.3243844192549947e-07,
+      "loss": 0.7575,
+      "step": 13855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5842940101832703,
+      "learning_rate": 3.3206691456159233e-07,
+      "loss": 0.633,
+      "step": 13856
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.872752481882602,
+      "learning_rate": 3.316955877891659e-07,
+      "loss": 0.7811,
+      "step": 13857
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7523329139595003,
+      "learning_rate": 3.313244616241773e-07,
+      "loss": 0.7383,
+      "step": 13858
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2479662267671885,
+      "learning_rate": 3.3095353608257385e-07,
+      "loss": 0.7098,
+      "step": 13859
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.697904184732168,
+      "learning_rate": 3.3058281118029553e-07,
+      "loss": 0.5926,
+      "step": 13860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2742581465724347,
+      "learning_rate": 3.302122869332736e-07,
+      "loss": 0.746,
+      "step": 13861
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.780397452057552,
+      "learning_rate": 3.2984196335742844e-07,
+      "loss": 0.6998,
+      "step": 13862
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6256959861872669,
+      "learning_rate": 3.294718404686775e-07,
+      "loss": 0.6999,
+      "step": 13863
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5357526875766387,
+      "learning_rate": 3.2910191828292083e-07,
+      "loss": 0.6801,
+      "step": 13864
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5671200645162193,
+      "learning_rate": 3.2873219681605894e-07,
+      "loss": 0.7308,
+      "step": 13865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2430805465851076,
+      "learning_rate": 3.283626760839775e-07,
+      "loss": 0.6092,
+      "step": 13866
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5916469933249928,
+      "learning_rate": 3.279933561025567e-07,
+      "loss": 0.7776,
+      "step": 13867
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.829061044475579,
+      "learning_rate": 3.276242368876664e-07,
+      "loss": 0.7492,
+      "step": 13868
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6434452036040639,
+      "learning_rate": 3.2725531845516744e-07,
+      "loss": 0.7607,
+      "step": 13869
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.26761202662012,
+      "learning_rate": 3.2688660082091594e-07,
+      "loss": 0.7088,
+      "step": 13870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7424451351293881,
+      "learning_rate": 3.2651808400075426e-07,
+      "loss": 0.7094,
+      "step": 13871
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8882523688499322,
+      "learning_rate": 3.2614976801051745e-07,
+      "loss": 0.7583,
+      "step": 13872
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8953886649726157,
+      "learning_rate": 3.257816528660357e-07,
+      "loss": 0.7216,
+      "step": 13873
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1275202852737438,
+      "learning_rate": 3.254137385831263e-07,
+      "loss": 0.6879,
+      "step": 13874
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5799898459502082,
+      "learning_rate": 3.250460251775994e-07,
+      "loss": 0.7154,
+      "step": 13875
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9152891811841666,
+      "learning_rate": 3.246785126652563e-07,
+      "loss": 0.7281,
+      "step": 13876
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7455258917594638,
+      "learning_rate": 3.243112010618893e-07,
+      "loss": 0.7756,
+      "step": 13877
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9186978780582205,
+      "learning_rate": 3.239440903832852e-07,
+      "loss": 0.5678,
+      "step": 13878
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9933698347186157,
+      "learning_rate": 3.2357718064521594e-07,
+      "loss": 0.7081,
+      "step": 13879
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1360066663377895,
+      "learning_rate": 3.2321047186345047e-07,
+      "loss": 0.8272,
+      "step": 13880
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9237940078465234,
+      "learning_rate": 3.2284396405374787e-07,
+      "loss": 0.6872,
+      "step": 13881
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6288374280222615,
+      "learning_rate": 3.2247765723185556e-07,
+      "loss": 0.6594,
+      "step": 13882
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.612463218985487,
+      "learning_rate": 3.221115514135159e-07,
+      "loss": 0.5665,
+      "step": 13883
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7660501081387803,
+      "learning_rate": 3.217456466144614e-07,
+      "loss": 0.8046,
+      "step": 13884
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.680830573515881,
+      "learning_rate": 3.213799428504155e-07,
+      "loss": 0.6761,
+      "step": 13885
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9236692189801659,
+      "learning_rate": 3.210144401370935e-07,
+      "loss": 0.6823,
+      "step": 13886
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.056464195906274,
+      "learning_rate": 3.206491384902005e-07,
+      "loss": 0.636,
+      "step": 13887
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.492851560568587,
+      "learning_rate": 3.202840379254374e-07,
+      "loss": 0.7014,
+      "step": 13888
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7896186881738858,
+      "learning_rate": 3.199191384584893e-07,
+      "loss": 0.7002,
+      "step": 13889
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.152051852095717,
+      "learning_rate": 3.1955444010504044e-07,
+      "loss": 0.6607,
+      "step": 13890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5868988705018336,
+      "learning_rate": 3.19189942880761e-07,
+      "loss": 0.708,
+      "step": 13891
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0422137579741597,
+      "learning_rate": 3.18825646801314e-07,
+      "loss": 0.6945,
+      "step": 13892
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8331864914461207,
+      "learning_rate": 3.184615518823547e-07,
+      "loss": 0.6378,
+      "step": 13893
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4457937077548815,
+      "learning_rate": 3.180976581395295e-07,
+      "loss": 0.6071,
+      "step": 13894
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.151006923880501,
+      "learning_rate": 3.177339655884737e-07,
+      "loss": 0.6618,
+      "step": 13895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4983172872387205,
+      "learning_rate": 3.173704742448186e-07,
+      "loss": 0.7236,
+      "step": 13896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7493894432081858,
+      "learning_rate": 3.170071841241823e-07,
+      "loss": 0.5234,
+      "step": 13897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6456097586748728,
+      "learning_rate": 3.1664409524217677e-07,
+      "loss": 0.7084,
+      "step": 13898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9819929494821775,
+      "learning_rate": 3.1628120761440616e-07,
+      "loss": 0.7791,
+      "step": 13899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0912697831024014,
+      "learning_rate": 3.1591852125646196e-07,
+      "loss": 0.6534,
+      "step": 13900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.16552198654474,
+      "learning_rate": 3.1555603618393106e-07,
+      "loss": 0.8075,
+      "step": 13901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7610852149983725,
+      "learning_rate": 3.151937524123905e-07,
+      "loss": 0.6903,
+      "step": 13902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.794987845214736,
+      "learning_rate": 3.148316699574078e-07,
+      "loss": 0.7459,
+      "step": 13903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5581527719083663,
+      "learning_rate": 3.144697888345427e-07,
+      "loss": 0.6694,
+      "step": 13904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9356711238012185,
+      "learning_rate": 3.1410810905934564e-07,
+      "loss": 0.7803,
+      "step": 13905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3582468531159968,
+      "learning_rate": 3.137466306473602e-07,
+      "loss": 0.7581,
+      "step": 13906
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.715283414091849,
+      "learning_rate": 3.1338535361411737e-07,
+      "loss": 0.6082,
+      "step": 13907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5057237755412762,
+      "learning_rate": 3.1302427797514465e-07,
+      "loss": 0.6697,
+      "step": 13908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.948139829502076,
+      "learning_rate": 3.1266340374595693e-07,
+      "loss": 0.6203,
+      "step": 13909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.724321287289265,
+      "learning_rate": 3.1230273094206233e-07,
+      "loss": 0.7708,
+      "step": 13910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.789902740976854,
+      "learning_rate": 3.1194225957895895e-07,
+      "loss": 0.708,
+      "step": 13911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.040163815696575,
+      "learning_rate": 3.115819896721378e-07,
+      "loss": 0.6639,
+      "step": 13912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9392401676505695,
+      "learning_rate": 3.1122192123707985e-07,
+      "loss": 0.688,
+      "step": 13913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7835477760619094,
+      "learning_rate": 3.108620542892593e-07,
+      "loss": 0.6255,
+      "step": 13914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.235353746252407,
+      "learning_rate": 3.105023888441383e-07,
+      "loss": 0.6484,
+      "step": 13915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3539515049986839,
+      "learning_rate": 3.1014292491717444e-07,
+      "loss": 0.6022,
+      "step": 13916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.529721135695161,
+      "learning_rate": 3.0978366252381376e-07,
+      "loss": 0.7684,
+      "step": 13917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7326729694387066,
+      "learning_rate": 3.0942460167949495e-07,
+      "loss": 0.788,
+      "step": 13918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.659162310693652,
+      "learning_rate": 3.0906574239964795e-07,
+      "loss": 0.6181,
+      "step": 13919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8663489706988186,
+      "learning_rate": 3.08707084699692e-07,
+      "loss": 0.6227,
+      "step": 13920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6522432152356799,
+      "learning_rate": 3.083486285950421e-07,
+      "loss": 0.6381,
+      "step": 13921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7836793834163769,
+      "learning_rate": 3.0799037410109976e-07,
+      "loss": 0.7056,
+      "step": 13922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8668412715204865,
+      "learning_rate": 3.076323212332605e-07,
+      "loss": 0.6287,
+      "step": 13923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6850509206027748,
+      "learning_rate": 3.072744700069119e-07,
+      "loss": 0.7108,
+      "step": 13924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8958227770332443,
+      "learning_rate": 3.06916820437429e-07,
+      "loss": 0.7901,
+      "step": 13925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7909593925210767,
+      "learning_rate": 3.065593725401833e-07,
+      "loss": 0.7881,
+      "step": 13926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0203291612402499,
+      "learning_rate": 3.0620212633053424e-07,
+      "loss": 0.6643,
+      "step": 13927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0632508498776096,
+      "learning_rate": 3.0584508182383346e-07,
+      "loss": 0.6486,
+      "step": 13928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7723482392860548,
+      "learning_rate": 3.054882390354241e-07,
+      "loss": 0.6978,
+      "step": 13929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7847747171016533,
+      "learning_rate": 3.0513159798063906e-07,
+      "loss": 0.7659,
+      "step": 13930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.080418924729355,
+      "learning_rate": 3.0477515867480655e-07,
+      "loss": 0.808,
+      "step": 13931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.60360992806739,
+      "learning_rate": 3.0441892113324265e-07,
+      "loss": 0.6974,
+      "step": 13932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7341650404382405,
+      "learning_rate": 3.0406288537125403e-07,
+      "loss": 0.7788,
+      "step": 13933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6805444770540887,
+      "learning_rate": 3.0370705140414293e-07,
+      "loss": 0.6749,
+      "step": 13934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.498798796732144,
+      "learning_rate": 3.0335141924719813e-07,
+      "loss": 0.6577,
+      "step": 13935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.07861761373819,
+      "learning_rate": 3.029959889157036e-07,
+      "loss": 0.7769,
+      "step": 13936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.767906760073703,
+      "learning_rate": 3.026407604249315e-07,
+      "loss": 0.8731,
+      "step": 13937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9262492578196853,
+      "learning_rate": 3.0228573379014745e-07,
+      "loss": 0.7234,
+      "step": 13938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8134065894277474,
+      "learning_rate": 3.019309090266087e-07,
+      "loss": 0.8258,
+      "step": 13939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8054008322650414,
+      "learning_rate": 3.015762861495608e-07,
+      "loss": 0.772,
+      "step": 13940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.729752851955382,
+      "learning_rate": 3.0122186517424435e-07,
+      "loss": 0.6795,
+      "step": 13941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4840646116154168,
+      "learning_rate": 3.0086764611589047e-07,
+      "loss": 0.7494,
+      "step": 13942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4205623231109246,
+      "learning_rate": 3.0051362898971704e-07,
+      "loss": 0.7458,
+      "step": 13943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6354851104693482,
+      "learning_rate": 3.0015981381094073e-07,
+      "loss": 0.838,
+      "step": 13944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.825327612690608,
+      "learning_rate": 2.9980620059476384e-07,
+      "loss": 0.7542,
+      "step": 13945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7770206976085172,
+      "learning_rate": 2.9945278935638255e-07,
+      "loss": 0.7301,
+      "step": 13946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4869764841822946,
+      "learning_rate": 2.990995801109836e-07,
+      "loss": 0.6114,
+      "step": 13947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4721895133996963,
+      "learning_rate": 2.987465728737443e-07,
+      "loss": 0.6534,
+      "step": 13948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9435409692634154,
+      "learning_rate": 2.9839376765983583e-07,
+      "loss": 0.6214,
+      "step": 13949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6727347850882202,
+      "learning_rate": 2.980411644844189e-07,
+      "loss": 0.6526,
+      "step": 13950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6992308005026233,
+      "learning_rate": 2.976887633626435e-07,
+      "loss": 0.738,
+      "step": 13951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6779658811610332,
+      "learning_rate": 2.973365643096554e-07,
+      "loss": 0.6401,
+      "step": 13952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7278260205550706,
+      "learning_rate": 2.9698456734058856e-07,
+      "loss": 0.7058,
+      "step": 13953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2890213830505113,
+      "learning_rate": 2.9663277247056923e-07,
+      "loss": 0.6317,
+      "step": 13954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6142416730019402,
+      "learning_rate": 2.9628117971471416e-07,
+      "loss": 0.583,
+      "step": 13955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6319767452021487,
+      "learning_rate": 2.959297890881319e-07,
+      "loss": 0.6303,
+      "step": 13956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7977038397601035,
+      "learning_rate": 2.955786006059247e-07,
+      "loss": 0.6871,
+      "step": 13957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.266357289781446,
+      "learning_rate": 2.952276142831806e-07,
+      "loss": 0.6464,
+      "step": 13958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1819452579941199,
+      "learning_rate": 2.9487683013498523e-07,
+      "loss": 0.6336,
+      "step": 13959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9130460957355062,
+      "learning_rate": 2.9452624817641163e-07,
+      "loss": 0.7542,
+      "step": 13960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5927161501447062,
+      "learning_rate": 2.9417586842252375e-07,
+      "loss": 0.7144,
+      "step": 13961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8740316110178177,
+      "learning_rate": 2.9382569088837966e-07,
+      "loss": 0.7965,
+      "step": 13962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5531071176103886,
+      "learning_rate": 2.934757155890272e-07,
+      "loss": 0.633,
+      "step": 13963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.753044132918518,
+      "learning_rate": 2.93125942539505e-07,
+      "loss": 0.691,
+      "step": 13964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7501032792982354,
+      "learning_rate": 2.9277637175484376e-07,
+      "loss": 0.6106,
+      "step": 13965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.745809953037248,
+      "learning_rate": 2.924270032500648e-07,
+      "loss": 0.8121,
+      "step": 13966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9344686845379822,
+      "learning_rate": 2.9207783704018223e-07,
+      "loss": 0.5652,
+      "step": 13967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.71324862356857,
+      "learning_rate": 2.9172887314020017e-07,
+      "loss": 0.7561,
+      "step": 13968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8344700192573498,
+      "learning_rate": 2.913801115651144e-07,
+      "loss": 0.7279,
+      "step": 13969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5838883060778737,
+      "learning_rate": 2.910315523299123e-07,
+      "loss": 0.6314,
+      "step": 13970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8323789282289704,
+      "learning_rate": 2.906831954495715e-07,
+      "loss": 0.7101,
+      "step": 13971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4545159468640592,
+      "learning_rate": 2.9033504093906207e-07,
+      "loss": 0.5478,
+      "step": 13972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6303485011124816,
+      "learning_rate": 2.8998708881334437e-07,
+      "loss": 0.6869,
+      "step": 13973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0670523581882976,
+      "learning_rate": 2.896393390873714e-07,
+      "loss": 0.6859,
+      "step": 13974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6722003552453384,
+      "learning_rate": 2.892917917760873e-07,
+      "loss": 0.7865,
+      "step": 13975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5397294156625203,
+      "learning_rate": 2.8894444689442526e-07,
+      "loss": 0.685,
+      "step": 13976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7583994432134495,
+      "learning_rate": 2.885973044573126e-07,
+      "loss": 0.7494,
+      "step": 13977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.92404931641739,
+      "learning_rate": 2.8825036447966745e-07,
+      "loss": 0.7066,
+      "step": 13978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.282210908169475,
+      "learning_rate": 2.8790362697639685e-07,
+      "loss": 0.7243,
+      "step": 13979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5252191544414535,
+      "learning_rate": 2.8755709196240264e-07,
+      "loss": 0.7201,
+      "step": 13980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9372032284340803,
+      "learning_rate": 2.8721075945257414e-07,
+      "loss": 0.6751,
+      "step": 13981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3431499796829458,
+      "learning_rate": 2.868646294617966e-07,
+      "loss": 0.6266,
+      "step": 13982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5946756845703023,
+      "learning_rate": 2.86518702004942e-07,
+      "loss": 0.7101,
+      "step": 13983
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8157633870147385,
+      "learning_rate": 2.8617297709687577e-07,
+      "loss": 0.7852,
+      "step": 13984
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.729156702754403,
+      "learning_rate": 2.858274547524559e-07,
+      "loss": 0.7043,
+      "step": 13985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2647331069319436,
+      "learning_rate": 2.854821349865289e-07,
+      "loss": 0.7403,
+      "step": 13986
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7124379510959946,
+      "learning_rate": 2.8513701781393456e-07,
+      "loss": 0.8274,
+      "step": 13987
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1257804644849732,
+      "learning_rate": 2.8479210324950266e-07,
+      "loss": 0.6148,
+      "step": 13988
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8097788073528698,
+      "learning_rate": 2.8444739130805587e-07,
+      "loss": 0.6523,
+      "step": 13989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6158153410469693,
+      "learning_rate": 2.841028820044067e-07,
+      "loss": 0.7134,
+      "step": 13990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6653223831973438,
+      "learning_rate": 2.837585753533589e-07,
+      "loss": 0.6189,
+      "step": 13991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8980340526820756,
+      "learning_rate": 2.8341447136970954e-07,
+      "loss": 0.8207,
+      "step": 13992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5734277621973933,
+      "learning_rate": 2.8307057006824514e-07,
+      "loss": 0.5982,
+      "step": 13993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8892091463170468,
+      "learning_rate": 2.827268714637421e-07,
+      "loss": 0.6859,
+      "step": 13994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0959645691486855,
+      "learning_rate": 2.823833755709721e-07,
+      "loss": 0.7254,
+      "step": 13995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7307858813774484,
+      "learning_rate": 2.820400824046954e-07,
+      "loss": 0.708,
+      "step": 13996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9058855059024917,
+      "learning_rate": 2.816969919796636e-07,
+      "loss": 0.7158,
+      "step": 13997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8689588309231944,
+      "learning_rate": 2.813541043106205e-07,
+      "loss": 0.6581,
+      "step": 13998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7641064022964537,
+      "learning_rate": 2.810114194122998e-07,
+      "loss": 0.8242,
+      "step": 13999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.564383306890229,
+      "learning_rate": 2.806689372994292e-07,
+      "loss": 0.603,
+      "step": 14000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7012681896151345,
+      "learning_rate": 2.803266579867242e-07,
+      "loss": 0.7952,
+      "step": 14001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6673788558372429,
+      "learning_rate": 2.7998458148889354e-07,
+      "loss": 0.7084,
+      "step": 14002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7987473806862164,
+      "learning_rate": 2.796427078206382e-07,
+      "loss": 0.695,
+      "step": 14003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.865223085382591,
+      "learning_rate": 2.793010369966487e-07,
+      "loss": 0.7081,
+      "step": 14004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6696378482700758,
+      "learning_rate": 2.789595690316066e-07,
+      "loss": 0.6258,
+      "step": 14005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.70841586948917,
+      "learning_rate": 2.7861830394018627e-07,
+      "loss": 0.6607,
+      "step": 14006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6814674792893367,
+      "learning_rate": 2.7827724173705273e-07,
+      "loss": 0.6507,
+      "step": 14007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.182697866711336,
+      "learning_rate": 2.779363824368619e-07,
+      "loss": 0.6772,
+      "step": 14008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.599505692926617,
+      "learning_rate": 2.7759572605426057e-07,
+      "loss": 0.6263,
+      "step": 14009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.547910698117605,
+      "learning_rate": 2.7725527260388905e-07,
+      "loss": 0.7917,
+      "step": 14010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9880620840203933,
+      "learning_rate": 2.7691502210037744e-07,
+      "loss": 0.6792,
+      "step": 14011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8512599649641337,
+      "learning_rate": 2.7657497455834457e-07,
+      "loss": 0.6463,
+      "step": 14012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6826526278831058,
+      "learning_rate": 2.762351299924054e-07,
+      "loss": 0.7574,
+      "step": 14013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1920190385895795,
+      "learning_rate": 2.7589548841716274e-07,
+      "loss": 0.6734,
+      "step": 14014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7242115346433016,
+      "learning_rate": 2.755560498472126e-07,
+      "loss": 0.8124,
+      "step": 14015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.508890198231387,
+      "learning_rate": 2.752168142971406e-07,
+      "loss": 0.6454,
+      "step": 14016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6749478271088123,
+      "learning_rate": 2.7487778178152457e-07,
+      "loss": 0.6929,
+      "step": 14017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7527690252199384,
+      "learning_rate": 2.7453895231493445e-07,
+      "loss": 0.6692,
+      "step": 14018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8643336153893015,
+      "learning_rate": 2.7420032591192856e-07,
+      "loss": 0.7549,
+      "step": 14019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9874760348174312,
+      "learning_rate": 2.7386190258706033e-07,
+      "loss": 0.7544,
+      "step": 14020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7813204536556497,
+      "learning_rate": 2.735236823548715e-07,
+      "loss": 0.8215,
+      "step": 14021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.288250859242323,
+      "learning_rate": 2.73185665229897e-07,
+      "loss": 0.73,
+      "step": 14022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.561178571164522,
+      "learning_rate": 2.7284785122666136e-07,
+      "loss": 0.6009,
+      "step": 14023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8977891125370345,
+      "learning_rate": 2.7251024035968134e-07,
+      "loss": 0.6948,
+      "step": 14024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5297917096381553,
+      "learning_rate": 2.721728326434653e-07,
+      "loss": 0.7235,
+      "step": 14025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8904962889247812,
+      "learning_rate": 2.718356280925116e-07,
+      "loss": 0.6184,
+      "step": 14026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8070835776156442,
+      "learning_rate": 2.71498626721311e-07,
+      "loss": 0.7935,
+      "step": 14027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5378105507055602,
+      "learning_rate": 2.711618285443457e-07,
+      "loss": 0.7331,
+      "step": 14028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6209105953944192,
+      "learning_rate": 2.7082523357608856e-07,
+      "loss": 0.7639,
+      "step": 14029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4867166154357643,
+      "learning_rate": 2.7048884183100313e-07,
+      "loss": 0.6662,
+      "step": 14030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6123929841673121,
+      "learning_rate": 2.70152653323546e-07,
+      "loss": 0.7045,
+      "step": 14031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7247855315049723,
+      "learning_rate": 2.6981666806816185e-07,
+      "loss": 0.6482,
+      "step": 14032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8635454087991523,
+      "learning_rate": 2.694808860792919e-07,
+      "loss": 0.7803,
+      "step": 14033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.709980817767303,
+      "learning_rate": 2.6914530737136346e-07,
+      "loss": 0.7356,
+      "step": 14034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9043409086978903,
+      "learning_rate": 2.6880993195879614e-07,
+      "loss": 0.583,
+      "step": 14035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7598323589336167,
+      "learning_rate": 2.684747598560045e-07,
+      "loss": 0.6512,
+      "step": 14036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6133073721680575,
+      "learning_rate": 2.6813979107738864e-07,
+      "loss": 0.7031,
+      "step": 14037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.529875459706574,
+      "learning_rate": 2.678050256373449e-07,
+      "loss": 0.597,
+      "step": 14038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.561773092381844,
+      "learning_rate": 2.674704635502584e-07,
+      "loss": 0.7641,
+      "step": 14039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6450786031574771,
+      "learning_rate": 2.6713610483050654e-07,
+      "loss": 0.6614,
+      "step": 14040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4976688988701836,
+      "learning_rate": 2.668019494924562e-07,
+      "loss": 0.7102,
+      "step": 14041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7532199651180682,
+      "learning_rate": 2.6646799755046746e-07,
+      "loss": 0.6771,
+      "step": 14042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8113595014233976,
+      "learning_rate": 2.661342490188917e-07,
+      "loss": 0.7699,
+      "step": 14043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7652214982066765,
+      "learning_rate": 2.658007039120697e-07,
+      "loss": 0.6815,
+      "step": 14044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.600752509070468,
+      "learning_rate": 2.654673622443349e-07,
+      "loss": 0.689,
+      "step": 14045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8053605281612988,
+      "learning_rate": 2.65134224030012e-07,
+      "loss": 0.7574,
+      "step": 14046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8965155174047956,
+      "learning_rate": 2.6480128928341675e-07,
+      "loss": 0.6684,
+      "step": 14047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6333740344084677,
+      "learning_rate": 2.6446855801885606e-07,
+      "loss": 0.8311,
+      "step": 14048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.859770961331028,
+      "learning_rate": 2.64136030250628e-07,
+      "loss": 0.7729,
+      "step": 14049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7033507800792613,
+      "learning_rate": 2.638037059930215e-07,
+      "loss": 0.8221,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3097032455803523,
+      "learning_rate": 2.634715852603187e-07,
+      "loss": 0.7538,
+      "step": 14051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6469301260236258,
+      "learning_rate": 2.631396680667908e-07,
+      "loss": 0.7308,
+      "step": 14052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3447881535326252,
+      "learning_rate": 2.6280795442669925e-07,
+      "loss": 0.5923,
+      "step": 14053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7235876803781223,
+      "learning_rate": 2.6247644435430263e-07,
+      "loss": 0.6134,
+      "step": 14054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6851153161392793,
+      "learning_rate": 2.621451378638418e-07,
+      "loss": 0.7353,
+      "step": 14055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.861121062838136,
+      "learning_rate": 2.618140349695575e-07,
+      "loss": 0.7225,
+      "step": 14056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.722720125717672,
+      "learning_rate": 2.6148313568567674e-07,
+      "loss": 0.7689,
+      "step": 14057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7733534701686235,
+      "learning_rate": 2.611524400264187e-07,
+      "loss": 0.8317,
+      "step": 14058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0713352700320868,
+      "learning_rate": 2.6082194800599424e-07,
+      "loss": 0.6975,
+      "step": 14059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8860069892037512,
+      "learning_rate": 2.6049165963860425e-07,
+      "loss": 0.6527,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7184078614761555,
+      "learning_rate": 2.601615749384451e-07,
+      "loss": 0.6661,
+      "step": 14061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7094454226750548,
+      "learning_rate": 2.5983169391969823e-07,
+      "loss": 0.6671,
+      "step": 14062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.097569692592239,
+      "learning_rate": 2.595020165965401e-07,
+      "loss": 0.7006,
+      "step": 14063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8577225189949278,
+      "learning_rate": 2.591725429831382e-07,
+      "loss": 0.6915,
+      "step": 14064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5669721711074316,
+      "learning_rate": 2.5884327309365074e-07,
+      "loss": 0.694,
+      "step": 14065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7148448056550059,
+      "learning_rate": 2.5851420694222697e-07,
+      "loss": 0.7276,
+      "step": 14066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6732990354937765,
+      "learning_rate": 2.581853445430077e-07,
+      "loss": 0.7281,
+      "step": 14067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8414270353849669,
+      "learning_rate": 2.5785668591012393e-07,
+      "loss": 0.7059,
+      "step": 14068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7556804116578195,
+      "learning_rate": 2.57528231057701e-07,
+      "loss": 0.682,
+      "step": 14069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1489941263156007,
+      "learning_rate": 2.571999799998509e-07,
+      "loss": 0.7485,
+      "step": 14070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6363253403042732,
+      "learning_rate": 2.568719327506808e-07,
+      "loss": 0.6348,
+      "step": 14071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0699560179264322,
+      "learning_rate": 2.565440893242882e-07,
+      "loss": 0.6643,
+      "step": 14072
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7292698674095228,
+      "learning_rate": 2.5621644973475914e-07,
+      "loss": 0.6564,
+      "step": 14073
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9434432683527212,
+      "learning_rate": 2.558890139961745e-07,
+      "loss": 0.7192,
+      "step": 14074
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6281582905551937,
+      "learning_rate": 2.5556178212260474e-07,
+      "loss": 0.6097,
+      "step": 14075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5529851832076138,
+      "learning_rate": 2.552347541281114e-07,
+      "loss": 0.7321,
+      "step": 14076
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6260366188255992,
+      "learning_rate": 2.549079300267482e-07,
+      "loss": 0.6704,
+      "step": 14077
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.66106448261476,
+      "learning_rate": 2.5458130983255835e-07,
+      "loss": 0.7333,
+      "step": 14078
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7803341013255922,
+      "learning_rate": 2.5425489355957956e-07,
+      "loss": 0.5889,
+      "step": 14079
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.535314167506126,
+      "learning_rate": 2.539286812218361e-07,
+      "loss": 0.6506,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8769411238817342,
+      "learning_rate": 2.5360267283334796e-07,
+      "loss": 0.6501,
+      "step": 14081
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4773389254943863,
+      "learning_rate": 2.5327686840812326e-07,
+      "loss": 0.7871,
+      "step": 14082
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7136709459373336,
+      "learning_rate": 2.5295126796016366e-07,
+      "loss": 0.7163,
+      "step": 14083
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.651952046754631,
+      "learning_rate": 2.526258715034602e-07,
+      "loss": 0.6683,
+      "step": 14084
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.116642949350574,
+      "learning_rate": 2.523006790519955e-07,
+      "loss": 0.8431,
+      "step": 14085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8811341168013265,
+      "learning_rate": 2.5197569061974404e-07,
+      "loss": 0.75,
+      "step": 14086
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7579400920323738,
+      "learning_rate": 2.5165090622067343e-07,
+      "loss": 0.7525,
+      "step": 14087
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7217041394243648,
+      "learning_rate": 2.513263258687365e-07,
+      "loss": 0.8051,
+      "step": 14088
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.062505367516983,
+      "learning_rate": 2.510019495778837e-07,
+      "loss": 0.6768,
+      "step": 14089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2288862612959761,
+      "learning_rate": 2.5067777736205444e-07,
+      "loss": 0.6883,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.59929974765775,
+      "learning_rate": 2.503538092351782e-07,
+      "loss": 0.6367,
+      "step": 14091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6780057350830362,
+      "learning_rate": 2.5003004521117646e-07,
+      "loss": 0.6353,
+      "step": 14092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.703828715619746,
+      "learning_rate": 2.4970648530396157e-07,
+      "loss": 0.7784,
+      "step": 14093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0148405729588075,
+      "learning_rate": 2.4938312952744016e-07,
+      "loss": 0.7732,
+      "step": 14094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7602512167950242,
+      "learning_rate": 2.490599778955055e-07,
+      "loss": 0.6998,
+      "step": 14095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9544515795264386,
+      "learning_rate": 2.4873703042204323e-07,
+      "loss": 0.7319,
+      "step": 14096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8488768299983498,
+      "learning_rate": 2.484142871209344e-07,
+      "loss": 0.6103,
+      "step": 14097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.996063533041763,
+      "learning_rate": 2.480917480060441e-07,
+      "loss": 0.7764,
+      "step": 14098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7876459111596024,
+      "learning_rate": 2.477694130912356e-07,
+      "loss": 0.6669,
+      "step": 14099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.792085786539587,
+      "learning_rate": 2.474472823903584e-07,
+      "loss": 0.5526,
+      "step": 14100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.206025240151984,
+      "learning_rate": 2.4712535591725706e-07,
+      "loss": 0.6794,
+      "step": 14101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2392451874698573,
+      "learning_rate": 2.4680363368576376e-07,
+      "loss": 0.6841,
+      "step": 14102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6808115623638502,
+      "learning_rate": 2.464821157097036e-07,
+      "loss": 0.7101,
+      "step": 14103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9743379186331365,
+      "learning_rate": 2.461608020028944e-07,
+      "loss": 0.8341,
+      "step": 14104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6113579735454828,
+      "learning_rate": 2.458396925791434e-07,
+      "loss": 0.6561,
+      "step": 14105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8559799743235603,
+      "learning_rate": 2.4551878745224745e-07,
+      "loss": 0.6576,
+      "step": 14106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.957640602869256,
+      "learning_rate": 2.4519808663599876e-07,
+      "loss": 0.7937,
+      "step": 14107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0571645579172615,
+      "learning_rate": 2.44877590144178e-07,
+      "loss": 0.7417,
+      "step": 14108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5369423010575156,
+      "learning_rate": 2.445572979905575e-07,
+      "loss": 0.6936,
+      "step": 14109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6463297639994317,
+      "learning_rate": 2.442372101889007e-07,
+      "loss": 0.7518,
+      "step": 14110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5582769669334167,
+      "learning_rate": 2.4391732675296207e-07,
+      "loss": 0.6127,
+      "step": 14111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.654721530927425,
+      "learning_rate": 2.4359764769648907e-07,
+      "loss": 0.7245,
+      "step": 14112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.578744200516921,
+      "learning_rate": 2.4327817303321786e-07,
+      "loss": 0.7298,
+      "step": 14113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6604918333621561,
+      "learning_rate": 2.4295890277687695e-07,
+      "loss": 0.7287,
+      "step": 14114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8250726239265993,
+      "learning_rate": 2.4263983694118754e-07,
+      "loss": 0.7534,
+      "step": 14115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2413624850822447,
+      "learning_rate": 2.423209755398587e-07,
+      "loss": 0.6551,
+      "step": 14116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1267254731501009,
+      "learning_rate": 2.4200231858659384e-07,
+      "loss": 0.5652,
+      "step": 14117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7657681278458701,
+      "learning_rate": 2.416838660950854e-07,
+      "loss": 0.7329,
+      "step": 14118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.781317890791791,
+      "learning_rate": 2.4136561807901916e-07,
+      "loss": 0.7144,
+      "step": 14119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6410165342270222,
+      "learning_rate": 2.410475745520702e-07,
+      "loss": 0.6916,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7365344180061655,
+      "learning_rate": 2.4072973552790535e-07,
+      "loss": 0.5801,
+      "step": 14121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8286626061342626,
+      "learning_rate": 2.4041210102018375e-07,
+      "loss": 0.7006,
+      "step": 14122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0729181661860816,
+      "learning_rate": 2.40094671042555e-07,
+      "loss": 0.7259,
+      "step": 14123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.612218776988137,
+      "learning_rate": 2.397774456086577e-07,
+      "loss": 0.6935,
+      "step": 14124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3384618155177437,
+      "learning_rate": 2.3946042473212526e-07,
+      "loss": 0.6641,
+      "step": 14125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8310427524274424,
+      "learning_rate": 2.391436084265814e-07,
+      "loss": 0.657,
+      "step": 14126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.020077837966421,
+      "learning_rate": 2.3882699670563947e-07,
+      "loss": 0.7079,
+      "step": 14127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.106776591052957,
+      "learning_rate": 2.385105895829054e-07,
+      "loss": 0.8308,
+      "step": 14128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5612156856754666,
+      "learning_rate": 2.3819438707197495e-07,
+      "loss": 0.6355,
+      "step": 14129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4746156089709566,
+      "learning_rate": 2.378783891864378e-07,
+      "loss": 0.7653,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5791162299983805,
+      "learning_rate": 2.3756259593987086e-07,
+      "loss": 0.6989,
+      "step": 14131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2972927912984198,
+      "learning_rate": 2.372470073458466e-07,
+      "loss": 0.6431,
+      "step": 14132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.851833793293082,
+      "learning_rate": 2.3693162341792532e-07,
+      "loss": 0.7716,
+      "step": 14133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5951629845366264,
+      "learning_rate": 2.3661644416966057e-07,
+      "loss": 0.7352,
+      "step": 14134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.872912031973824,
+      "learning_rate": 2.3630146961459543e-07,
+      "loss": 0.6831,
+      "step": 14135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9143960129753492,
+      "learning_rate": 2.3598669976626576e-07,
+      "loss": 0.6395,
+      "step": 14136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7264425671247643,
+      "learning_rate": 2.3567213463819738e-07,
+      "loss": 0.6577,
+      "step": 14137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6687963350003607,
+      "learning_rate": 2.353577742439084e-07,
+      "loss": 0.7005,
+      "step": 14138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8926931007910448,
+      "learning_rate": 2.3504361859690628e-07,
+      "loss": 0.7604,
+      "step": 14139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9615121353634986,
+      "learning_rate": 2.347296677106925e-07,
+      "loss": 0.6584,
+      "step": 14140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7908607660455576,
+      "learning_rate": 2.344159215987585e-07,
+      "loss": 0.6836,
+      "step": 14141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3404464627776063,
+      "learning_rate": 2.341023802745851e-07,
+      "loss": 0.6567,
+      "step": 14142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6305577376276115,
+      "learning_rate": 2.3378904375164714e-07,
+      "loss": 0.7795,
+      "step": 14143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.627566302516718,
+      "learning_rate": 2.3347591204340881e-07,
+      "loss": 0.6594,
+      "step": 14144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.692761167450683,
+      "learning_rate": 2.3316298516332604e-07,
+      "loss": 0.7384,
+      "step": 14145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0017643176713185,
+      "learning_rate": 2.3285026312484583e-07,
+      "loss": 0.6585,
+      "step": 14146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4657862255575163,
+      "learning_rate": 2.3253774594140633e-07,
+      "loss": 0.7344,
+      "step": 14147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7792497835197016,
+      "learning_rate": 2.32225433626439e-07,
+      "loss": 0.7149,
+      "step": 14148
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8163719122317876,
+      "learning_rate": 2.3191332619336204e-07,
+      "loss": 0.8348,
+      "step": 14149
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6304721482615367,
+      "learning_rate": 2.3160142365558914e-07,
+      "loss": 0.6027,
+      "step": 14150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6945782761729675,
+      "learning_rate": 2.312897260265229e-07,
+      "loss": 0.7172,
+      "step": 14151
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8726301101618108,
+      "learning_rate": 2.309782333195576e-07,
+      "loss": 0.6661,
+      "step": 14152
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5429056501201943,
+      "learning_rate": 2.3066694554807867e-07,
+      "loss": 0.665,
+      "step": 14153
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.55314540837943,
+      "learning_rate": 2.3035586272546207e-07,
+      "loss": 0.7395,
+      "step": 14154
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8389159876084216,
+      "learning_rate": 2.300449848650782e-07,
+      "loss": 0.7324,
+      "step": 14155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.736065297064127,
+      "learning_rate": 2.2973431198028417e-07,
+      "loss": 0.7394,
+      "step": 14156
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6096004572502158,
+      "learning_rate": 2.2942384408442986e-07,
+      "loss": 0.7112,
+      "step": 14157
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5244449875316164,
+      "learning_rate": 2.2911358119085792e-07,
+      "loss": 0.7496,
+      "step": 14158
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9430130267505186,
+      "learning_rate": 2.2880352331290102e-07,
+      "loss": 0.7876,
+      "step": 14159
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8366244041351973,
+      "learning_rate": 2.2849367046388294e-07,
+      "loss": 0.7836,
+      "step": 14160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4741183868907293,
+      "learning_rate": 2.2818402265711858e-07,
+      "loss": 0.6319,
+      "step": 14161
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3151382871479234,
+      "learning_rate": 2.27874579905914e-07,
+      "loss": 0.6725,
+      "step": 14162
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1367783512901322,
+      "learning_rate": 2.2756534222356687e-07,
+      "loss": 0.6907,
+      "step": 14163
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1150968451825347,
+      "learning_rate": 2.2725630962336542e-07,
+      "loss": 0.7103,
+      "step": 14164
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.766566930580808,
+      "learning_rate": 2.2694748211858964e-07,
+      "loss": 0.6345,
+      "step": 14165
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6394451485642065,
+      "learning_rate": 2.2663885972251165e-07,
+      "loss": 0.6998,
+      "step": 14166
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8975228454910378,
+      "learning_rate": 2.2633044244839143e-07,
+      "loss": 0.8553,
+      "step": 14167
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.916691121708568,
+      "learning_rate": 2.2602223030948445e-07,
+      "loss": 0.7281,
+      "step": 14168
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7157668436733848,
+      "learning_rate": 2.2571422331903458e-07,
+      "loss": 0.8463,
+      "step": 14169
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0938312424688885,
+      "learning_rate": 2.2540642149027735e-07,
+      "loss": 0.6337,
+      "step": 14170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1716982026675558,
+      "learning_rate": 2.2509882483643942e-07,
+      "loss": 0.6226,
+      "step": 14171
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5152696844722477,
+      "learning_rate": 2.2479143337073904e-07,
+      "loss": 0.6268,
+      "step": 14172
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5504555939625944,
+      "learning_rate": 2.2448424710638684e-07,
+      "loss": 0.6493,
+      "step": 14173
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6992248398060825,
+      "learning_rate": 2.2417726605658164e-07,
+      "loss": 0.5492,
+      "step": 14174
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1999871394638895,
+      "learning_rate": 2.2387049023451458e-07,
+      "loss": 0.6424,
+      "step": 14175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.377740708810333,
+      "learning_rate": 2.2356391965337064e-07,
+      "loss": 0.8302,
+      "step": 14176
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.090246145852581,
+      "learning_rate": 2.2325755432632267e-07,
+      "loss": 0.7526,
+      "step": 14177
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6363254926144775,
+      "learning_rate": 2.2295139426653622e-07,
+      "loss": 0.6543,
+      "step": 14178
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.130188698419129,
+      "learning_rate": 2.226454394871669e-07,
+      "loss": 0.7366,
+      "step": 14179
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0927004019847322,
+      "learning_rate": 2.2233969000136357e-07,
+      "loss": 0.6051,
+      "step": 14180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4226791548544222,
+      "learning_rate": 2.2203414582226357e-07,
+      "loss": 0.7013,
+      "step": 14181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8020678079905073,
+      "learning_rate": 2.2172880696299692e-07,
+      "loss": 0.6881,
+      "step": 14182
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7626069998630889,
+      "learning_rate": 2.2142367343668592e-07,
+      "loss": 0.6812,
+      "step": 14183
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9430325231774135,
+      "learning_rate": 2.2111874525644228e-07,
+      "loss": 0.7179,
+      "step": 14184
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7182174188468127,
+      "learning_rate": 2.2081402243536833e-07,
+      "loss": 0.7622,
+      "step": 14185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0326912805717106,
+      "learning_rate": 2.2050950498656078e-07,
+      "loss": 0.7018,
+      "step": 14186
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5506112572555422,
+      "learning_rate": 2.2020519292310361e-07,
+      "loss": 0.6147,
+      "step": 14187
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2982934872485692,
+      "learning_rate": 2.199010862580747e-07,
+      "loss": 0.6238,
+      "step": 14188
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8848125484333063,
+      "learning_rate": 2.1959718500454196e-07,
+      "loss": 0.7025,
+      "step": 14189
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7508646841246722,
+      "learning_rate": 2.1929348917556436e-07,
+      "loss": 0.6903,
+      "step": 14190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7573039411820213,
+      "learning_rate": 2.189899987841937e-07,
+      "loss": 0.8278,
+      "step": 14191
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.610134489314608,
+      "learning_rate": 2.186867138434695e-07,
+      "loss": 0.776,
+      "step": 14192
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.995935538699339,
+      "learning_rate": 2.183836343664264e-07,
+      "loss": 0.5831,
+      "step": 14193
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5063349598344755,
+      "learning_rate": 2.1808076036608783e-07,
+      "loss": 0.7074,
+      "step": 14194
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5632811756254075,
+      "learning_rate": 2.1777809185546894e-07,
+      "loss": 0.6496,
+      "step": 14195
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.850372970809137,
+      "learning_rate": 2.17475628847576e-07,
+      "loss": 0.7717,
+      "step": 14196
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.661339854677665,
+      "learning_rate": 2.171733713554064e-07,
+      "loss": 0.6987,
+      "step": 14197
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3417151125698348,
+      "learning_rate": 2.1687131939194862e-07,
+      "loss": 0.6694,
+      "step": 14198
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5220147594289495,
+      "learning_rate": 2.165694729701834e-07,
+      "loss": 0.6751,
+      "step": 14199
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.827899347136854,
+      "learning_rate": 2.1626783210308032e-07,
+      "loss": 0.8064,
+      "step": 14200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6522839747639473,
+      "learning_rate": 2.15966396803603e-07,
+      "loss": 0.7435,
+      "step": 14201
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.696362375828108,
+      "learning_rate": 2.1566516708470487e-07,
+      "loss": 0.6433,
+      "step": 14202
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5407935440800238,
+      "learning_rate": 2.1536414295932896e-07,
+      "loss": 0.6049,
+      "step": 14203
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.196208847670163,
+      "learning_rate": 2.1506332444041212e-07,
+      "loss": 0.8038,
+      "step": 14204
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7826046902274921,
+      "learning_rate": 2.1476271154088013e-07,
+      "loss": 0.5875,
+      "step": 14205
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.783233677214367,
+      "learning_rate": 2.1446230427365323e-07,
+      "loss": 0.7282,
+      "step": 14206
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.067914534170073,
+      "learning_rate": 2.1416210265163828e-07,
+      "loss": 0.7355,
+      "step": 14207
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7140835808363684,
+      "learning_rate": 2.1386210668773554e-07,
+      "loss": 0.7265,
+      "step": 14208
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.033347337000365,
+      "learning_rate": 2.1356231639483917e-07,
+      "loss": 0.7182,
+      "step": 14209
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.689685250815345,
+      "learning_rate": 2.1326273178582822e-07,
+      "loss": 0.6587,
+      "step": 14210
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8367389449594533,
+      "learning_rate": 2.1296335287357918e-07,
+      "loss": 0.6936,
+      "step": 14211
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7522235647051865,
+      "learning_rate": 2.1266417967095665e-07,
+      "loss": 0.6601,
+      "step": 14212
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.790068658933856,
+      "learning_rate": 2.123652121908154e-07,
+      "loss": 0.7049,
+      "step": 14213
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7712821622964054,
+      "learning_rate": 2.1206645044600404e-07,
+      "loss": 0.6324,
+      "step": 14214
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7321145463353442,
+      "learning_rate": 2.1176789444936065e-07,
+      "loss": 0.6736,
+      "step": 14215
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1422181996676446,
+      "learning_rate": 2.114695442137149e-07,
+      "loss": 0.619,
+      "step": 14216
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.144536285148104,
+      "learning_rate": 2.1117139975188716e-07,
+      "loss": 0.7816,
+      "step": 14217
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1165645723339175,
+      "learning_rate": 2.1087346107668937e-07,
+      "loss": 0.6288,
+      "step": 14218
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.797664453966137,
+      "learning_rate": 2.1057572820092576e-07,
+      "loss": 0.8014,
+      "step": 14219
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.822365244282465,
+      "learning_rate": 2.1027820113738884e-07,
+      "loss": 0.743,
+      "step": 14220
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2543854144297697,
+      "learning_rate": 2.099808798988656e-07,
+      "loss": 0.8226,
+      "step": 14221
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3798629414315,
+      "learning_rate": 2.0968376449813198e-07,
+      "loss": 0.6384,
+      "step": 14222
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8372812651516615,
+      "learning_rate": 2.093868549479544e-07,
+      "loss": 0.6808,
+      "step": 14223
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.902847962841981,
+      "learning_rate": 2.0909015126109488e-07,
+      "loss": 0.5673,
+      "step": 14224
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7002540304148657,
+      "learning_rate": 2.0879365345030044e-07,
+      "loss": 0.7258,
+      "step": 14225
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.7927617292392295,
+      "learning_rate": 2.0849736152831256e-07,
+      "loss": 0.6912,
+      "step": 14226
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0565034617136917,
+      "learning_rate": 2.0820127550786551e-07,
+      "loss": 0.6554,
+      "step": 14227
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6410284050695778,
+      "learning_rate": 2.0790539540168074e-07,
+      "loss": 0.7509,
+      "step": 14228
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0940896616138565,
+      "learning_rate": 2.0760972122247425e-07,
+      "loss": 0.762,
+      "step": 14229
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5174916223754022,
+      "learning_rate": 2.0731425298295083e-07,
+      "loss": 0.6628,
+      "step": 14230
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7721059387361235,
+      "learning_rate": 2.070189906958081e-07,
+      "loss": 0.6603,
+      "step": 14231
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1738442518128647,
+      "learning_rate": 2.0672393437373372e-07,
+      "loss": 0.6537,
+      "step": 14232
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8969983686289995,
+      "learning_rate": 2.06429084029407e-07,
+      "loss": 0.678,
+      "step": 14233
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.809630471867197,
+      "learning_rate": 2.061344396754994e-07,
+      "loss": 0.7343,
+      "step": 14234
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.548484526853518,
+      "learning_rate": 2.058400013246703e-07,
+      "loss": 0.7428,
+      "step": 14235
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6143990545555271,
+      "learning_rate": 2.0554576898957346e-07,
+      "loss": 0.786,
+      "step": 14236
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7675672926581665,
+      "learning_rate": 2.052517426828532e-07,
+      "loss": 0.8437,
+      "step": 14237
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9091718918770877,
+      "learning_rate": 2.0495792241714386e-07,
+      "loss": 0.7393,
+      "step": 14238
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7126609075092507,
+      "learning_rate": 2.04664308205072e-07,
+      "loss": 0.6526,
+      "step": 14239
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.105293171374907,
+      "learning_rate": 2.0437090005925475e-07,
+      "loss": 0.6618,
+      "step": 14240
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8186584033227926,
+      "learning_rate": 2.0407769799229925e-07,
+      "loss": 0.7085,
+      "step": 14241
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1267885126736823,
+      "learning_rate": 2.0378470201680767e-07,
+      "loss": 0.6527,
+      "step": 14242
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.743721110891099,
+      "learning_rate": 2.0349191214536767e-07,
+      "loss": 0.6315,
+      "step": 14243
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0702335021798044,
+      "learning_rate": 2.0319932839056365e-07,
+      "loss": 0.6464,
+      "step": 14244
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4552019929402575,
+      "learning_rate": 2.029069507649678e-07,
+      "loss": 0.7303,
+      "step": 14245
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7053381472893794,
+      "learning_rate": 2.0261477928114282e-07,
+      "loss": 0.6836,
+      "step": 14246
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.015978932378789,
+      "learning_rate": 2.0232281395164587e-07,
+      "loss": 0.5747,
+      "step": 14247
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.808637921091522,
+      "learning_rate": 2.0203105478902252e-07,
+      "loss": 0.7999,
+      "step": 14248
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.08509480338063,
+      "learning_rate": 2.0173950180581047e-07,
+      "loss": 0.654,
+      "step": 14249
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.540407247997902,
+      "learning_rate": 2.0144815501453863e-07,
+      "loss": 0.6834,
+      "step": 14250
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7146723290540151,
+      "learning_rate": 2.0115701442772528e-07,
+      "loss": 0.7139,
+      "step": 14251
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8047886438520886,
+      "learning_rate": 2.0086608005788376e-07,
+      "loss": 0.6812,
+      "step": 14252
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8892678655379422,
+      "learning_rate": 2.0057535191751465e-07,
+      "loss": 0.8161,
+      "step": 14253
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0532679276023662,
+      "learning_rate": 2.002848300191118e-07,
+      "loss": 0.6331,
+      "step": 14254
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6787482150866189,
+      "learning_rate": 1.9999451437515915e-07,
+      "loss": 0.6569,
+      "step": 14255
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1737043895078245,
+      "learning_rate": 1.997044049981328e-07,
+      "loss": 0.7004,
+      "step": 14256
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6224314397662027,
+      "learning_rate": 1.994145019004984e-07,
+      "loss": 0.7356,
+      "step": 14257
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7653192998993346,
+      "learning_rate": 1.9912480509471476e-07,
+      "loss": 0.6536,
+      "step": 14258
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5268831448446551,
+      "learning_rate": 1.988353145932298e-07,
+      "loss": 0.6648,
+      "step": 14259
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7259083182400765,
+      "learning_rate": 1.985460304084852e-07,
+      "loss": 0.8434,
+      "step": 14260
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0604025180709973,
+      "learning_rate": 1.9825695255290934e-07,
+      "loss": 0.6404,
+      "step": 14261
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5677678853754378,
+      "learning_rate": 1.9796808103892783e-07,
+      "loss": 0.7076,
+      "step": 14262
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7319243541758131,
+      "learning_rate": 1.9767941587895246e-07,
+      "loss": 0.7799,
+      "step": 14263
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8129133764083576,
+      "learning_rate": 1.9739095708538714e-07,
+      "loss": 0.8083,
+      "step": 14264
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8199273721101616,
+      "learning_rate": 1.9710270467062865e-07,
+      "loss": 0.6239,
+      "step": 14265
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6372871704448078,
+      "learning_rate": 1.9681465864706372e-07,
+      "loss": 0.7281,
+      "step": 14266
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3868162632563483,
+      "learning_rate": 1.9652681902707028e-07,
+      "loss": 0.6932,
+      "step": 14267
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7674316533187473,
+      "learning_rate": 1.9623918582301726e-07,
+      "loss": 0.7461,
+      "step": 14268
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.580427892624354,
+      "learning_rate": 1.9595175904726481e-07,
+      "loss": 0.6439,
+      "step": 14269
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6994881466037681,
+      "learning_rate": 1.9566453871216417e-07,
+      "loss": 0.6754,
+      "step": 14270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5994257921175994,
+      "learning_rate": 1.953775248300588e-07,
+      "loss": 0.6539,
+      "step": 14271
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.777589458453853,
+      "learning_rate": 1.9509071741328155e-07,
+      "loss": 0.777,
+      "step": 14272
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.855042328196936,
+      "learning_rate": 1.9480411647415708e-07,
+      "loss": 0.7516,
+      "step": 14273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5843932226409545,
+      "learning_rate": 1.9451772202500163e-07,
+      "loss": 0.6226,
+      "step": 14274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5879078531486228,
+      "learning_rate": 1.94231534078122e-07,
+      "loss": 0.7823,
+      "step": 14275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5802683098388561,
+      "learning_rate": 1.9394555264581672e-07,
+      "loss": 0.7512,
+      "step": 14276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5427902414316792,
+      "learning_rate": 1.936597777403737e-07,
+      "loss": 0.7366,
+      "step": 14277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0457086151525847,
+      "learning_rate": 1.9337420937407536e-07,
+      "loss": 0.5596,
+      "step": 14278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8851318099326615,
+      "learning_rate": 1.9308884755919132e-07,
+      "loss": 0.7932,
+      "step": 14279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.517597306369321,
+      "learning_rate": 1.9280369230798568e-07,
+      "loss": 0.669,
+      "step": 14280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2567871204394803,
+      "learning_rate": 1.925187436327114e-07,
+      "loss": 0.8014,
+      "step": 14281
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8015454523178884,
+      "learning_rate": 1.9223400154561313e-07,
+      "loss": 0.7294,
+      "step": 14282
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6622143124796818,
+      "learning_rate": 1.9194946605892773e-07,
+      "loss": 0.6648,
+      "step": 14283
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1795102164951254,
+      "learning_rate": 1.9166513718488155e-07,
+      "loss": 0.5841,
+      "step": 14284
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7475455607450627,
+      "learning_rate": 1.9138101493569428e-07,
+      "loss": 0.8114,
+      "step": 14285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6076212245170178,
+      "learning_rate": 1.9109709932357333e-07,
+      "loss": 0.6662,
+      "step": 14286
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8208291405616277,
+      "learning_rate": 1.9081339036071956e-07,
+      "loss": 0.7194,
+      "step": 14287
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8999901785815552,
+      "learning_rate": 1.9052988805932538e-07,
+      "loss": 0.6974,
+      "step": 14288
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9783307036931872,
+      "learning_rate": 1.902465924315733e-07,
+      "loss": 0.7615,
+      "step": 14289
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4843929518275871,
+      "learning_rate": 1.899635034896369e-07,
+      "loss": 0.6599,
+      "step": 14290
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7270011710870146,
+      "learning_rate": 1.8968062124568142e-07,
+      "loss": 0.5864,
+      "step": 14291
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5696691678043408,
+      "learning_rate": 1.8939794571186332e-07,
+      "loss": 0.5494,
+      "step": 14292
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.71613826418905,
+      "learning_rate": 1.8911547690032838e-07,
+      "loss": 0.6381,
+      "step": 14293
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7099887581515842,
+      "learning_rate": 1.8883321482321583e-07,
+      "loss": 0.6285,
+      "step": 14294
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7133696731135062,
+      "learning_rate": 1.8855115949265535e-07,
+      "loss": 0.636,
+      "step": 14295
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6620390173727346,
+      "learning_rate": 1.882693109207684e-07,
+      "loss": 0.8625,
+      "step": 14296
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6294856971277225,
+      "learning_rate": 1.879876691196636e-07,
+      "loss": 0.742,
+      "step": 14297
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.007136611898069,
+      "learning_rate": 1.8770623410144628e-07,
+      "loss": 0.7924,
+      "step": 14298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0420537861835513,
+      "learning_rate": 1.8742500587820955e-07,
+      "loss": 0.7245,
+      "step": 14299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.90278419206698,
+      "learning_rate": 1.8714398446203873e-07,
+      "loss": 0.738,
+      "step": 14300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2008533354377673,
+      "learning_rate": 1.8686316986500974e-07,
+      "loss": 0.7202,
+      "step": 14301
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7700868248306614,
+      "learning_rate": 1.86582562099189e-07,
+      "loss": 0.6194,
+      "step": 14302
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9879785087402289,
+      "learning_rate": 1.8630216117663747e-07,
+      "loss": 0.7772,
+      "step": 14303
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6981265726606218,
+      "learning_rate": 1.86021967109401e-07,
+      "loss": 0.745,
+      "step": 14304
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.717218624523267,
+      "learning_rate": 1.8574197990952225e-07,
+      "loss": 0.6384,
+      "step": 14305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.157717919184838,
+      "learning_rate": 1.8546219958903267e-07,
+      "loss": 0.7816,
+      "step": 14306
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6399605700090805,
+      "learning_rate": 1.8518262615995541e-07,
+      "loss": 0.7619,
+      "step": 14307
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6658886217949507,
+      "learning_rate": 1.8490325963430368e-07,
+      "loss": 0.8423,
+      "step": 14308
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.933814574597151,
+      "learning_rate": 1.8462410002408228e-07,
+      "loss": 0.8444,
+      "step": 14309
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.552500947078134,
+      "learning_rate": 1.843451473412877e-07,
+      "loss": 0.6499,
+      "step": 14310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5068507646027336,
+      "learning_rate": 1.8406640159790767e-07,
+      "loss": 0.6092,
+      "step": 14311
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9267295781869593,
+      "learning_rate": 1.837878628059192e-07,
+      "loss": 0.6884,
+      "step": 14312
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.832515333993335,
+      "learning_rate": 1.835095309772933e-07,
+      "loss": 0.7727,
+      "step": 14313
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.452246306122695,
+      "learning_rate": 1.8323140612399038e-07,
+      "loss": 0.7743,
+      "step": 14314
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0993030089949642,
+      "learning_rate": 1.829534882579598e-07,
+      "loss": 0.6988,
+      "step": 14315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8239801618011793,
+      "learning_rate": 1.8267577739114695e-07,
+      "loss": 0.7149,
+      "step": 14316
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.694457763791934,
+      "learning_rate": 1.8239827353548457e-07,
+      "loss": 0.6425,
+      "step": 14317
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7412890671982804,
+      "learning_rate": 1.8212097670289807e-07,
+      "loss": 0.7141,
+      "step": 14318
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.921735939834853,
+      "learning_rate": 1.8184388690530242e-07,
+      "loss": 0.6915,
+      "step": 14319
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6980370686424622,
+      "learning_rate": 1.8156700415460582e-07,
+      "loss": 0.675,
+      "step": 14320
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6428888734274716,
+      "learning_rate": 1.812903284627071e-07,
+      "loss": 0.6199,
+      "step": 14321
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.758175226815686,
+      "learning_rate": 1.8101385984149343e-07,
+      "loss": 0.6618,
+      "step": 14322
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.601035583424121,
+      "learning_rate": 1.8073759830284754e-07,
+      "loss": 0.6612,
+      "step": 14323
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4824526012248693,
+      "learning_rate": 1.804615438586399e-07,
+      "loss": 0.5753,
+      "step": 14324
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7817192140996085,
+      "learning_rate": 1.801856965207338e-07,
+      "loss": 0.6455,
+      "step": 14325
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6189295567163753,
+      "learning_rate": 1.79910056300982e-07,
+      "loss": 0.7738,
+      "step": 14326
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9722600867830711,
+      "learning_rate": 1.796346232112306e-07,
+      "loss": 0.6082,
+      "step": 14327
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.312706922089254,
+      "learning_rate": 1.7935939726331453e-07,
+      "loss": 0.6534,
+      "step": 14328
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6396737382850046,
+      "learning_rate": 1.7908437846906158e-07,
+      "loss": 0.8,
+      "step": 14329
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5276202608780405,
+      "learning_rate": 1.7880956684028894e-07,
+      "loss": 0.7077,
+      "step": 14330
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9199394169597213,
+      "learning_rate": 1.7853496238880718e-07,
+      "loss": 0.8144,
+      "step": 14331
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.33677089485638,
+      "learning_rate": 1.7826056512641633e-07,
+      "loss": 0.7762,
+      "step": 14332
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9527457638465677,
+      "learning_rate": 1.7798637506490745e-07,
+      "loss": 0.732,
+      "step": 14333
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1844638814368174,
+      "learning_rate": 1.7771239221606285e-07,
+      "loss": 0.7671,
+      "step": 14334
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1200007593435153,
+      "learning_rate": 1.7743861659165695e-07,
+      "loss": 0.6645,
+      "step": 14335
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9606043845776462,
+      "learning_rate": 1.7716504820345427e-07,
+      "loss": 0.6575,
+      "step": 14336
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6114955114350464,
+      "learning_rate": 1.768916870632098e-07,
+      "loss": 0.8076,
+      "step": 14337
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6616067509766586,
+      "learning_rate": 1.7661853318267142e-07,
+      "loss": 0.669,
+      "step": 14338
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.896557450851478,
+      "learning_rate": 1.7634558657357748e-07,
+      "loss": 0.674,
+      "step": 14339
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.944565379903548,
+      "learning_rate": 1.7607284724765583e-07,
+      "loss": 0.7081,
+      "step": 14340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5217242531320325,
+      "learning_rate": 1.7580031521662765e-07,
+      "loss": 0.6907,
+      "step": 14341
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.632830952567794,
+      "learning_rate": 1.7552799049220415e-07,
+      "loss": 0.7534,
+      "step": 14342
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7450724493573293,
+      "learning_rate": 1.752558730860876e-07,
+      "loss": 0.6693,
+      "step": 14343
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4681603173159825,
+      "learning_rate": 1.7498396300997146e-07,
+      "loss": 0.6119,
+      "step": 14344
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6957521420905883,
+      "learning_rate": 1.747122602755391e-07,
+      "loss": 0.6561,
+      "step": 14345
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5158617997623653,
+      "learning_rate": 1.7444076489446958e-07,
+      "loss": 0.83,
+      "step": 14346
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5417861119628788,
+      "learning_rate": 1.7416947687842633e-07,
+      "loss": 0.6708,
+      "step": 14347
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5095329941498283,
+      "learning_rate": 1.738983962390678e-07,
+      "loss": 0.7281,
+      "step": 14348
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1220101394967683,
+      "learning_rate": 1.736275229880441e-07,
+      "loss": 0.8123,
+      "step": 14349
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.248092387123723,
+      "learning_rate": 1.733568571369948e-07,
+      "loss": 0.8362,
+      "step": 14350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7347318235186764,
+      "learning_rate": 1.7308639869755062e-07,
+      "loss": 0.881,
+      "step": 14351
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.944840289661672,
+      "learning_rate": 1.728161476813345e-07,
+      "loss": 0.668,
+      "step": 14352
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2452446427395336,
+      "learning_rate": 1.7254610409995875e-07,
+      "loss": 0.7027,
+      "step": 14353
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.82133975161061,
+      "learning_rate": 1.7227626796502807e-07,
+      "loss": 0.7058,
+      "step": 14354
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.926818967055056,
+      "learning_rate": 1.720066392881381e-07,
+      "loss": 0.6791,
+      "step": 14355
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7131410220474923,
+      "learning_rate": 1.7173721808087573e-07,
+      "loss": 0.7535,
+      "step": 14356
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7939376774836862,
+      "learning_rate": 1.7146800435481837e-07,
+      "loss": 0.5718,
+      "step": 14357
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6542156425490044,
+      "learning_rate": 1.7119899812153394e-07,
+      "loss": 0.8066,
+      "step": 14358
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.133031251097968,
+      "learning_rate": 1.7093019939258327e-07,
+      "loss": 0.7192,
+      "step": 14359
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8177635069117033,
+      "learning_rate": 1.706616081795165e-07,
+      "loss": 0.7984,
+      "step": 14360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6353858257444422,
+      "learning_rate": 1.7039322449387664e-07,
+      "loss": 0.656,
+      "step": 14361
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2466096824827244,
+      "learning_rate": 1.7012504834719556e-07,
+      "loss": 0.5966,
+      "step": 14362
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8347205967962137,
+      "learning_rate": 1.6985707975099742e-07,
+      "loss": 0.7092,
+      "step": 14363
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6667985039815902,
+      "learning_rate": 1.6958931871679908e-07,
+      "loss": 0.7219,
+      "step": 14364
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1498891763550185,
+      "learning_rate": 1.6932176525610465e-07,
+      "loss": 0.733,
+      "step": 14365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.576859645191715,
+      "learning_rate": 1.6905441938041222e-07,
+      "loss": 0.7098,
+      "step": 14366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8421877599333052,
+      "learning_rate": 1.687872811012109e-07,
+      "loss": 0.6505,
+      "step": 14367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6050908256211371,
+      "learning_rate": 1.685203504299804e-07,
+      "loss": 0.7383,
+      "step": 14368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4534992856948148,
+      "learning_rate": 1.6825362737818985e-07,
+      "loss": 0.7191,
+      "step": 14369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.82348043028502,
+      "learning_rate": 1.6798711195730233e-07,
+      "loss": 0.8546,
+      "step": 14370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6413521729724823,
+      "learning_rate": 1.677208041787698e-07,
+      "loss": 0.6545,
+      "step": 14371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.714662814519082,
+      "learning_rate": 1.6745470405403697e-07,
+      "loss": 0.6626,
+      "step": 14372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9638569014366227,
+      "learning_rate": 1.6718881159453693e-07,
+      "loss": 0.5848,
+      "step": 14373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6703500861730678,
+      "learning_rate": 1.6692312681169775e-07,
+      "loss": 0.7352,
+      "step": 14374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7174923183270896,
+      "learning_rate": 1.6665764971693587e-07,
+      "loss": 0.7578,
+      "step": 14375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9046799962622865,
+      "learning_rate": 1.6639238032165883e-07,
+      "loss": 0.7082,
+      "step": 14376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5575815975451088,
+      "learning_rate": 1.6612731863726638e-07,
+      "loss": 0.6899,
+      "step": 14377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.885256335707503,
+      "learning_rate": 1.6586246467514833e-07,
+      "loss": 0.6997,
+      "step": 14378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7167435648425433,
+      "learning_rate": 1.6559781844668666e-07,
+      "loss": 0.7804,
+      "step": 14379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5831191453198545,
+      "learning_rate": 1.6533337996325337e-07,
+      "loss": 0.725,
+      "step": 14380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3792208101708465,
+      "learning_rate": 1.650691492362122e-07,
+      "loss": 0.7267,
+      "step": 14381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5260538305242415,
+      "learning_rate": 1.6480512627691792e-07,
+      "loss": 0.645,
+      "step": 14382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7889996851878351,
+      "learning_rate": 1.6454131109671534e-07,
+      "loss": 0.6292,
+      "step": 14383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1058903720674722,
+      "learning_rate": 1.6427770370694208e-07,
+      "loss": 0.5637,
+      "step": 14384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7921291764016536,
+      "learning_rate": 1.6401430411892572e-07,
+      "loss": 0.7274,
+      "step": 14385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.150078941850727,
+      "learning_rate": 1.63751112343985e-07,
+      "loss": 0.7204,
+      "step": 14386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9517392393334367,
+      "learning_rate": 1.6348812839342976e-07,
+      "loss": 0.7225,
+      "step": 14387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5623511105752688,
+      "learning_rate": 1.6322535227856152e-07,
+      "loss": 0.6831,
+      "step": 14388
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6897076266084878,
+      "learning_rate": 1.6296278401067122e-07,
+      "loss": 0.6926,
+      "step": 14389
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9227386022651896,
+      "learning_rate": 1.6270042360104377e-07,
+      "loss": 0.6918,
+      "step": 14390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.789427230270893,
+      "learning_rate": 1.624382710609512e-07,
+      "loss": 0.7068,
+      "step": 14391
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2059191232335842,
+      "learning_rate": 1.621763264016607e-07,
+      "loss": 0.6339,
+      "step": 14392
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7381540775088198,
+      "learning_rate": 1.6191458963442762e-07,
+      "loss": 0.7132,
+      "step": 14393
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5326324614660718,
+      "learning_rate": 1.6165306077049969e-07,
+      "loss": 0.6733,
+      "step": 14394
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.498106906627794,
+      "learning_rate": 1.6139173982111567e-07,
+      "loss": 0.6724,
+      "step": 14395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.772726174216742,
+      "learning_rate": 1.611306267975038e-07,
+      "loss": 0.7931,
+      "step": 14396
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.606450394037533,
+      "learning_rate": 1.6086972171088678e-07,
+      "loss": 0.7261,
+      "step": 14397
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.696646768366796,
+      "learning_rate": 1.606090245724745e-07,
+      "loss": 0.6536,
+      "step": 14398
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4585208974558863,
+      "learning_rate": 1.603485353934703e-07,
+      "loss": 0.7591,
+      "step": 14399
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7745945673722734,
+      "learning_rate": 1.6008825418506846e-07,
+      "loss": 0.8572,
+      "step": 14400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1002785810819309,
+      "learning_rate": 1.5982818095845286e-07,
+      "loss": 0.5861,
+      "step": 14401
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6891447258494743,
+      "learning_rate": 1.5956831572480013e-07,
+      "loss": 0.7781,
+      "step": 14402
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3637677097153458,
+      "learning_rate": 1.5930865849527744e-07,
+      "loss": 0.626,
+      "step": 14403
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6942945007327468,
+      "learning_rate": 1.5904920928104196e-07,
+      "loss": 0.7197,
+      "step": 14404
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8554529789687595,
+      "learning_rate": 1.5878996809324366e-07,
+      "loss": 0.6503,
+      "step": 14405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3250229782391965,
+      "learning_rate": 1.5853093494302195e-07,
+      "loss": 0.6524,
+      "step": 14406
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.889996544629961,
+      "learning_rate": 1.582721098415091e-07,
+      "loss": 0.8787,
+      "step": 14407
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.638510251358823,
+      "learning_rate": 1.5801349279982726e-07,
+      "loss": 0.6349,
+      "step": 14408
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0279037879439041,
+      "learning_rate": 1.577550838290881e-07,
+      "loss": 0.6252,
+      "step": 14409
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8642576919444789,
+      "learning_rate": 1.5749688294039778e-07,
+      "loss": 0.7536,
+      "step": 14410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8683006097036854,
+      "learning_rate": 1.5723889014485128e-07,
+      "loss": 0.7959,
+      "step": 14411
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8355989321924941,
+      "learning_rate": 1.5698110545353474e-07,
+      "loss": 0.7595,
+      "step": 14412
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.919561637647557,
+      "learning_rate": 1.567235288775265e-07,
+      "loss": 0.787,
+      "step": 14413
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8302273144213501,
+      "learning_rate": 1.564661604278944e-07,
+      "loss": 0.7361,
+      "step": 14414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7623474274608106,
+      "learning_rate": 1.5620900011569906e-07,
+      "loss": 0.8013,
+      "step": 14415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.140190358115246,
+      "learning_rate": 1.5595204795199047e-07,
+      "loss": 0.856,
+      "step": 14416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.090228076792378,
+      "learning_rate": 1.5569530394780986e-07,
+      "loss": 0.7234,
+      "step": 14417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2722783323637703,
+      "learning_rate": 1.5543876811419222e-07,
+      "loss": 0.6678,
+      "step": 14418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7317846524641474,
+      "learning_rate": 1.5518244046215936e-07,
+      "loss": 0.7223,
+      "step": 14419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8085860327831984,
+      "learning_rate": 1.5492632100272686e-07,
+      "loss": 0.7013,
+      "step": 14420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8525402955244927,
+      "learning_rate": 1.546704097469015e-07,
+      "loss": 0.7961,
+      "step": 14421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4435772872180204,
+      "learning_rate": 1.544147067056795e-07,
+      "loss": 0.6112,
+      "step": 14422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.701922269359028,
+      "learning_rate": 1.5415921189004922e-07,
+      "loss": 0.823,
+      "step": 14423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7431720531525747,
+      "learning_rate": 1.539039253109892e-07,
+      "loss": 0.6997,
+      "step": 14424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.780264212028311,
+      "learning_rate": 1.5364884697947113e-07,
+      "loss": 0.663,
+      "step": 14425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7422083714112897,
+      "learning_rate": 1.533939769064563e-07,
+      "loss": 0.6799,
+      "step": 14426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7802051738667752,
+      "learning_rate": 1.5313931510289482e-07,
+      "loss": 0.7443,
+      "step": 14427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3225084213432117,
+      "learning_rate": 1.528848615797318e-07,
+      "loss": 0.6045,
+      "step": 14428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6314670974348449,
+      "learning_rate": 1.526306163479019e-07,
+      "loss": 0.709,
+      "step": 14429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5346314978408504,
+      "learning_rate": 1.523765794183296e-07,
+      "loss": 0.5942,
+      "step": 14430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6119599113784058,
+      "learning_rate": 1.521227508019324e-07,
+      "loss": 0.6776,
+      "step": 14431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7121172954920891,
+      "learning_rate": 1.518691305096165e-07,
+      "loss": 0.6858,
+      "step": 14432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.661439248286549,
+      "learning_rate": 1.5161571855228264e-07,
+      "loss": 0.826,
+      "step": 14433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1257281583013317,
+      "learning_rate": 1.5136251494081822e-07,
+      "loss": 0.6929,
+      "step": 14434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2422020813454386,
+      "learning_rate": 1.5110951968610564e-07,
+      "loss": 0.7677,
+      "step": 14435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.664451981456185,
+      "learning_rate": 1.5085673279901626e-07,
+      "loss": 0.6664,
+      "step": 14436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8946386730413636,
+      "learning_rate": 1.5060415429041186e-07,
+      "loss": 0.6909,
+      "step": 14437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.037297549823884,
+      "learning_rate": 1.503517841711477e-07,
+      "loss": 0.7705,
+      "step": 14438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2460901563322864,
+      "learning_rate": 1.5009962245206845e-07,
+      "loss": 0.7581,
+      "step": 14439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8176986562455848,
+      "learning_rate": 1.4984766914400928e-07,
+      "loss": 0.684,
+      "step": 14440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7242303469337237,
+      "learning_rate": 1.4959592425779768e-07,
+      "loss": 0.58,
+      "step": 14441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9876997979874167,
+      "learning_rate": 1.4934438780425108e-07,
+      "loss": 0.6945,
+      "step": 14442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.752226881975699,
+      "learning_rate": 1.4909305979417976e-07,
+      "loss": 0.6314,
+      "step": 14443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7674399094981703,
+      "learning_rate": 1.488419402383834e-07,
+      "loss": 0.6771,
+      "step": 14444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7628698369692821,
+      "learning_rate": 1.4859102914765278e-07,
+      "loss": 0.6469,
+      "step": 14445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5835319912060963,
+      "learning_rate": 1.4834032653276986e-07,
+      "loss": 0.6941,
+      "step": 14446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4055863325410365,
+      "learning_rate": 1.4808983240450881e-07,
+      "loss": 0.7378,
+      "step": 14447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.674035046133028,
+      "learning_rate": 1.4783954677363376e-07,
+      "loss": 0.7575,
+      "step": 14448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6765290013218928,
+      "learning_rate": 1.4758946965089894e-07,
+      "loss": 0.7178,
+      "step": 14449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6774162776443122,
+      "learning_rate": 1.473396010470518e-07,
+      "loss": 0.6496,
+      "step": 14450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8441843191627598,
+      "learning_rate": 1.470899409728299e-07,
+      "loss": 0.6611,
+      "step": 14451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.038947758528806,
+      "learning_rate": 1.4684048943896022e-07,
+      "loss": 0.6761,
+      "step": 14452
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8747385025759986,
+      "learning_rate": 1.4659124645616418e-07,
+      "loss": 0.6328,
+      "step": 14453
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7254838074963674,
+      "learning_rate": 1.4634221203515097e-07,
+      "loss": 0.7121,
+      "step": 14454
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8382030788012391,
+      "learning_rate": 1.4609338618662318e-07,
+      "loss": 0.7177,
+      "step": 14455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6894055244449737,
+      "learning_rate": 1.458447689212722e-07,
+      "loss": 0.6451,
+      "step": 14456
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5910543392645065,
+      "learning_rate": 1.455963602497823e-07,
+      "loss": 0.665,
+      "step": 14457
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.53856811458033,
+      "learning_rate": 1.4534816018282882e-07,
+      "loss": 0.7797,
+      "step": 14458
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5411451449930615,
+      "learning_rate": 1.4510016873107657e-07,
+      "loss": 0.7662,
+      "step": 14459
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6944695984779952,
+      "learning_rate": 1.4485238590518147e-07,
+      "loss": 0.6515,
+      "step": 14460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.659686715639601,
+      "learning_rate": 1.4460481171579334e-07,
+      "loss": 0.7116,
+      "step": 14461
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5964124816882226,
+      "learning_rate": 1.4435744617354975e-07,
+      "loss": 0.6706,
+      "step": 14462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4762480406707152,
+      "learning_rate": 1.4411028928908054e-07,
+      "loss": 0.7486,
+      "step": 14463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7397196366904897,
+      "learning_rate": 1.4386334107300727e-07,
+      "loss": 0.6634,
+      "step": 14464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9083521010784831,
+      "learning_rate": 1.4361660153594137e-07,
+      "loss": 0.562,
+      "step": 14465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.775804472762118,
+      "learning_rate": 1.4337007068848607e-07,
+      "loss": 0.7031,
+      "step": 14466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4330499879879968,
+      "learning_rate": 1.4312374854123456e-07,
+      "loss": 0.6664,
+      "step": 14467
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8084993040474044,
+      "learning_rate": 1.4287763510477225e-07,
+      "loss": 0.6895,
+      "step": 14468
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8673857247144703,
+      "learning_rate": 1.4263173038967627e-07,
+      "loss": 0.7156,
+      "step": 14469
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9325969949899284,
+      "learning_rate": 1.42386034406512e-07,
+      "loss": 0.8511,
+      "step": 14470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1838995539759343,
+      "learning_rate": 1.421405471658388e-07,
+      "loss": 0.6372,
+      "step": 14471
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6838115442846948,
+      "learning_rate": 1.4189526867820491e-07,
+      "loss": 0.7035,
+      "step": 14472
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7477340469161844,
+      "learning_rate": 1.4165019895415133e-07,
+      "loss": 0.712,
+      "step": 14473
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5653145219516973,
+      "learning_rate": 1.4140533800420853e-07,
+      "loss": 0.5549,
+      "step": 14474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7505672149670806,
+      "learning_rate": 1.4116068583889865e-07,
+      "loss": 0.7016,
+      "step": 14475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7287334894324256,
+      "learning_rate": 1.409162424687366e-07,
+      "loss": 0.6865,
+      "step": 14476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.149731824017318,
+      "learning_rate": 1.4067200790422454e-07,
+      "loss": 0.7765,
+      "step": 14477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6159421884075302,
+      "learning_rate": 1.4042798215585796e-07,
+      "loss": 0.6277,
+      "step": 14478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6384098985948998,
+      "learning_rate": 1.401841652341246e-07,
+      "loss": 0.6078,
+      "step": 14479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4877270474184598,
+      "learning_rate": 1.3994055714950105e-07,
+      "loss": 0.6137,
+      "step": 14480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8512790256851268,
+      "learning_rate": 1.3969715791245565e-07,
+      "loss": 0.5787,
+      "step": 14481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0738883090534994,
+      "learning_rate": 1.3945396753344776e-07,
+      "loss": 0.6392,
+      "step": 14482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4949269853184723,
+      "learning_rate": 1.3921098602292793e-07,
+      "loss": 0.6889,
+      "step": 14483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.022709536814429,
+      "learning_rate": 1.389682133913378e-07,
+      "loss": 0.6314,
+      "step": 14484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.606345958903866,
+      "learning_rate": 1.3872564964910905e-07,
+      "loss": 0.6151,
+      "step": 14485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8245056122280605,
+      "learning_rate": 1.384832948066661e-07,
+      "loss": 0.6344,
+      "step": 14486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6641416562362896,
+      "learning_rate": 1.382411488744234e-07,
+      "loss": 0.552,
+      "step": 14487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9799880067095694,
+      "learning_rate": 1.3799921186278597e-07,
+      "loss": 0.7807,
+      "step": 14488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4968499251397644,
+      "learning_rate": 1.3775748378215047e-07,
+      "loss": 0.6162,
+      "step": 14489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.749796848965866,
+      "learning_rate": 1.3751596464290529e-07,
+      "loss": 0.6864,
+      "step": 14490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6442448000080059,
+      "learning_rate": 1.372746544554282e-07,
+      "loss": 0.6337,
+      "step": 14491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.692977575414849,
+      "learning_rate": 1.3703355323008872e-07,
+      "loss": 0.6265,
+      "step": 14492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8620120436694254,
+      "learning_rate": 1.3679266097724797e-07,
+      "loss": 0.6731,
+      "step": 14493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6134466982355782,
+      "learning_rate": 1.3655197770725826e-07,
+      "loss": 0.7476,
+      "step": 14494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7059292345169592,
+      "learning_rate": 1.363115034304602e-07,
+      "loss": 0.7003,
+      "step": 14495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6980247854608306,
+      "learning_rate": 1.3607123815718993e-07,
+      "loss": 0.6787,
+      "step": 14496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0600596030732519,
+      "learning_rate": 1.358311818977709e-07,
+      "loss": 0.6877,
+      "step": 14497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0154578821018716,
+      "learning_rate": 1.3559133466251874e-07,
+      "loss": 0.6722,
+      "step": 14498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4733932623439396,
+      "learning_rate": 1.3535169646174073e-07,
+      "loss": 0.6778,
+      "step": 14499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9251490166883523,
+      "learning_rate": 1.3511226730573478e-07,
+      "loss": 0.726,
+      "step": 14500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.658514472734709,
+      "learning_rate": 1.3487304720478933e-07,
+      "loss": 0.7116,
+      "step": 14501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.63364640188885,
+      "learning_rate": 1.346340361691839e-07,
+      "loss": 0.744,
+      "step": 14502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7525794249123048,
+      "learning_rate": 1.343952342091892e-07,
+      "loss": 0.6928,
+      "step": 14503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9081196147770119,
+      "learning_rate": 1.3415664133506812e-07,
+      "loss": 0.7262,
+      "step": 14504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9682048503068392,
+      "learning_rate": 1.3391825755707356e-07,
+      "loss": 0.7369,
+      "step": 14505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6874170814561817,
+      "learning_rate": 1.3368008288544843e-07,
+      "loss": 0.6007,
+      "step": 14506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0506691820501284,
+      "learning_rate": 1.3344211733042788e-07,
+      "loss": 0.7438,
+      "step": 14507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6342485499142587,
+      "learning_rate": 1.3320436090223765e-07,
+      "loss": 0.7055,
+      "step": 14508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.721079119778498,
+      "learning_rate": 1.3296681361109564e-07,
+      "loss": 0.7699,
+      "step": 14509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8050703684657678,
+      "learning_rate": 1.327294754672087e-07,
+      "loss": 0.7418,
+      "step": 14510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1473027676326333,
+      "learning_rate": 1.324923464807759e-07,
+      "loss": 0.7069,
+      "step": 14511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6406168763540283,
+      "learning_rate": 1.3225542666198854e-07,
+      "loss": 0.7545,
+      "step": 14512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.693311095985538,
+      "learning_rate": 1.3201871602102512e-07,
+      "loss": 0.6125,
+      "step": 14513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8698307526179883,
+      "learning_rate": 1.3178221456806028e-07,
+      "loss": 0.7776,
+      "step": 14514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6115716920910077,
+      "learning_rate": 1.3154592231325535e-07,
+      "loss": 0.6732,
+      "step": 14515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7968324117226004,
+      "learning_rate": 1.3130983926676445e-07,
+      "loss": 0.7589,
+      "step": 14516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0008625972114653,
+      "learning_rate": 1.310739654387333e-07,
+      "loss": 0.7142,
+      "step": 14517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1195005594786358,
+      "learning_rate": 1.308383008392977e-07,
+      "loss": 0.693,
+      "step": 14518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.494596430238232,
+      "learning_rate": 1.3060284547858403e-07,
+      "loss": 0.6962,
+      "step": 14519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8460137956428475,
+      "learning_rate": 1.3036759936671084e-07,
+      "loss": 0.7241,
+      "step": 14520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4386939779588395,
+      "learning_rate": 1.3013256251378725e-07,
+      "loss": 0.5618,
+      "step": 14521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7250481955182426,
+      "learning_rate": 1.2989773492991298e-07,
+      "loss": 0.6814,
+      "step": 14522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6488220335512507,
+      "learning_rate": 1.2966311662517994e-07,
+      "loss": 0.7407,
+      "step": 14523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6531366365286142,
+      "learning_rate": 1.2942870760966952e-07,
+      "loss": 0.7503,
+      "step": 14524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.58674887428705,
+      "learning_rate": 1.2919450789345477e-07,
+      "loss": 0.6675,
+      "step": 14525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7611155484679288,
+      "learning_rate": 1.289605174865993e-07,
+      "loss": 0.6619,
+      "step": 14526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.92247723848726,
+      "learning_rate": 1.2872673639915955e-07,
+      "loss": 0.6509,
+      "step": 14527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5955489528255256,
+      "learning_rate": 1.2849316464118078e-07,
+      "loss": 0.7415,
+      "step": 14528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4860583088076105,
+      "learning_rate": 1.282598022226994e-07,
+      "loss": 0.6832,
+      "step": 14529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.070471774723826,
+      "learning_rate": 1.2802664915374575e-07,
+      "loss": 0.8,
+      "step": 14530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1579572507415554,
+      "learning_rate": 1.2779370544433567e-07,
+      "loss": 0.6829,
+      "step": 14531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7764348305712234,
+      "learning_rate": 1.275609711044823e-07,
+      "loss": 0.6807,
+      "step": 14532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7574781867212852,
+      "learning_rate": 1.2732844614418482e-07,
+      "loss": 0.698,
+      "step": 14533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0218269154041035,
+      "learning_rate": 1.270961305734364e-07,
+      "loss": 0.8129,
+      "step": 14534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.365253875869033,
+      "learning_rate": 1.2686402440222013e-07,
+      "loss": 0.6311,
+      "step": 14535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5999918275183027,
+      "learning_rate": 1.2663212764050858e-07,
+      "loss": 0.6694,
+      "step": 14536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3492858525317728,
+      "learning_rate": 1.2640044029826937e-07,
+      "loss": 0.6929,
+      "step": 14537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.694560787398974,
+      "learning_rate": 1.261689623854573e-07,
+      "loss": 0.6496,
+      "step": 14538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7572875193390127,
+      "learning_rate": 1.2593769391201827e-07,
+      "loss": 0.8002,
+      "step": 14539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.788125136765289,
+      "learning_rate": 1.257066348878927e-07,
+      "loss": 0.684,
+      "step": 14540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7166568421914583,
+      "learning_rate": 1.254757853230082e-07,
+      "loss": 0.6819,
+      "step": 14541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.735226938579664,
+      "learning_rate": 1.252451452272857e-07,
+      "loss": 0.6148,
+      "step": 14542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1323838087917526,
+      "learning_rate": 1.2501471461063564e-07,
+      "loss": 0.6863,
+      "step": 14543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.136440757111275,
+      "learning_rate": 1.247844934829606e-07,
+      "loss": 0.7034,
+      "step": 14544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6470590892309915,
+      "learning_rate": 1.2455448185415386e-07,
+      "loss": 0.6761,
+      "step": 14545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7206778853173101,
+      "learning_rate": 1.2432467973409857e-07,
+      "loss": 0.7983,
+      "step": 14546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0868644654977113,
+      "learning_rate": 1.2409508713267072e-07,
+      "loss": 0.7687,
+      "step": 14547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7760342818426087,
+      "learning_rate": 1.238657040597374e-07,
+      "loss": 0.7041,
+      "step": 14548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6114762221773973,
+      "learning_rate": 1.2363653052515302e-07,
+      "loss": 0.7087,
+      "step": 14549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7287818897003642,
+      "learning_rate": 1.234075665387674e-07,
+      "loss": 0.6511,
+      "step": 14550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5768070466360207,
+      "learning_rate": 1.2317881211041993e-07,
+      "loss": 0.6431,
+      "step": 14551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.861590280480963,
+      "learning_rate": 1.2295026724993998e-07,
+      "loss": 0.7579,
+      "step": 14552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9165392682484863,
+      "learning_rate": 1.2272193196714854e-07,
+      "loss": 0.5922,
+      "step": 14553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7342231129322399,
+      "learning_rate": 1.2249380627185781e-07,
+      "loss": 0.6593,
+      "step": 14554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9832355335990903,
+      "learning_rate": 1.2226589017387158e-07,
+      "loss": 0.7168,
+      "step": 14555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.939032409575401,
+      "learning_rate": 1.2203818368298315e-07,
+      "loss": 0.814,
+      "step": 14556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8925395179732365,
+      "learning_rate": 1.2181068680897745e-07,
+      "loss": 0.6488,
+      "step": 14557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9266236352843493,
+      "learning_rate": 1.2158339956163113e-07,
+      "loss": 0.7524,
+      "step": 14558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7185059804296619,
+      "learning_rate": 1.2135632195071133e-07,
+      "loss": 0.755,
+      "step": 14559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7047719270962134,
+      "learning_rate": 1.211294539859753e-07,
+      "loss": 0.7302,
+      "step": 14560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.734723927775591,
+      "learning_rate": 1.2090279567717246e-07,
+      "loss": 0.611,
+      "step": 14561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6431414771111559,
+      "learning_rate": 1.2067634703404218e-07,
+      "loss": 0.6935,
+      "step": 14562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9009668454784079,
+      "learning_rate": 1.204501080663173e-07,
+      "loss": 0.7032,
+      "step": 14563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7555826310991263,
+      "learning_rate": 1.202240787837178e-07,
+      "loss": 0.585,
+      "step": 14564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8308896495065858,
+      "learning_rate": 1.1999825919595808e-07,
+      "loss": 0.7426,
+      "step": 14565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1989733353727654,
+      "learning_rate": 1.1977264931274158e-07,
+      "loss": 0.6749,
+      "step": 14566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.458993777965213,
+      "learning_rate": 1.1954724914376215e-07,
+      "loss": 0.739,
+      "step": 14567
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5898100709177803,
+      "learning_rate": 1.1932205869870762e-07,
+      "loss": 0.7202,
+      "step": 14568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5202766716832337,
+      "learning_rate": 1.1909707798725412e-07,
+      "loss": 0.6452,
+      "step": 14569
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9834383203250512,
+      "learning_rate": 1.1887230701906893e-07,
+      "loss": 0.6495,
+      "step": 14570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8592626171651112,
+      "learning_rate": 1.186477458038121e-07,
+      "loss": 0.7,
+      "step": 14571
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.093239782707676,
+      "learning_rate": 1.1842339435113259e-07,
+      "loss": 0.7655,
+      "step": 14572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8893042349617206,
+      "learning_rate": 1.1819925267067211e-07,
+      "loss": 0.7331,
+      "step": 14573
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.079531861251197,
+      "learning_rate": 1.1797532077206187e-07,
+      "loss": 0.6363,
+      "step": 14574
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5998816110489573,
+      "learning_rate": 1.177515986649247e-07,
+      "loss": 0.7292,
+      "step": 14575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5980296033568233,
+      "learning_rate": 1.1752808635887514e-07,
+      "loss": 0.7092,
+      "step": 14576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2400539678268998,
+      "learning_rate": 1.1730478386351719e-07,
+      "loss": 0.662,
+      "step": 14577
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.458139602316113,
+      "learning_rate": 1.1708169118844759e-07,
+      "loss": 0.7044,
+      "step": 14578
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1494596912118393,
+      "learning_rate": 1.1685880834325203e-07,
+      "loss": 0.7882,
+      "step": 14579
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9434109457772972,
+      "learning_rate": 1.1663613533750839e-07,
+      "loss": 0.6218,
+      "step": 14580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5951589825902057,
+      "learning_rate": 1.1641367218078736e-07,
+      "loss": 0.6892,
+      "step": 14581
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.684787370159744,
+      "learning_rate": 1.1619141888264574e-07,
+      "loss": 0.6339,
+      "step": 14582
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6338774028253484,
+      "learning_rate": 1.1596937545263587e-07,
+      "loss": 0.6962,
+      "step": 14583
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0038061179660283,
+      "learning_rate": 1.1574754190030014e-07,
+      "loss": 0.5156,
+      "step": 14584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.885522328265593,
+      "learning_rate": 1.1552591823516978e-07,
+      "loss": 0.7502,
+      "step": 14585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7665055349753027,
+      "learning_rate": 1.1530450446676889e-07,
+      "loss": 0.5797,
+      "step": 14586
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.61216809541497,
+      "learning_rate": 1.1508330060461204e-07,
+      "loss": 0.7542,
+      "step": 14587
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9757377307498915,
+      "learning_rate": 1.1486230665820552e-07,
+      "loss": 0.6445,
+      "step": 14588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5024036152358298,
+      "learning_rate": 1.1464152263704565e-07,
+      "loss": 0.7159,
+      "step": 14589
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7350623912079783,
+      "learning_rate": 1.1442094855061925e-07,
+      "loss": 0.7657,
+      "step": 14590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7519341393940129,
+      "learning_rate": 1.1420058440840598e-07,
+      "loss": 0.7487,
+      "step": 14591
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.710426849848372,
+      "learning_rate": 1.1398043021987438e-07,
+      "loss": 0.7283,
+      "step": 14592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2723994824312705,
+      "learning_rate": 1.1376048599448519e-07,
+      "loss": 0.7305,
+      "step": 14593
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8897734223562743,
+      "learning_rate": 1.1354075174169088e-07,
+      "loss": 0.727,
+      "step": 14594
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.772565258171332,
+      "learning_rate": 1.1332122747093277e-07,
+      "loss": 0.7124,
+      "step": 14595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7871103810627755,
+      "learning_rate": 1.1310191319164498e-07,
+      "loss": 0.7119,
+      "step": 14596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9614753111942167,
+      "learning_rate": 1.1288280891325165e-07,
+      "loss": 0.6805,
+      "step": 14597
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7170758929544694,
+      "learning_rate": 1.1266391464516802e-07,
+      "loss": 0.7637,
+      "step": 14598
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5675029640289193,
+      "learning_rate": 1.12445230396801e-07,
+      "loss": 0.6627,
+      "step": 14599
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.709383095691629,
+      "learning_rate": 1.1222675617754752e-07,
+      "loss": 0.6687,
+      "step": 14600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5443427161560705,
+      "learning_rate": 1.1200849199679564e-07,
+      "loss": 0.6935,
+      "step": 14601
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0565467928947614,
+      "learning_rate": 1.1179043786392507e-07,
+      "loss": 0.6902,
+      "step": 14602
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6046346003641558,
+      "learning_rate": 1.1157259378830665e-07,
+      "loss": 0.7094,
+      "step": 14603
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.082834352831491,
+      "learning_rate": 1.1135495977930011e-07,
+      "loss": 0.703,
+      "step": 14604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7164819565373979,
+      "learning_rate": 1.1113753584625853e-07,
+      "loss": 0.7482,
+      "step": 14605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1040666120290377,
+      "learning_rate": 1.1092032199852609e-07,
+      "loss": 0.6426,
+      "step": 14606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5834918994062481,
+      "learning_rate": 1.1070331824543534e-07,
+      "loss": 0.642,
+      "step": 14607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5541720585517853,
+      "learning_rate": 1.1048652459631216e-07,
+      "loss": 0.6736,
+      "step": 14608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6083430434971864,
+      "learning_rate": 1.1026994106047296e-07,
+      "loss": 0.775,
+      "step": 14609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0082384244277267,
+      "learning_rate": 1.1005356764722364e-07,
+      "loss": 0.8443,
+      "step": 14610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7223329503612175,
+      "learning_rate": 1.0983740436586343e-07,
+      "loss": 0.7128,
+      "step": 14611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8576206103189035,
+      "learning_rate": 1.0962145122568046e-07,
+      "loss": 0.7773,
+      "step": 14612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.321028088433607,
+      "learning_rate": 1.0940570823595565e-07,
+      "loss": 0.6824,
+      "step": 14613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8802874584064848,
+      "learning_rate": 1.0919017540595933e-07,
+      "loss": 0.6123,
+      "step": 14614
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8863588995161842,
+      "learning_rate": 1.0897485274495357e-07,
+      "loss": 0.8708,
+      "step": 14615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6916292126548513,
+      "learning_rate": 1.0875974026219149e-07,
+      "loss": 0.6784,
+      "step": 14616
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6979077070884894,
+      "learning_rate": 1.0854483796691629e-07,
+      "loss": 0.7551,
+      "step": 14617
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6514970670378604,
+      "learning_rate": 1.0833014586836388e-07,
+      "loss": 0.6256,
+      "step": 14618
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8378049242187635,
+      "learning_rate": 1.0811566397575912e-07,
+      "loss": 0.8103,
+      "step": 14619
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.541475438280303,
+      "learning_rate": 1.0790139229831909e-07,
+      "loss": 0.7175,
+      "step": 14620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0732805721418075,
+      "learning_rate": 1.0768733084525196e-07,
+      "loss": 0.6543,
+      "step": 14621
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.65868914796924,
+      "learning_rate": 1.0747347962575539e-07,
+      "loss": 0.6587,
+      "step": 14622
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9588864223968687,
+      "learning_rate": 1.0725983864901978e-07,
+      "loss": 0.6823,
+      "step": 14623
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6240859110194685,
+      "learning_rate": 1.0704640792422616e-07,
+      "loss": 0.5834,
+      "step": 14624
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0775281706681037,
+      "learning_rate": 1.0683318746054439e-07,
+      "loss": 0.7112,
+      "step": 14625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6034191095507622,
+      "learning_rate": 1.066201772671388e-07,
+      "loss": 0.6592,
+      "step": 14626
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4703970643275877,
+      "learning_rate": 1.0640737735316264e-07,
+      "loss": 0.6368,
+      "step": 14627
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6782094335930957,
+      "learning_rate": 1.0619478772776027e-07,
+      "loss": 0.6862,
+      "step": 14628
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6217270824201515,
+      "learning_rate": 1.0598240840006658e-07,
+      "loss": 0.6854,
+      "step": 14629
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.521743828100136,
+      "learning_rate": 1.0577023937920816e-07,
+      "loss": 0.6495,
+      "step": 14630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1847684209621354,
+      "learning_rate": 1.055582806743033e-07,
+      "loss": 0.691,
+      "step": 14631
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4858900767637697,
+      "learning_rate": 1.0534653229445913e-07,
+      "loss": 0.6863,
+      "step": 14632
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9117453768867907,
+      "learning_rate": 1.0513499424877505e-07,
+      "loss": 0.7499,
+      "step": 14633
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9744977008485581,
+      "learning_rate": 1.0492366654634211e-07,
+      "loss": 0.6877,
+      "step": 14634
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6250874950256557,
+      "learning_rate": 1.0471254919624141e-07,
+      "loss": 0.6213,
+      "step": 14635
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5767547670646875,
+      "learning_rate": 1.0450164220754456e-07,
+      "loss": 0.6212,
+      "step": 14636
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7056499581958224,
+      "learning_rate": 1.0429094558931485e-07,
+      "loss": 0.735,
+      "step": 14637
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.140375713357697,
+      "learning_rate": 1.0408045935060674e-07,
+      "loss": 0.6353,
+      "step": 14638
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5876044726573288,
+      "learning_rate": 1.0387018350046519e-07,
+      "loss": 0.6659,
+      "step": 14639
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8547045914352087,
+      "learning_rate": 1.0366011804792631e-07,
+      "loss": 0.7501,
+      "step": 14640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7338191334328055,
+      "learning_rate": 1.0345026300201622e-07,
+      "loss": 0.6774,
+      "step": 14641
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6722688058636495,
+      "learning_rate": 1.0324061837175436e-07,
+      "loss": 0.6919,
+      "step": 14642
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0115956791473433,
+      "learning_rate": 1.0303118416614798e-07,
+      "loss": 0.6518,
+      "step": 14643
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6184586018973561,
+      "learning_rate": 1.0282196039419823e-07,
+      "loss": 0.7657,
+      "step": 14644
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8173719190279467,
+      "learning_rate": 1.0261294706489511e-07,
+      "loss": 0.7346,
+      "step": 14645
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.588653449844697,
+      "learning_rate": 1.0240414418722088e-07,
+      "loss": 0.7993,
+      "step": 14646
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.838977560348005,
+      "learning_rate": 1.021955517701484e-07,
+      "loss": 0.83,
+      "step": 14647
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.723779464469964,
+      "learning_rate": 1.0198716982264045e-07,
+      "loss": 0.6419,
+      "step": 14648
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.094099695174284,
+      "learning_rate": 1.0177899835365323e-07,
+      "loss": 0.7062,
+      "step": 14649
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.741086318238862,
+      "learning_rate": 1.015710373721307e-07,
+      "loss": 0.7251,
+      "step": 14650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8834233494761248,
+      "learning_rate": 1.0136328688700958e-07,
+      "loss": 0.7222,
+      "step": 14651
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7141513934850694,
+      "learning_rate": 1.0115574690721886e-07,
+      "loss": 0.605,
+      "step": 14652
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4184529554631988,
+      "learning_rate": 1.0094841744167583e-07,
+      "loss": 0.6627,
+      "step": 14653
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6791576371144197,
+      "learning_rate": 1.0074129849928948e-07,
+      "loss": 0.7451,
+      "step": 14654
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6374496956530515,
+      "learning_rate": 1.005343900889616e-07,
+      "loss": 0.6533,
+      "step": 14655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5856510993134385,
+      "learning_rate": 1.0032769221958227e-07,
+      "loss": 0.7633,
+      "step": 14656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5535005737744285,
+      "learning_rate": 1.0012120490003385e-07,
+      "loss": 0.607,
+      "step": 14657
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.767413863386712,
+      "learning_rate": 9.99149281391898e-08,
+      "loss": 0.5317,
+      "step": 14658
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7026295111399048,
+      "learning_rate": 9.970886194591467e-08,
+      "loss": 0.6848,
+      "step": 14659
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.703433532593276,
+      "learning_rate": 9.950300632906363e-08,
+      "loss": 0.9176,
+      "step": 14660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7510177137803886,
+      "learning_rate": 9.92973612974818e-08,
+      "loss": 0.8259,
+      "step": 14661
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6717445698555342,
+      "learning_rate": 9.909192686000713e-08,
+      "loss": 0.6849,
+      "step": 14662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0943801350686972,
+      "learning_rate": 9.888670302546754e-08,
+      "loss": 0.7064,
+      "step": 14663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.047251069712306,
+      "learning_rate": 9.8681689802681e-08,
+      "loss": 0.5756,
+      "step": 14664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1685680747023073,
+      "learning_rate": 9.847688720045878e-08,
+      "loss": 0.688,
+      "step": 14665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5793415563294757,
+      "learning_rate": 9.827229522759996e-08,
+      "loss": 0.7308,
+      "step": 14666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5508819574350408,
+      "learning_rate": 9.806791389289861e-08,
+      "loss": 0.7318,
+      "step": 14667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.673258323090414,
+      "learning_rate": 9.786374320513547e-08,
+      "loss": 0.6758,
+      "step": 14668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6381446327511688,
+      "learning_rate": 9.765978317308522e-08,
+      "loss": 0.6647,
+      "step": 14669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.626125809595372,
+      "learning_rate": 9.745603380551194e-08,
+      "loss": 0.6548,
+      "step": 14670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5754254692458027,
+      "learning_rate": 9.72524951111714e-08,
+      "loss": 0.6865,
+      "step": 14671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5020247067986425,
+      "learning_rate": 9.704916709881052e-08,
+      "loss": 0.6635,
+      "step": 14672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6421985587567605,
+      "learning_rate": 9.684604977716616e-08,
+      "loss": 0.6737,
+      "step": 14673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8290218633483348,
+      "learning_rate": 9.664314315496692e-08,
+      "loss": 0.7334,
+      "step": 14674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5549903639584834,
+      "learning_rate": 9.644044724093249e-08,
+      "loss": 0.7515,
+      "step": 14675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6885166439269974,
+      "learning_rate": 9.6237962043772e-08,
+      "loss": 0.7719,
+      "step": 14676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.566208133549958,
+      "learning_rate": 9.603568757218795e-08,
+      "loss": 0.6113,
+      "step": 14677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7522672364845724,
+      "learning_rate": 9.583362383487283e-08,
+      "loss": 0.7198,
+      "step": 14678
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7922953174647778,
+      "learning_rate": 9.5631770840508e-08,
+      "loss": 0.6485,
+      "step": 14679
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7740900447708117,
+      "learning_rate": 9.543012859776934e-08,
+      "loss": 0.7643,
+      "step": 14680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5892359830252187,
+      "learning_rate": 9.5228697115321e-08,
+      "loss": 0.7594,
+      "step": 14681
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5538194491308703,
+      "learning_rate": 9.502747640181887e-08,
+      "loss": 0.6407,
+      "step": 14682
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6371535607721595,
+      "learning_rate": 9.482646646591042e-08,
+      "loss": 0.7887,
+      "step": 14683
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4755668267420197,
+      "learning_rate": 9.462566731623213e-08,
+      "loss": 0.6367,
+      "step": 14684
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7554752934215891,
+      "learning_rate": 9.442507896141484e-08,
+      "loss": 0.7109,
+      "step": 14685
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.899950147934226,
+      "learning_rate": 9.422470141007667e-08,
+      "loss": 0.7061,
+      "step": 14686
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6004057898337827,
+      "learning_rate": 9.402453467082906e-08,
+      "loss": 0.7812,
+      "step": 14687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5934930537065968,
+      "learning_rate": 9.382457875227346e-08,
+      "loss": 0.7891,
+      "step": 14688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5493026131337468,
+      "learning_rate": 9.36248336630019e-08,
+      "loss": 0.7329,
+      "step": 14689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6803870697859908,
+      "learning_rate": 9.342529941159862e-08,
+      "loss": 0.6446,
+      "step": 14690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7679687739069687,
+      "learning_rate": 9.322597600663785e-08,
+      "loss": 0.6447,
+      "step": 14691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8415595926497705,
+      "learning_rate": 9.3026863456685e-08,
+      "loss": 0.8494,
+      "step": 14692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0334089539893068,
+      "learning_rate": 9.282796177029596e-08,
+      "loss": 0.7672,
+      "step": 14693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1475751124206797,
+      "learning_rate": 9.262927095601782e-08,
+      "loss": 0.5999,
+      "step": 14694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.215732531953298,
+      "learning_rate": 9.243079102238983e-08,
+      "loss": 0.716,
+      "step": 14695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5633854039214374,
+      "learning_rate": 9.22325219779402e-08,
+      "loss": 0.7141,
+      "step": 14696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.556639825853594,
+      "learning_rate": 9.20344638311893e-08,
+      "loss": 0.6304,
+      "step": 14697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0841456100366194,
+      "learning_rate": 9.183661659064869e-08,
+      "loss": 0.6307,
+      "step": 14698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5166674966607587,
+      "learning_rate": 9.163898026481876e-08,
+      "loss": 0.6992,
+      "step": 14699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5945379536562732,
+      "learning_rate": 9.144155486219442e-08,
+      "loss": 0.6183,
+      "step": 14700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6378531068775048,
+      "learning_rate": 9.124434039125774e-08,
+      "loss": 0.7125,
+      "step": 14701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7211121593061458,
+      "learning_rate": 9.104733686048417e-08,
+      "loss": 0.6756,
+      "step": 14702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6759380192417772,
+      "learning_rate": 9.085054427833972e-08,
+      "loss": 0.7076,
+      "step": 14703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7875312488400124,
+      "learning_rate": 9.065396265327986e-08,
+      "loss": 0.752,
+      "step": 14704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.69743151473023,
+      "learning_rate": 9.045759199375393e-08,
+      "loss": 0.8198,
+      "step": 14705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1493887529008924,
+      "learning_rate": 9.02614323081985e-08,
+      "loss": 0.622,
+      "step": 14706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.613468605467621,
+      "learning_rate": 9.006548360504463e-08,
+      "loss": 0.7083,
+      "step": 14707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5904051751365578,
+      "learning_rate": 8.986974589271169e-08,
+      "loss": 0.63,
+      "step": 14708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6225593286432072,
+      "learning_rate": 8.967421917961072e-08,
+      "loss": 0.6498,
+      "step": 14709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5435368311449793,
+      "learning_rate": 8.947890347414556e-08,
+      "loss": 0.7343,
+      "step": 14710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8223311241798381,
+      "learning_rate": 8.928379878470727e-08,
+      "loss": 0.8183,
+      "step": 14711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6693533193091812,
+      "learning_rate": 8.908890511968083e-08,
+      "loss": 0.8158,
+      "step": 14712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0819356930293713,
+      "learning_rate": 8.889422248744173e-08,
+      "loss": 0.7571,
+      "step": 14713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0278881773981907,
+      "learning_rate": 8.869975089635552e-08,
+      "loss": 0.6722,
+      "step": 14714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3396700267081718,
+      "learning_rate": 8.85054903547794e-08,
+      "loss": 0.6562,
+      "step": 14715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5817588259258777,
+      "learning_rate": 8.831144087106003e-08,
+      "loss": 0.6981,
+      "step": 14716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4219223226936857,
+      "learning_rate": 8.811760245353739e-08,
+      "loss": 0.6013,
+      "step": 14717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7302113403369033,
+      "learning_rate": 8.792397511054097e-08,
+      "loss": 0.7509,
+      "step": 14718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8801060669324612,
+      "learning_rate": 8.773055885039072e-08,
+      "loss": 0.7656,
+      "step": 14719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8238781883488004,
+      "learning_rate": 8.753735368139838e-08,
+      "loss": 0.8536,
+      "step": 14720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1534560917557128,
+      "learning_rate": 8.734435961186782e-08,
+      "loss": 0.595,
+      "step": 14721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8361231710094612,
+      "learning_rate": 8.715157665008967e-08,
+      "loss": 0.6449,
+      "step": 14722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7786158022337712,
+      "learning_rate": 8.69590048043506e-08,
+      "loss": 0.7419,
+      "step": 14723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5729442505469982,
+      "learning_rate": 8.676664408292457e-08,
+      "loss": 0.7878,
+      "step": 14724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4669270829986165,
+      "learning_rate": 8.657449449407884e-08,
+      "loss": 0.6567,
+      "step": 14725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0516000648168125,
+      "learning_rate": 8.63825560460696e-08,
+      "loss": 0.7386,
+      "step": 14726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5972343729518788,
+      "learning_rate": 8.619082874714413e-08,
+      "loss": 0.6848,
+      "step": 14727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.699935053697992,
+      "learning_rate": 8.599931260554417e-08,
+      "loss": 0.7151,
+      "step": 14728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7338610933697705,
+      "learning_rate": 8.580800762949704e-08,
+      "loss": 0.7964,
+      "step": 14729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8366689307297295,
+      "learning_rate": 8.561691382722337e-08,
+      "loss": 0.7304,
+      "step": 14730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0277546789914829,
+      "learning_rate": 8.542603120693716e-08,
+      "loss": 0.5508,
+      "step": 14731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.135107301144639,
+      "learning_rate": 8.523535977683905e-08,
+      "loss": 0.606,
+      "step": 14732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0045711311177072,
+      "learning_rate": 8.504489954512363e-08,
+      "loss": 0.7304,
+      "step": 14733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0149651485137463,
+      "learning_rate": 8.485465051997488e-08,
+      "loss": 0.7843,
+      "step": 14734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5764475177623665,
+      "learning_rate": 8.466461270956794e-08,
+      "loss": 0.7946,
+      "step": 14735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7485733491028292,
+      "learning_rate": 8.447478612207072e-08,
+      "loss": 0.8566,
+      "step": 14736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.080115395427779,
+      "learning_rate": 8.428517076563836e-08,
+      "loss": 0.6553,
+      "step": 14737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6429920644037814,
+      "learning_rate": 8.409576664842045e-08,
+      "loss": 0.6582,
+      "step": 14738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6207443193049382,
+      "learning_rate": 8.39065737785566e-08,
+      "loss": 0.7297,
+      "step": 14739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.627040492197516,
+      "learning_rate": 8.371759216417474e-08,
+      "loss": 0.694,
+      "step": 14740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6213972194361215,
+      "learning_rate": 8.352882181339783e-08,
+      "loss": 0.7451,
+      "step": 14741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3512009996838932,
+      "learning_rate": 8.334026273433659e-08,
+      "loss": 0.707,
+      "step": 14742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6409398004230549,
+      "learning_rate": 8.315191493509456e-08,
+      "loss": 0.697,
+      "step": 14743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9184099990716803,
+      "learning_rate": 8.296377842376524e-08,
+      "loss": 0.6549,
+      "step": 14744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8757475572257178,
+      "learning_rate": 8.277585320843273e-08,
+      "loss": 0.6685,
+      "step": 14745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7637357039453658,
+      "learning_rate": 8.258813929717391e-08,
+      "loss": 0.769,
+      "step": 14746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4989516367100657,
+      "learning_rate": 8.240063669805343e-08,
+      "loss": 0.6571,
+      "step": 14747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7846292702936155,
+      "learning_rate": 8.221334541912984e-08,
+      "loss": 0.7599,
+      "step": 14748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.994751083538798,
+      "learning_rate": 8.202626546845172e-08,
+      "loss": 0.6681,
+      "step": 14749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4045086978947476,
+      "learning_rate": 8.183939685405762e-08,
+      "loss": 0.5999,
+      "step": 14750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.823586018053325,
+      "learning_rate": 8.165273958397835e-08,
+      "loss": 0.7627,
+      "step": 14751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.585582958158203,
+      "learning_rate": 8.146629366623415e-08,
+      "loss": 0.7177,
+      "step": 14752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8847999498660508,
+      "learning_rate": 8.128005910883751e-08,
+      "loss": 0.6671,
+      "step": 14753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.557691718522696,
+      "learning_rate": 8.109403591979148e-08,
+      "loss": 0.7049,
+      "step": 14754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6987334862057943,
+      "learning_rate": 8.090822410708965e-08,
+      "loss": 0.8381,
+      "step": 14755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.897797380846139,
+      "learning_rate": 8.072262367871675e-08,
+      "loss": 0.7716,
+      "step": 14756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.16483564735988,
+      "learning_rate": 8.053723464264863e-08,
+      "loss": 0.6416,
+      "step": 14757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.843920150636355,
+      "learning_rate": 8.035205700685167e-08,
+      "loss": 0.7691,
+      "step": 14758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.682637739708094,
+      "learning_rate": 8.016709077928397e-08,
+      "loss": 0.725,
+      "step": 14759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2221372356231497,
+      "learning_rate": 7.998233596789251e-08,
+      "loss": 0.5565,
+      "step": 14760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.618706164609398,
+      "learning_rate": 7.97977925806187e-08,
+      "loss": 0.7355,
+      "step": 14761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6174603714254934,
+      "learning_rate": 7.961346062539066e-08,
+      "loss": 0.7173,
+      "step": 14762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6058986945349034,
+      "learning_rate": 7.942934011013037e-08,
+      "loss": 0.7338,
+      "step": 14763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5248403191349484,
+      "learning_rate": 7.924543104275095e-08,
+      "loss": 0.6588,
+      "step": 14764
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1968099484616386,
+      "learning_rate": 7.906173343115386e-08,
+      "loss": 0.6959,
+      "step": 14765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7042706140074313,
+      "learning_rate": 7.88782472832339e-08,
+      "loss": 0.681,
+      "step": 14766
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6305115109785238,
+      "learning_rate": 7.86949726068753e-08,
+      "loss": 0.6972,
+      "step": 14767
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6692132290083346,
+      "learning_rate": 7.851190940995402e-08,
+      "loss": 0.828,
+      "step": 14768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9479224300001736,
+      "learning_rate": 7.832905770033705e-08,
+      "loss": 0.6566,
+      "step": 14769
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8047814205404134,
+      "learning_rate": 7.814641748588148e-08,
+      "loss": 0.7597,
+      "step": 14770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3430128485211674,
+      "learning_rate": 7.796398877443545e-08,
+      "loss": 0.6001,
+      "step": 14771
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8016178594141703,
+      "learning_rate": 7.778177157383993e-08,
+      "loss": 0.8131,
+      "step": 14772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7991030065413187,
+      "learning_rate": 7.759976589192309e-08,
+      "loss": 0.6356,
+      "step": 14773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5781980043879955,
+      "learning_rate": 7.7417971736507e-08,
+      "loss": 0.7087,
+      "step": 14774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1250968264221455,
+      "learning_rate": 7.723638911540432e-08,
+      "loss": 0.774,
+      "step": 14775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6027372024635531,
+      "learning_rate": 7.705501803641769e-08,
+      "loss": 0.6769,
+      "step": 14776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1058191323851263,
+      "learning_rate": 7.687385850734086e-08,
+      "loss": 0.6951,
+      "step": 14777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1022829063078516,
+      "learning_rate": 7.669291053595818e-08,
+      "loss": 0.6871,
+      "step": 14778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.661861575851772,
+      "learning_rate": 7.651217413004674e-08,
+      "loss": 0.6186,
+      "step": 14779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.08950744609698,
+      "learning_rate": 7.633164929737202e-08,
+      "loss": 0.7392,
+      "step": 14780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0780463959457327,
+      "learning_rate": 7.615133604569113e-08,
+      "loss": 0.6911,
+      "step": 14781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.536922981583939,
+      "learning_rate": 7.597123438275399e-08,
+      "loss": 0.7578,
+      "step": 14782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6041154675729314,
+      "learning_rate": 7.579134431629942e-08,
+      "loss": 0.5954,
+      "step": 14783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.617992310968703,
+      "learning_rate": 7.561166585405789e-08,
+      "loss": 0.7128,
+      "step": 14784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7267027354323334,
+      "learning_rate": 7.543219900374987e-08,
+      "loss": 0.7389,
+      "step": 14785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5199376592348686,
+      "learning_rate": 7.525294377308812e-08,
+      "loss": 0.6215,
+      "step": 14786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5401713013234328,
+      "learning_rate": 7.507390016977478e-08,
+      "loss": 0.759,
+      "step": 14787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0253220561510434,
+      "learning_rate": 7.489506820150483e-08,
+      "loss": 0.6929,
+      "step": 14788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1759164515880005,
+      "learning_rate": 7.47164478759621e-08,
+      "loss": 0.65,
+      "step": 14789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.535121062127182,
+      "learning_rate": 7.453803920082381e-08,
+      "loss": 0.6327,
+      "step": 14790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7439975355869632,
+      "learning_rate": 7.435984218375436e-08,
+      "loss": 0.8161,
+      "step": 14791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.710550828135564,
+      "learning_rate": 7.418185683241319e-08,
+      "loss": 0.7989,
+      "step": 14792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0007763854981317,
+      "learning_rate": 7.400408315444807e-08,
+      "loss": 0.7135,
+      "step": 14793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1698279689531197,
+      "learning_rate": 7.382652115749789e-08,
+      "loss": 0.6632,
+      "step": 14794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8639778133885299,
+      "learning_rate": 7.364917084919376e-08,
+      "loss": 0.7308,
+      "step": 14795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.756783540258687,
+      "learning_rate": 7.347203223715571e-08,
+      "loss": 0.7346,
+      "step": 14796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6348963531110141,
+      "learning_rate": 7.329510532899708e-08,
+      "loss": 0.7535,
+      "step": 14797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9765447908465759,
+      "learning_rate": 7.311839013231959e-08,
+      "loss": 0.7783,
+      "step": 14798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8470694777515464,
+      "learning_rate": 7.294188665471769e-08,
+      "loss": 0.7036,
+      "step": 14799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6770552710071431,
+      "learning_rate": 7.276559490377644e-08,
+      "loss": 0.8147,
+      "step": 14800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9139982710224426,
+      "learning_rate": 7.258951488707089e-08,
+      "loss": 0.6754,
+      "step": 14801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3338279265910076,
+      "learning_rate": 7.241364661216776e-08,
+      "loss": 0.7204,
+      "step": 14802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.213562302743063,
+      "learning_rate": 7.223799008662491e-08,
+      "loss": 0.702,
+      "step": 14803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.81433412253376,
+      "learning_rate": 7.206254531799018e-08,
+      "loss": 0.6235,
+      "step": 14804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6041461296462336,
+      "learning_rate": 7.188731231380253e-08,
+      "loss": 0.7592,
+      "step": 14805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9262690887554654,
+      "learning_rate": 7.171229108159262e-08,
+      "loss": 0.6832,
+      "step": 14806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.964720409913378,
+      "learning_rate": 7.153748162888164e-08,
+      "loss": 0.6884,
+      "step": 14807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2076170957995407,
+      "learning_rate": 7.136288396318192e-08,
+      "loss": 0.6801,
+      "step": 14808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1582883695944295,
+      "learning_rate": 7.118849809199524e-08,
+      "loss": 0.7505,
+      "step": 14809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.732944415733336,
+      "learning_rate": 7.10143240228156e-08,
+      "loss": 0.6468,
+      "step": 14810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6945807010506104,
+      "learning_rate": 7.084036176312814e-08,
+      "loss": 0.7036,
+      "step": 14811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9942717006277682,
+      "learning_rate": 7.066661132040853e-08,
+      "loss": 0.8772,
+      "step": 14812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1773338771171336,
+      "learning_rate": 7.049307270212246e-08,
+      "loss": 0.7366,
+      "step": 14813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0954265233231673,
+      "learning_rate": 7.031974591572732e-08,
+      "loss": 0.7339,
+      "step": 14814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4887820928020972,
+      "learning_rate": 7.014663096867214e-08,
+      "loss": 0.6766,
+      "step": 14815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8403272742604841,
+      "learning_rate": 6.99737278683954e-08,
+      "loss": 0.7458,
+      "step": 14816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4885779549144909,
+      "learning_rate": 6.980103662232728e-08,
+      "loss": 0.6729,
+      "step": 14817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7453534501230406,
+      "learning_rate": 6.962855723788909e-08,
+      "loss": 0.6947,
+      "step": 14818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7299353672255964,
+      "learning_rate": 6.945628972249208e-08,
+      "loss": 0.7235,
+      "step": 14819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7428101395280806,
+      "learning_rate": 6.928423408353979e-08,
+      "loss": 0.6024,
+      "step": 14820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.630856010982197,
+      "learning_rate": 6.91123903284241e-08,
+      "loss": 0.729,
+      "step": 14821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6291782664424181,
+      "learning_rate": 6.894075846453185e-08,
+      "loss": 0.7353,
+      "step": 14822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6906809535348961,
+      "learning_rate": 6.876933849923662e-08,
+      "loss": 0.6728,
+      "step": 14823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3173075135444778,
+      "learning_rate": 6.859813043990526e-08,
+      "loss": 0.6576,
+      "step": 14824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6710605723065488,
+      "learning_rate": 6.842713429389525e-08,
+      "loss": 0.7507,
+      "step": 14825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6770506039504705,
+      "learning_rate": 6.825635006855458e-08,
+      "loss": 0.7187,
+      "step": 14826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6232304438322958,
+      "learning_rate": 6.808577777122183e-08,
+      "loss": 0.7568,
+      "step": 14827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9432816750640631,
+      "learning_rate": 6.79154174092278e-08,
+      "loss": 0.7089,
+      "step": 14828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9567233088822757,
+      "learning_rate": 6.77452689898922e-08,
+      "loss": 0.7249,
+      "step": 14829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3334495280006147,
+      "learning_rate": 6.757533252052695e-08,
+      "loss": 0.6938,
+      "step": 14830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5976030980984244,
+      "learning_rate": 6.74056080084351e-08,
+      "loss": 0.6929,
+      "step": 14831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7571532285819933,
+      "learning_rate": 6.723609546090915e-08,
+      "loss": 0.705,
+      "step": 14832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.146026679846405,
+      "learning_rate": 6.706679488523494e-08,
+      "loss": 0.5603,
+      "step": 14833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5540448434785898,
+      "learning_rate": 6.689770628868609e-08,
+      "loss": 0.7169,
+      "step": 14834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6342971779204245,
+      "learning_rate": 6.672882967852956e-08,
+      "loss": 0.6801,
+      "step": 14835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2196344048620358,
+      "learning_rate": 6.656016506202234e-08,
+      "loss": 0.7297,
+      "step": 14836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6926123713654946,
+      "learning_rate": 6.639171244641252e-08,
+      "loss": 0.7622,
+      "step": 14837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8647811760564892,
+      "learning_rate": 6.622347183893818e-08,
+      "loss": 0.6369,
+      "step": 14838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5877630967180671,
+      "learning_rate": 6.605544324682855e-08,
+      "loss": 0.667,
+      "step": 14839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.082014039319625,
+      "learning_rate": 6.58876266773062e-08,
+      "loss": 0.6172,
+      "step": 14840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6523835584707272,
+      "learning_rate": 6.572002213758144e-08,
+      "loss": 0.6759,
+      "step": 14841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7129458301151699,
+      "learning_rate": 6.555262963485576e-08,
+      "loss": 0.7539,
+      "step": 14842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.714754687602705,
+      "learning_rate": 6.53854491763234e-08,
+      "loss": 0.6617,
+      "step": 14843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.988968454130983,
+      "learning_rate": 6.521848076916859e-08,
+      "loss": 0.7938,
+      "step": 14844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.76237525418858,
+      "learning_rate": 6.505172442056562e-08,
+      "loss": 0.6748,
+      "step": 14845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.916849960931398,
+      "learning_rate": 6.488518013768096e-08,
+      "loss": 0.8212,
+      "step": 14846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6501771433780783,
+      "learning_rate": 6.471884792767169e-08,
+      "loss": 0.6881,
+      "step": 14847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0511528759523563,
+      "learning_rate": 6.455272779768429e-08,
+      "loss": 0.5903,
+      "step": 14848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6922536566976458,
+      "learning_rate": 6.438681975485805e-08,
+      "loss": 0.7099,
+      "step": 14849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.991510560659862,
+      "learning_rate": 6.422112380632228e-08,
+      "loss": 0.6973,
+      "step": 14850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.83660278258036,
+      "learning_rate": 6.40556399591985e-08,
+      "loss": 0.6609,
+      "step": 14851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1787241633674803,
+      "learning_rate": 6.389036822059546e-08,
+      "loss": 0.7375,
+      "step": 14852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.23862554638847,
+      "learning_rate": 6.372530859761694e-08,
+      "loss": 0.7593,
+      "step": 14853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6901099936402568,
+      "learning_rate": 6.356046109735614e-08,
+      "loss": 0.6623,
+      "step": 14854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5435239052236236,
+      "learning_rate": 6.339582572689573e-08,
+      "loss": 0.6538,
+      "step": 14855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.943277598398202,
+      "learning_rate": 6.323140249331116e-08,
+      "loss": 0.6913,
+      "step": 14856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8681952531624926,
+      "learning_rate": 6.306719140366791e-08,
+      "loss": 0.7175,
+      "step": 14857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9339166510000514,
+      "learning_rate": 6.290319246502363e-08,
+      "loss": 0.7315,
+      "step": 14858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.836836104096995,
+      "learning_rate": 6.273940568442327e-08,
+      "loss": 0.8287,
+      "step": 14859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6870335328617738,
+      "learning_rate": 6.25758310689073e-08,
+      "loss": 0.6907,
+      "step": 14860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7949890386242207,
+      "learning_rate": 6.241246862550398e-08,
+      "loss": 0.7017,
+      "step": 14861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6583224042666378,
+      "learning_rate": 6.224931836123327e-08,
+      "loss": 0.8529,
+      "step": 14862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9082894274008124,
+      "learning_rate": 6.208638028310677e-08,
+      "loss": 0.6645,
+      "step": 14863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5581345994000364,
+      "learning_rate": 6.192365439812553e-08,
+      "loss": 0.6275,
+      "step": 14864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5598702660391375,
+      "learning_rate": 6.176114071328287e-08,
+      "loss": 0.7554,
+      "step": 14865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6225020571701172,
+      "learning_rate": 6.159883923556154e-08,
+      "loss": 0.7567,
+      "step": 14866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6531596471155645,
+      "learning_rate": 6.143674997193649e-08,
+      "loss": 0.7168,
+      "step": 14867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8252164179736927,
+      "learning_rate": 6.127487292937328e-08,
+      "loss": 0.7485,
+      "step": 14868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6169369063475847,
+      "learning_rate": 6.111320811482802e-08,
+      "loss": 0.5824,
+      "step": 14869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5490052775162417,
+      "learning_rate": 6.095175553524735e-08,
+      "loss": 0.6761,
+      "step": 14870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1975831967387194,
+      "learning_rate": 6.079051519757018e-08,
+      "loss": 0.7106,
+      "step": 14871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6979043024030969,
+      "learning_rate": 6.062948710872373e-08,
+      "loss": 0.5474,
+      "step": 14872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5950497758501936,
+      "learning_rate": 6.046867127563027e-08,
+      "loss": 0.6407,
+      "step": 14873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6436798184310395,
+      "learning_rate": 6.030806770519815e-08,
+      "loss": 0.8446,
+      "step": 14874
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9468336339026724,
+      "learning_rate": 6.014767640432905e-08,
+      "loss": 0.7024,
+      "step": 14875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5142818571382923,
+      "learning_rate": 5.998749737991693e-08,
+      "loss": 0.8674,
+      "step": 14876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8579788884673951,
+      "learning_rate": 5.98275306388435e-08,
+      "loss": 0.7906,
+      "step": 14877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0310340136069065,
+      "learning_rate": 5.96677761879838e-08,
+      "loss": 0.6273,
+      "step": 14878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6458338335613454,
+      "learning_rate": 5.9508234034202364e-08,
+      "loss": 0.6542,
+      "step": 14879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5251060886877401,
+      "learning_rate": 5.9348904184354795e-08,
+      "loss": 0.6314,
+      "step": 14880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3949707676170027,
+      "learning_rate": 5.918978664528841e-08,
+      "loss": 0.746,
+      "step": 14881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9555919497642045,
+      "learning_rate": 5.903088142384106e-08,
+      "loss": 0.7875,
+      "step": 14882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.918893939948121,
+      "learning_rate": 5.887218852684007e-08,
+      "loss": 0.7795,
+      "step": 14883
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.90442036611093,
+      "learning_rate": 5.871370796110665e-08,
+      "loss": 0.7164,
+      "step": 14884
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7491098140020096,
+      "learning_rate": 5.855543973344924e-08,
+      "loss": 0.6524,
+      "step": 14885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5618454840098408,
+      "learning_rate": 5.8397383850670176e-08,
+      "loss": 0.6,
+      "step": 14886
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.488635596372504,
+      "learning_rate": 5.823954031956069e-08,
+      "loss": 0.641,
+      "step": 14887
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9489756261319529,
+      "learning_rate": 5.808190914690426e-08,
+      "loss": 0.7258,
+      "step": 14888
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5475427016518835,
+      "learning_rate": 5.7924490339474335e-08,
+      "loss": 0.7246,
+      "step": 14889
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6944930356284424,
+      "learning_rate": 5.776728390403552e-08,
+      "loss": 0.6437,
+      "step": 14890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.283255258375448,
+      "learning_rate": 5.761028984734407e-08,
+      "loss": 0.6697,
+      "step": 14891
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6250221452451485,
+      "learning_rate": 5.74535081761457e-08,
+      "loss": 0.677,
+      "step": 14892
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8885246023631475,
+      "learning_rate": 5.7296938897177245e-08,
+      "loss": 0.7021,
+      "step": 14893
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5250567451290091,
+      "learning_rate": 5.7140582017167764e-08,
+      "loss": 0.6997,
+      "step": 14894
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8506005166450263,
+      "learning_rate": 5.6984437542835225e-08,
+      "loss": 0.6021,
+      "step": 14895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1587159635918547,
+      "learning_rate": 5.682850548089036e-08,
+      "loss": 0.6772,
+      "step": 14896
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.39005939777141,
+      "learning_rate": 5.6672785838033925e-08,
+      "loss": 0.6135,
+      "step": 14897
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4528168797451626,
+      "learning_rate": 5.651727862095668e-08,
+      "loss": 0.6348,
+      "step": 14898
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7728728578933501,
+      "learning_rate": 5.636198383634217e-08,
+      "loss": 0.6629,
+      "step": 14899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4571278422471843,
+      "learning_rate": 5.620690149086283e-08,
+      "loss": 0.6553,
+      "step": 14900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8389720579480633,
+      "learning_rate": 5.6052031591183884e-08,
+      "loss": 0.6915,
+      "step": 14901
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8338679442306143,
+      "learning_rate": 5.5897374143960013e-08,
+      "loss": 0.623,
+      "step": 14902
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2267777094236167,
+      "learning_rate": 5.574292915583646e-08,
+      "loss": 0.6786,
+      "step": 14903
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5552000175714285,
+      "learning_rate": 5.558869663345123e-08,
+      "loss": 0.6876,
+      "step": 14904
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5844038149927626,
+      "learning_rate": 5.543467658343127e-08,
+      "loss": 0.6935,
+      "step": 14905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8166815597956887,
+      "learning_rate": 5.5280869012395154e-08,
+      "loss": 0.7089,
+      "step": 14906
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.010330454905795,
+      "learning_rate": 5.5127273926953165e-08,
+      "loss": 0.8528,
+      "step": 14907
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1202679501024801,
+      "learning_rate": 5.4973891333704453e-08,
+      "loss": 0.7197,
+      "step": 14908
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3406209082686307,
+      "learning_rate": 5.482072123924098e-08,
+      "loss": 0.6448,
+      "step": 14909
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.096927408027728,
+      "learning_rate": 5.46677636501447e-08,
+      "loss": 0.7053,
+      "step": 14910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.84704351446542,
+      "learning_rate": 5.4515018572988687e-08,
+      "loss": 0.691,
+      "step": 14911
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7936623839216017,
+      "learning_rate": 5.436248601433658e-08,
+      "loss": 0.8262,
+      "step": 14912
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1142316967320582,
+      "learning_rate": 5.4210165980742025e-08,
+      "loss": 0.673,
+      "step": 14913
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.268065950359636,
+      "learning_rate": 5.405805847875256e-08,
+      "loss": 0.6504,
+      "step": 14914
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5536331890778958,
+      "learning_rate": 5.3906163514902966e-08,
+      "loss": 0.8376,
+      "step": 14915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1062789974993181,
+      "learning_rate": 5.3754481095721344e-08,
+      "loss": 0.6557,
+      "step": 14916
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.511565916510537,
+      "learning_rate": 5.3603011227725265e-08,
+      "loss": 0.5437,
+      "step": 14917
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5597681024045236,
+      "learning_rate": 5.3451753917423964e-08,
+      "loss": 0.7386,
+      "step": 14918
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.631145934304313,
+      "learning_rate": 5.330070917131724e-08,
+      "loss": 0.7268,
+      "step": 14919
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6231513533066335,
+      "learning_rate": 5.3149876995896575e-08,
+      "loss": 0.6404,
+      "step": 14920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5092525793065152,
+      "learning_rate": 5.299925739764233e-08,
+      "loss": 0.6859,
+      "step": 14921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.749511518432324,
+      "learning_rate": 5.284885038302823e-08,
+      "loss": 0.643,
+      "step": 14922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9272340689871805,
+      "learning_rate": 5.269865595851631e-08,
+      "loss": 0.6328,
+      "step": 14923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.112075058947361,
+      "learning_rate": 5.2548674130561974e-08,
+      "loss": 0.5713,
+      "step": 14924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8878142096815178,
+      "learning_rate": 5.239890490560895e-08,
+      "loss": 0.6607,
+      "step": 14925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5415446479179828,
+      "learning_rate": 5.224934829009376e-08,
+      "loss": 0.6943,
+      "step": 14926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6149095175357375,
+      "learning_rate": 5.21000042904446e-08,
+      "loss": 0.6575,
+      "step": 14927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9757083619627598,
+      "learning_rate": 5.195087291307632e-08,
+      "loss": 0.6966,
+      "step": 14928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7315699857898976,
+      "learning_rate": 5.1801954164399925e-08,
+      "loss": 0.7171,
+      "step": 14929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7199723033461722,
+      "learning_rate": 5.165324805081362e-08,
+      "loss": 0.7072,
+      "step": 14930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6346627062358268,
+      "learning_rate": 5.1504754578707294e-08,
+      "loss": 0.5957,
+      "step": 14931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2833413265853388,
+      "learning_rate": 5.135647375446307e-08,
+      "loss": 0.6701,
+      "step": 14932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.53771722768885,
+      "learning_rate": 5.120840558445195e-08,
+      "loss": 0.7441,
+      "step": 14933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8829234414458327,
+      "learning_rate": 5.106055007503774e-08,
+      "loss": 0.7154,
+      "step": 14934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.474863324364915,
+      "learning_rate": 5.091290723257314e-08,
+      "loss": 0.645,
+      "step": 14935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2884691907352632,
+      "learning_rate": 5.076547706340307e-08,
+      "loss": 0.5631,
+      "step": 14936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6640517918760878,
+      "learning_rate": 5.061825957386302e-08,
+      "loss": 0.6873,
+      "step": 14937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6149079470023642,
+      "learning_rate": 5.047125477027959e-08,
+      "loss": 0.7061,
+      "step": 14938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1899590276954202,
+      "learning_rate": 5.0324462658969395e-08,
+      "loss": 0.6234,
+      "step": 14939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6400312733840063,
+      "learning_rate": 5.017788324624073e-08,
+      "loss": 0.6552,
+      "step": 14940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.909285813349263,
+      "learning_rate": 5.003151653839189e-08,
+      "loss": 0.6341,
+      "step": 14941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.601720242838217,
+      "learning_rate": 4.988536254171284e-08,
+      "loss": 0.7363,
+      "step": 14942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7405454413509078,
+      "learning_rate": 4.9739421262484656e-08,
+      "loss": 0.7898,
+      "step": 14943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9016872445492308,
+      "learning_rate": 4.959369270697789e-08,
+      "loss": 0.7533,
+      "step": 14944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6501736778856873,
+      "learning_rate": 4.944817688145642e-08,
+      "loss": 0.6699,
+      "step": 14945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.195488382367502,
+      "learning_rate": 4.93028737921708e-08,
+      "loss": 0.618,
+      "step": 14946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0096146447530936,
+      "learning_rate": 4.915778344536715e-08,
+      "loss": 0.8757,
+      "step": 14947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9012479934971869,
+      "learning_rate": 4.9012905847279913e-08,
+      "loss": 0.7519,
+      "step": 14948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5670615976918987,
+      "learning_rate": 4.886824100413412e-08,
+      "loss": 0.5781,
+      "step": 14949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.681542955457265,
+      "learning_rate": 4.872378892214702e-08,
+      "loss": 0.746,
+      "step": 14950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.810444255198452,
+      "learning_rate": 4.8579549607525866e-08,
+      "loss": 0.7524,
+      "step": 14951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5074261225185102,
+      "learning_rate": 4.843552306646904e-08,
+      "loss": 0.6445,
+      "step": 14952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.93480081956848,
+      "learning_rate": 4.829170930516491e-08,
+      "loss": 0.7008,
+      "step": 14953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5584885548542537,
+      "learning_rate": 4.814810832979411e-08,
+      "loss": 0.7479,
+      "step": 14954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5952828379604755,
+      "learning_rate": 4.8004720146527795e-08,
+      "loss": 0.6824,
+      "step": 14955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1214075129836734,
+      "learning_rate": 4.786154476152716e-08,
+      "loss": 0.7737,
+      "step": 14956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0519097667625497,
+      "learning_rate": 4.771858218094505e-08,
+      "loss": 0.6737,
+      "step": 14957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.683267467778827,
+      "learning_rate": 4.7575832410924893e-08,
+      "loss": 0.7737,
+      "step": 14958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7590049549462523,
+      "learning_rate": 4.743329545760122e-08,
+      "loss": 0.7021,
+      "step": 14959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7687269630391673,
+      "learning_rate": 4.7290971327098035e-08,
+      "loss": 0.8012,
+      "step": 14960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5236122717661391,
+      "learning_rate": 4.71488600255321e-08,
+      "loss": 0.715,
+      "step": 14961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8272254239590433,
+      "learning_rate": 4.7006961559010745e-08,
+      "loss": 0.6622,
+      "step": 14962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9473348364942555,
+      "learning_rate": 4.686527593363133e-08,
+      "loss": 0.707,
+      "step": 14963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8460354815354967,
+      "learning_rate": 4.67238031554812e-08,
+      "loss": 0.6791,
+      "step": 14964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.889715568385557,
+      "learning_rate": 4.6582543230641596e-08,
+      "loss": 0.8138,
+      "step": 14965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.635143102691409,
+      "learning_rate": 4.644149616518212e-08,
+      "loss": 0.6423,
+      "step": 14966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7045422124077085,
+      "learning_rate": 4.630066196516347e-08,
+      "loss": 0.6389,
+      "step": 14967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.677833188913711,
+      "learning_rate": 4.6160040636637484e-08,
+      "loss": 0.5736,
+      "step": 14968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8460410529207463,
+      "learning_rate": 4.6019632185647645e-08,
+      "loss": 0.6968,
+      "step": 14969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0593658654216096,
+      "learning_rate": 4.587943661822747e-08,
+      "loss": 0.6509,
+      "step": 14970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1457196931387323,
+      "learning_rate": 4.5739453940401024e-08,
+      "loss": 0.7242,
+      "step": 14971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8208399002496913,
+      "learning_rate": 4.5599684158184055e-08,
+      "loss": 0.675,
+      "step": 14972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7159374563250986,
+      "learning_rate": 4.5460127277582863e-08,
+      "loss": 0.6586,
+      "step": 14973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5709110784122942,
+      "learning_rate": 4.532078330459433e-08,
+      "loss": 0.7179,
+      "step": 14974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4925989813103338,
+      "learning_rate": 4.518165224520643e-08,
+      "loss": 0.7251,
+      "step": 14975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7116791730199055,
+      "learning_rate": 4.504273410539772e-08,
+      "loss": 0.6958,
+      "step": 14976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6184650014222846,
+      "learning_rate": 4.490402889113787e-08,
+      "loss": 0.5982,
+      "step": 14977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9538016380408196,
+      "learning_rate": 4.4765536608388227e-08,
+      "loss": 0.685,
+      "step": 14978
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2130791452249707,
+      "learning_rate": 4.4627257263098465e-08,
+      "loss": 0.6957,
+      "step": 14979
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8594587704728962,
+      "learning_rate": 4.448919086121217e-08,
+      "loss": 0.6645,
+      "step": 14980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9002407404891108,
+      "learning_rate": 4.435133740866182e-08,
+      "loss": 0.7487,
+      "step": 14981
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.242722344027821,
+      "learning_rate": 4.4213696911371006e-08,
+      "loss": 0.6274,
+      "step": 14982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7951428794555142,
+      "learning_rate": 4.4076269375255e-08,
+      "loss": 0.6886,
+      "step": 14983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6391587756832628,
+      "learning_rate": 4.393905480621907e-08,
+      "loss": 0.7042,
+      "step": 14984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7036657168726959,
+      "learning_rate": 4.380205321016018e-08,
+      "loss": 0.6713,
+      "step": 14985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0686580144775486,
+      "learning_rate": 4.366526459296472e-08,
+      "loss": 0.6183,
+      "step": 14986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0834797700956116,
+      "learning_rate": 4.352868896051077e-08,
+      "loss": 0.7036,
+      "step": 14987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7137125182923825,
+      "learning_rate": 4.3392326318668076e-08,
+      "loss": 0.5951,
+      "step": 14988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.125760515751485,
+      "learning_rate": 4.3256176673295846e-08,
+      "loss": 0.7212,
+      "step": 14989
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6972373670213394,
+      "learning_rate": 4.312024003024551e-08,
+      "loss": 0.7656,
+      "step": 14990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5397179566050825,
+      "learning_rate": 4.2984516395357944e-08,
+      "loss": 0.7698,
+      "step": 14991
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5653009714561912,
+      "learning_rate": 4.284900577446516e-08,
+      "loss": 0.6896,
+      "step": 14992
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.421807959788813,
+      "learning_rate": 4.271370817339082e-08,
+      "loss": 0.6436,
+      "step": 14993
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9275751222111615,
+      "learning_rate": 4.2578623597949174e-08,
+      "loss": 0.6948,
+      "step": 14994
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.764752938656084,
+      "learning_rate": 4.244375205394502e-08,
+      "loss": 0.676,
+      "step": 14995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.070677175219681,
+      "learning_rate": 4.230909354717372e-08,
+      "loss": 0.6697,
+      "step": 14996
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6712737351707072,
+      "learning_rate": 4.217464808342175e-08,
+      "loss": 0.7585,
+      "step": 14997
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7376448577233528,
+      "learning_rate": 4.204041566846673e-08,
+      "loss": 0.7765,
+      "step": 14998
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1194387648303175,
+      "learning_rate": 4.1906396308077356e-08,
+      "loss": 0.6116,
+      "step": 14999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8552637246404173,
+      "learning_rate": 4.177259000801237e-08,
+      "loss": 0.725,
+      "step": 15000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6072501262636294,
+      "learning_rate": 4.163899677402161e-08,
+      "loss": 0.6237,
+      "step": 15001
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7053678789513074,
+      "learning_rate": 4.150561661184604e-08,
+      "loss": 0.6873,
+      "step": 15002
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4011135539169572,
+      "learning_rate": 4.1372449527217754e-08,
+      "loss": 0.5173,
+      "step": 15003
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6995370041973832,
+      "learning_rate": 4.123949552585826e-08,
+      "loss": 0.7713,
+      "step": 15004
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7599046206798452,
+      "learning_rate": 4.1106754613481347e-08,
+      "loss": 0.748,
+      "step": 15005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0958378138355163,
+      "learning_rate": 4.097422679579188e-08,
+      "loss": 0.6875,
+      "step": 15006
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8517135608156223,
+      "learning_rate": 4.0841912078483645e-08,
+      "loss": 0.6962,
+      "step": 15007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.447358984544923,
+      "learning_rate": 4.0709810467243204e-08,
+      "loss": 0.6461,
+      "step": 15008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5692383637066485,
+      "learning_rate": 4.0577921967747126e-08,
+      "loss": 0.7512,
+      "step": 15009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3448763889680486,
+      "learning_rate": 4.0446246585663093e-08,
+      "loss": 0.6488,
+      "step": 15010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7835287155613726,
+      "learning_rate": 4.031478432664937e-08,
+      "loss": 0.7237,
+      "step": 15011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7218931065993524,
+      "learning_rate": 4.018353519635476e-08,
+      "loss": 0.7202,
+      "step": 15012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6223581320205565,
+      "learning_rate": 4.0052499200420314e-08,
+      "loss": 0.7,
+      "step": 15013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7692773759230067,
+      "learning_rate": 3.9921676344475966e-08,
+      "loss": 0.8362,
+      "step": 15014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6867703159706233,
+      "learning_rate": 3.979106663414389e-08,
+      "loss": 0.6905,
+      "step": 15015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.594807335301339,
+      "learning_rate": 3.9660670075036823e-08,
+      "loss": 0.6409,
+      "step": 15016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3119766278353782,
+      "learning_rate": 3.953048667275805e-08,
+      "loss": 0.5677,
+      "step": 15017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8246098907527826,
+      "learning_rate": 3.9400516432901434e-08,
+      "loss": 0.6302,
+      "step": 15018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5234939567833605,
+      "learning_rate": 3.927075936105307e-08,
+      "loss": 0.654,
+      "step": 15019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6898996724708872,
+      "learning_rate": 3.914121546278793e-08,
+      "loss": 0.7056,
+      "step": 15020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6317882504188144,
+      "learning_rate": 3.901188474367323e-08,
+      "loss": 0.7268,
+      "step": 15021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6846923823422224,
+      "learning_rate": 3.8882767209266756e-08,
+      "loss": 0.5755,
+      "step": 15022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7608117966762948,
+      "learning_rate": 3.875386286511684e-08,
+      "loss": 0.7447,
+      "step": 15023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7509911345866291,
+      "learning_rate": 3.8625171716762385e-08,
+      "loss": 0.6987,
+      "step": 15024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.769495484167332,
+      "learning_rate": 3.849669376973397e-08,
+      "loss": 0.7228,
+      "step": 15025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2105747145364447,
+      "learning_rate": 3.836842902955273e-08,
+      "loss": 0.6581,
+      "step": 15026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8828680482858107,
+      "learning_rate": 3.8240377501729817e-08,
+      "loss": 0.7436,
+      "step": 15027
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9388278156012269,
+      "learning_rate": 3.8112539191768604e-08,
+      "loss": 0.7087,
+      "step": 15028
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8035577051501703,
+      "learning_rate": 3.7984914105162474e-08,
+      "loss": 0.7181,
+      "step": 15029
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8113331498644498,
+      "learning_rate": 3.785750224739482e-08,
+      "loss": 0.7499,
+      "step": 15030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5745561802902226,
+      "learning_rate": 3.7730303623942364e-08,
+      "loss": 0.7258,
+      "step": 15031
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2159639868720586,
+      "learning_rate": 3.760331824027019e-08,
+      "loss": 0.6303,
+      "step": 15032
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3946278021993779,
+      "learning_rate": 3.7476546101835045e-08,
+      "loss": 0.5822,
+      "step": 15033
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.731244474287305,
+      "learning_rate": 3.7349987214084784e-08,
+      "loss": 0.67,
+      "step": 15034
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9387842839104612,
+      "learning_rate": 3.722364158245839e-08,
+      "loss": 0.6817,
+      "step": 15035
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.566193949970245,
+      "learning_rate": 3.709750921238486e-08,
+      "loss": 0.6993,
+      "step": 15036
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.442579174741549,
+      "learning_rate": 3.697159010928375e-08,
+      "loss": 0.7583,
+      "step": 15037
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5477576511848632,
+      "learning_rate": 3.6845884278566836e-08,
+      "loss": 0.6955,
+      "step": 15038
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0700150827925294,
+      "learning_rate": 3.672039172563646e-08,
+      "loss": 0.6802,
+      "step": 15039
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6159325371839004,
+      "learning_rate": 3.659511245588387e-08,
+      "loss": 0.674,
+      "step": 15040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7363391984578713,
+      "learning_rate": 3.647004647469365e-08,
+      "loss": 0.801,
+      "step": 15041
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6261186294278214,
+      "learning_rate": 3.634519378744039e-08,
+      "loss": 0.5791,
+      "step": 15042
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7945449892768948,
+      "learning_rate": 3.622055439948813e-08,
+      "loss": 0.6865,
+      "step": 15043
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6255097840448824,
+      "learning_rate": 3.609612831619369e-08,
+      "loss": 0.6851,
+      "step": 15044
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.603958252973634,
+      "learning_rate": 3.597191554290391e-08,
+      "loss": 0.7428,
+      "step": 15045
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5990203507705851,
+      "learning_rate": 3.584791608495675e-08,
+      "loss": 0.6847,
+      "step": 15046
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6326371264756814,
+      "learning_rate": 3.5724129947680156e-08,
+      "loss": 0.7948,
+      "step": 15047
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5783157519206261,
+      "learning_rate": 3.560055713639321e-08,
+      "loss": 0.8063,
+      "step": 15048
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6388029663324684,
+      "learning_rate": 3.547719765640778e-08,
+      "loss": 0.7571,
+      "step": 15049
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.151393089690464,
+      "learning_rate": 3.5354051513022405e-08,
+      "loss": 0.7958,
+      "step": 15050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8692213857104731,
+      "learning_rate": 3.523111871153118e-08,
+      "loss": 0.7182,
+      "step": 15051
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4403777116130285,
+      "learning_rate": 3.510839925721543e-08,
+      "loss": 0.6494,
+      "step": 15052
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8561008223103483,
+      "learning_rate": 3.4985893155349835e-08,
+      "loss": 0.7734,
+      "step": 15053
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.736277641894853,
+      "learning_rate": 3.4863600411197404e-08,
+      "loss": 0.5995,
+      "step": 15054
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3664724671448676,
+      "learning_rate": 3.474152103001449e-08,
+      "loss": 0.7039,
+      "step": 15055
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1068857654347828,
+      "learning_rate": 3.461965501704634e-08,
+      "loss": 0.6458,
+      "step": 15056
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0086272354997645,
+      "learning_rate": 3.449800237753043e-08,
+      "loss": 0.6874,
+      "step": 15057
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1447192013775231,
+      "learning_rate": 3.4376563116693685e-08,
+      "loss": 0.6201,
+      "step": 15058
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9948351919890851,
+      "learning_rate": 3.425533723975527e-08,
+      "loss": 0.6859,
+      "step": 15059
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2177065848519553,
+      "learning_rate": 3.413432475192435e-08,
+      "loss": 0.6142,
+      "step": 15060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.691936687134409,
+      "learning_rate": 3.401352565840121e-08,
+      "loss": 0.7078,
+      "step": 15061
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9036390532376168,
+      "learning_rate": 3.3892939964376704e-08,
+      "loss": 0.6911,
+      "step": 15062
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0237268886056152,
+      "learning_rate": 3.377256767503223e-08,
+      "loss": 0.772,
+      "step": 15063
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0434240103063335,
+      "learning_rate": 3.365240879554144e-08,
+      "loss": 0.799,
+      "step": 15064
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2034875387248114,
+      "learning_rate": 3.353246333106741e-08,
+      "loss": 0.6179,
+      "step": 15065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8756867204574448,
+      "learning_rate": 3.3412731286763256e-08,
+      "loss": 0.7321,
+      "step": 15066
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.832358298789196,
+      "learning_rate": 3.329321266777652e-08,
+      "loss": 0.7953,
+      "step": 15067
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6369604169400909,
+      "learning_rate": 3.3173907479240875e-08,
+      "loss": 0.6162,
+      "step": 15068
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6898153042469382,
+      "learning_rate": 3.3054815726285e-08,
+      "loss": 0.7544,
+      "step": 15069
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0704827009427764,
+      "learning_rate": 3.2935937414025345e-08,
+      "loss": 0.8013,
+      "step": 15070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.626546411230416,
+      "learning_rate": 3.281727254757061e-08,
+      "loss": 0.9006,
+      "step": 15071
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2246426997877817,
+      "learning_rate": 3.269882113202005e-08,
+      "loss": 0.743,
+      "step": 15072
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9743800487343481,
+      "learning_rate": 3.258058317246404e-08,
+      "loss": 0.6122,
+      "step": 15073
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8756062209667463,
+      "learning_rate": 3.2462558673983516e-08,
+      "loss": 0.7164,
+      "step": 15074
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8154086755360574,
+      "learning_rate": 3.234474764165052e-08,
+      "loss": 0.7241,
+      "step": 15075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7493883461984843,
+      "learning_rate": 3.222715008052657e-08,
+      "loss": 0.7739,
+      "step": 15076
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8233266998401623,
+      "learning_rate": 3.2109765995665957e-08,
+      "loss": 0.651,
+      "step": 15077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6825911897905224,
+      "learning_rate": 3.1992595392112966e-08,
+      "loss": 0.6902,
+      "step": 15078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.817139064549965,
+      "learning_rate": 3.1875638274902476e-08,
+      "loss": 0.6666,
+      "step": 15079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.267832881491834,
+      "learning_rate": 3.1758894649060456e-08,
+      "loss": 0.7373,
+      "step": 15080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8731993316367734,
+      "learning_rate": 3.164236451960345e-08,
+      "loss": 0.7449,
+      "step": 15081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5892482998105382,
+      "learning_rate": 3.152604789153912e-08,
+      "loss": 0.7228,
+      "step": 15082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8480063050630666,
+      "learning_rate": 3.140994476986625e-08,
+      "loss": 0.7237,
+      "step": 15083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7349015693960748,
+      "learning_rate": 3.129405515957307e-08,
+      "loss": 0.7732,
+      "step": 15084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9393002709603961,
+      "learning_rate": 3.117837906564114e-08,
+      "loss": 0.6669,
+      "step": 15085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.72268386123379,
+      "learning_rate": 3.106291649303983e-08,
+      "loss": 0.7531,
+      "step": 15086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6450865338391054,
+      "learning_rate": 3.094766744673128e-08,
+      "loss": 0.7226,
+      "step": 15087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0257364225215415,
+      "learning_rate": 3.083263193166819e-08,
+      "loss": 0.6333,
+      "step": 15088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0506744837425686,
+      "learning_rate": 3.071780995279439e-08,
+      "loss": 0.6758,
+      "step": 15089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6136925066064143,
+      "learning_rate": 3.060320151504315e-08,
+      "loss": 0.6298,
+      "step": 15090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5109034937071328,
+      "learning_rate": 3.0488806623339414e-08,
+      "loss": 0.7444,
+      "step": 15091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7940833220705006,
+      "learning_rate": 3.0374625282599826e-08,
+      "loss": 0.6793,
+      "step": 15092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.223370211799864,
+      "learning_rate": 3.026065749773044e-08,
+      "loss": 0.7854,
+      "step": 15093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1138626452050298,
+      "learning_rate": 3.014690327362846e-08,
+      "loss": 0.6248,
+      "step": 15094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1751413961082515,
+      "learning_rate": 3.003336261518275e-08,
+      "loss": 0.6784,
+      "step": 15095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6106062503822864,
+      "learning_rate": 2.992003552727219e-08,
+      "loss": 0.7139,
+      "step": 15096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7070868255202707,
+      "learning_rate": 2.980692201476676e-08,
+      "loss": 0.6359,
+      "step": 15097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.429776528719208,
+      "learning_rate": 2.969402208252703e-08,
+      "loss": 0.697,
+      "step": 15098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.256280019496336,
+      "learning_rate": 2.9581335735404672e-08,
+      "loss": 0.6353,
+      "step": 15099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7579356459209383,
+      "learning_rate": 2.9468862978241917e-08,
+      "loss": 0.7152,
+      "step": 15100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.923621035635826,
+      "learning_rate": 2.9356603815872132e-08,
+      "loss": 0.6703,
+      "step": 15101
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5440770511741997,
+      "learning_rate": 2.9244558253119226e-08,
+      "loss": 0.7126,
+      "step": 15102
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.640086050311251,
+      "learning_rate": 2.9132726294798242e-08,
+      "loss": 0.7154,
+      "step": 15103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9825764081341448,
+      "learning_rate": 2.9021107945714777e-08,
+      "loss": 0.6788,
+      "step": 15104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6192990180086484,
+      "learning_rate": 2.8909703210664995e-08,
+      "loss": 0.6719,
+      "step": 15105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7228648754283684,
+      "learning_rate": 2.8798512094436738e-08,
+      "loss": 0.6457,
+      "step": 15106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.549412713023184,
+      "learning_rate": 2.8687534601807842e-08,
+      "loss": 0.7294,
+      "step": 15107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7894615551259279,
+      "learning_rate": 2.8576770737547276e-08,
+      "loss": 0.7276,
+      "step": 15108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.862888338460288,
+      "learning_rate": 2.8466220506414565e-08,
+      "loss": 0.7143,
+      "step": 15109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.628921369159612,
+      "learning_rate": 2.8355883913160907e-08,
+      "loss": 0.7758,
+      "step": 15110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.661517239032183,
+      "learning_rate": 2.8245760962527514e-08,
+      "loss": 0.8128,
+      "step": 15111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8826464163369474,
+      "learning_rate": 2.813585165924615e-08,
+      "loss": 0.6791,
+      "step": 15112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5841347796865044,
+      "learning_rate": 2.802615600804026e-08,
+      "loss": 0.6303,
+      "step": 15113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4857401303442614,
+      "learning_rate": 2.79166740136233e-08,
+      "loss": 0.6583,
+      "step": 15114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.758984888960159,
+      "learning_rate": 2.780740568070095e-08,
+      "loss": 0.7678,
+      "step": 15115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8472628755090026,
+      "learning_rate": 2.769835101396723e-08,
+      "loss": 0.6677,
+      "step": 15116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8563498890594237,
+      "learning_rate": 2.75895100181095e-08,
+      "loss": 0.6944,
+      "step": 15117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8101055755233029,
+      "learning_rate": 2.7480882697805133e-08,
+      "loss": 0.798,
+      "step": 15118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3627932841045496,
+      "learning_rate": 2.7372469057721506e-08,
+      "loss": 0.7156,
+      "step": 15119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.721510145065034,
+      "learning_rate": 2.7264269102517117e-08,
+      "loss": 0.6666,
+      "step": 15120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1079469968072377,
+      "learning_rate": 2.7156282836842683e-08,
+      "loss": 0.6718,
+      "step": 15121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0056607488070695,
+      "learning_rate": 2.704851026533728e-08,
+      "loss": 0.7652,
+      "step": 15122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6215778115321153,
+      "learning_rate": 2.6940951392632753e-08,
+      "loss": 0.6294,
+      "step": 15123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6119317489217362,
+      "learning_rate": 2.6833606223351515e-08,
+      "loss": 0.6746,
+      "step": 15124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.166480857958241,
+      "learning_rate": 2.6726474762105992e-08,
+      "loss": 0.7349,
+      "step": 15125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6866474776931373,
+      "learning_rate": 2.661955701349972e-08,
+      "loss": 0.7319,
+      "step": 15126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9019262054692767,
+      "learning_rate": 2.6512852982127357e-08,
+      "loss": 0.6837,
+      "step": 15127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9481107707094076,
+      "learning_rate": 2.6406362672574125e-08,
+      "loss": 0.7478,
+      "step": 15128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5676461204908572,
+      "learning_rate": 2.6300086089416366e-08,
+      "loss": 0.7493,
+      "step": 15129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.658165917578451,
+      "learning_rate": 2.6194023237221533e-08,
+      "loss": 0.6043,
+      "step": 15130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6560811237627677,
+      "learning_rate": 2.6088174120545985e-08,
+      "loss": 0.584,
+      "step": 15131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3148868582076774,
+      "learning_rate": 2.5982538743939413e-08,
+      "loss": 0.7096,
+      "step": 15132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.930160194798084,
+      "learning_rate": 2.5877117111940964e-08,
+      "loss": 0.8013,
+      "step": 15133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9707771946889912,
+      "learning_rate": 2.577190922908035e-08,
+      "loss": 0.7711,
+      "step": 15134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.651894942416351,
+      "learning_rate": 2.5666915099879508e-08,
+      "loss": 0.5877,
+      "step": 15135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.683684051668555,
+      "learning_rate": 2.556213472884983e-08,
+      "loss": 0.7676,
+      "step": 15136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.225541259250911,
+      "learning_rate": 2.5457568120493825e-08,
+      "loss": 0.6937,
+      "step": 15137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7497863248065824,
+      "learning_rate": 2.5353215279304566e-08,
+      "loss": 0.6809,
+      "step": 15138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9389105625332034,
+      "learning_rate": 2.5249076209767353e-08,
+      "loss": 0.5302,
+      "step": 15139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1609455589086897,
+      "learning_rate": 2.5145150916356386e-08,
+      "loss": 0.692,
+      "step": 15140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3751718857775148,
+      "learning_rate": 2.5041439403537537e-08,
+      "loss": 0.6906,
+      "step": 15141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5833689797425963,
+      "learning_rate": 2.493794167576835e-08,
+      "loss": 0.7026,
+      "step": 15142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5683941017956655,
+      "learning_rate": 2.4834657737495828e-08,
+      "loss": 0.6989,
+      "step": 15143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7432812100773079,
+      "learning_rate": 2.473158759315808e-08,
+      "loss": 0.6665,
+      "step": 15144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7303379595114576,
+      "learning_rate": 2.4628731247184902e-08,
+      "loss": 0.6937,
+      "step": 15145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2664620819326355,
+      "learning_rate": 2.452608870399553e-08,
+      "loss": 0.7078,
+      "step": 15146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5791975895194241,
+      "learning_rate": 2.4423659968001446e-08,
+      "loss": 0.5911,
+      "step": 15147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9337042736694763,
+      "learning_rate": 2.4321445043603565e-08,
+      "loss": 0.6939,
+      "step": 15148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.963493969841467,
+      "learning_rate": 2.421944393519504e-08,
+      "loss": 0.7892,
+      "step": 15149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8788301852080438,
+      "learning_rate": 2.411765664715793e-08,
+      "loss": 0.7019,
+      "step": 15150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0456605707907793,
+      "learning_rate": 2.401608318386761e-08,
+      "loss": 0.726,
+      "step": 15151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7388642073082774,
+      "learning_rate": 2.3914723549687825e-08,
+      "loss": 0.5843,
+      "step": 15152
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8857255915931046,
+      "learning_rate": 2.381357774897508e-08,
+      "loss": 0.7018,
+      "step": 15153
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.251741589900394,
+      "learning_rate": 2.3712645786075905e-08,
+      "loss": 0.7084,
+      "step": 15154
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7343812707364394,
+      "learning_rate": 2.3611927665326275e-08,
+      "loss": 0.786,
+      "step": 15155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5891301991263427,
+      "learning_rate": 2.3511423391055498e-08,
+      "loss": 0.5735,
+      "step": 15156
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.182463366759183,
+      "learning_rate": 2.341113296758235e-08,
+      "loss": 0.6958,
+      "step": 15157
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.650745732499472,
+      "learning_rate": 2.3311056399216158e-08,
+      "loss": 0.6741,
+      "step": 15158
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7413925697088308,
+      "learning_rate": 2.3211193690257373e-08,
+      "loss": 0.6173,
+      "step": 15159
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.753089327054124,
+      "learning_rate": 2.3111544844997003e-08,
+      "loss": 0.7154,
+      "step": 15160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7415958745862703,
+      "learning_rate": 2.3012109867718845e-08,
+      "loss": 0.7121,
+      "step": 15161
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7640646918293066,
+      "learning_rate": 2.291288876269393e-08,
+      "loss": 0.612,
+      "step": 15162
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.200658167971177,
+      "learning_rate": 2.2813881534186622e-08,
+      "loss": 0.6275,
+      "step": 15163
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.52912934646632,
+      "learning_rate": 2.271508818645185e-08,
+      "loss": 0.7136,
+      "step": 15164
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1218026945702373,
+      "learning_rate": 2.2616508723734555e-08,
+      "loss": 0.6204,
+      "step": 15165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.827182755749929,
+      "learning_rate": 2.2518143150271344e-08,
+      "loss": 0.7324,
+      "step": 15166
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2316122793392512,
+      "learning_rate": 2.2419991470289393e-08,
+      "loss": 0.6946,
+      "step": 15167
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.65361974277326,
+      "learning_rate": 2.232205368800533e-08,
+      "loss": 0.6592,
+      "step": 15168
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8153884383305823,
+      "learning_rate": 2.222432980762912e-08,
+      "loss": 0.7348,
+      "step": 15169
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5268097645815901,
+      "learning_rate": 2.2126819833358516e-08,
+      "loss": 0.7988,
+      "step": 15170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7776911788964072,
+      "learning_rate": 2.2029523769385718e-08,
+      "loss": 0.6468,
+      "step": 15171
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6806491882094001,
+      "learning_rate": 2.1932441619890722e-08,
+      "loss": 0.7435,
+      "step": 15172
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8545070910078512,
+      "learning_rate": 2.1835573389045185e-08,
+      "loss": 0.6919,
+      "step": 15173
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6831601043283386,
+      "learning_rate": 2.1738919081012446e-08,
+      "loss": 0.7606,
+      "step": 15174
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6305003417792163,
+      "learning_rate": 2.1642478699944736e-08,
+      "loss": 0.7014,
+      "step": 15175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6235744718860998,
+      "learning_rate": 2.1546252249988186e-08,
+      "loss": 0.7664,
+      "step": 15176
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7350039378893107,
+      "learning_rate": 2.1450239735276158e-08,
+      "loss": 0.7175,
+      "step": 15177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6600174376180437,
+      "learning_rate": 2.135444115993479e-08,
+      "loss": 0.805,
+      "step": 15178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0432469543857303,
+      "learning_rate": 2.1258856528081906e-08,
+      "loss": 0.6145,
+      "step": 15179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7963886352418479,
+      "learning_rate": 2.116348584382366e-08,
+      "loss": 0.8445,
+      "step": 15180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7944732176415445,
+      "learning_rate": 2.1068329111259e-08,
+      "loss": 0.6651,
+      "step": 15181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.263345953428395,
+      "learning_rate": 2.097338633447743e-08,
+      "loss": 0.7203,
+      "step": 15182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3495833533730623,
+      "learning_rate": 2.087865751755791e-08,
+      "loss": 0.5931,
+      "step": 15183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6549582891796484,
+      "learning_rate": 2.0784142664571626e-08,
+      "loss": 0.6521,
+      "step": 15184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8735777251271974,
+      "learning_rate": 2.0689841779579777e-08,
+      "loss": 0.6739,
+      "step": 15185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0379026309799153,
+      "learning_rate": 2.059575486663523e-08,
+      "loss": 0.6371,
+      "step": 15186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7955156429208394,
+      "learning_rate": 2.0501881929780865e-08,
+      "loss": 0.8057,
+      "step": 15187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5279770097500833,
+      "learning_rate": 2.040822297305012e-08,
+      "loss": 0.5797,
+      "step": 15188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5468929967308291,
+      "learning_rate": 2.031477800046866e-08,
+      "loss": 0.636,
+      "step": 15189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5820783729340164,
+      "learning_rate": 2.0221547016051614e-08,
+      "loss": 0.6643,
+      "step": 15190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.772099992358545,
+      "learning_rate": 2.012853002380466e-08,
+      "loss": 0.6545,
+      "step": 15191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.619020969041474,
+      "learning_rate": 2.003572702772627e-08,
+      "loss": 0.6548,
+      "step": 15192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.55189113905674,
+      "learning_rate": 1.9943138031803255e-08,
+      "loss": 0.6912,
+      "step": 15193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5665053491537948,
+      "learning_rate": 1.9850763040014654e-08,
+      "loss": 0.7152,
+      "step": 15194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8464093968632662,
+      "learning_rate": 1.9758602056330623e-08,
+      "loss": 0.6321,
+      "step": 15195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7051478316444548,
+      "learning_rate": 1.9666655084710772e-08,
+      "loss": 0.7311,
+      "step": 15196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0602356433156912,
+      "learning_rate": 1.957492212910639e-08,
+      "loss": 0.6085,
+      "step": 15197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9301149161417648,
+      "learning_rate": 1.9483403193459873e-08,
+      "loss": 0.6477,
+      "step": 15198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.634807255330644,
+      "learning_rate": 1.939209828170363e-08,
+      "loss": 0.647,
+      "step": 15199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0618653376875122,
+      "learning_rate": 1.9301007397761197e-08,
+      "loss": 0.6491,
+      "step": 15200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4539400453522728,
+      "learning_rate": 1.9210130545547213e-08,
+      "loss": 0.7688,
+      "step": 15201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.097584217926621,
+      "learning_rate": 1.9119467728966888e-08,
+      "loss": 0.6821,
+      "step": 15202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8323436052511246,
+      "learning_rate": 1.9029018951915446e-08,
+      "loss": 0.6455,
+      "step": 15203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5511298752302058,
+      "learning_rate": 1.8938784218281435e-08,
+      "loss": 0.6742,
+      "step": 15204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4030339243493,
+      "learning_rate": 1.884876353194065e-08,
+      "loss": 0.7543,
+      "step": 15205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2074323061681957,
+      "learning_rate": 1.875895689676166e-08,
+      "loss": 0.6592,
+      "step": 15206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0517407170187982,
+      "learning_rate": 1.866936431660471e-08,
+      "loss": 0.7185,
+      "step": 15207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7588616437941458,
+      "learning_rate": 1.8579985795318946e-08,
+      "loss": 0.7272,
+      "step": 15208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6987106642511458,
+      "learning_rate": 1.849082133674518e-08,
+      "loss": 0.8073,
+      "step": 15209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9315036664247562,
+      "learning_rate": 1.8401870944715903e-08,
+      "loss": 0.8304,
+      "step": 15210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5366650562990376,
+      "learning_rate": 1.8313134623051955e-08,
+      "loss": 0.7554,
+      "step": 15211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8181859231253872,
+      "learning_rate": 1.822461237556805e-08,
+      "loss": 0.6359,
+      "step": 15212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8641723155888175,
+      "learning_rate": 1.813630420606727e-08,
+      "loss": 0.712,
+      "step": 15213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7300337263392365,
+      "learning_rate": 1.80482101183449e-08,
+      "loss": 0.7206,
+      "step": 15214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5520943922375277,
+      "learning_rate": 1.7960330116186807e-08,
+      "loss": 0.687,
+      "step": 15215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.975795247864811,
+      "learning_rate": 1.7872664203368305e-08,
+      "loss": 0.6535,
+      "step": 15216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7750526872108965,
+      "learning_rate": 1.7785212383656934e-08,
+      "loss": 0.7569,
+      "step": 15217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6663925684461616,
+      "learning_rate": 1.7697974660811357e-08,
+      "loss": 0.6879,
+      "step": 15218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3963926722344306,
+      "learning_rate": 1.761095103858024e-08,
+      "loss": 0.6257,
+      "step": 15219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7262501330038371,
+      "learning_rate": 1.7524141520702808e-08,
+      "loss": 0.6008,
+      "step": 15220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9735702133667872,
+      "learning_rate": 1.743754611090942e-08,
+      "loss": 0.6984,
+      "step": 15221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7383536983976478,
+      "learning_rate": 1.7351164812921538e-08,
+      "loss": 0.6693,
+      "step": 15222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6657827857458942,
+      "learning_rate": 1.726499763045064e-08,
+      "loss": 0.783,
+      "step": 15223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5591469357903982,
+      "learning_rate": 1.717904456720043e-08,
+      "loss": 0.723,
+      "step": 15224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2424078903339333,
+      "learning_rate": 1.7093305626864065e-08,
+      "loss": 0.6902,
+      "step": 15225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6318863164219184,
+      "learning_rate": 1.7007780813125817e-08,
+      "loss": 0.618,
+      "step": 15226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.121281019028498,
+      "learning_rate": 1.692247012966053e-08,
+      "loss": 0.808,
+      "step": 15227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.745865853618489,
+      "learning_rate": 1.6837373580135265e-08,
+      "loss": 0.6748,
+      "step": 15228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9959318824503356,
+      "learning_rate": 1.675249116820543e-08,
+      "loss": 0.7476,
+      "step": 15229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.1054495301721685,
+      "learning_rate": 1.666782289751978e-08,
+      "loss": 0.677,
+      "step": 15230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4663432663573976,
+      "learning_rate": 1.6583368771716513e-08,
+      "loss": 0.8641,
+      "step": 15231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8650861074929945,
+      "learning_rate": 1.6499128794423836e-08,
+      "loss": 0.6294,
+      "step": 15232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7236659043168576,
+      "learning_rate": 1.6415102969263297e-08,
+      "loss": 0.6322,
+      "step": 15233
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5178862963982946,
+      "learning_rate": 1.6331291299844233e-08,
+      "loss": 0.7197,
+      "step": 15234
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9525402265650795,
+      "learning_rate": 1.6247693789768758e-08,
+      "loss": 0.7287,
+      "step": 15235
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5886304860130807,
+      "learning_rate": 1.616431044262956e-08,
+      "loss": 0.6729,
+      "step": 15236
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.782174022724717,
+      "learning_rate": 1.6081141262009326e-08,
+      "loss": 0.7552,
+      "step": 15237
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9611262774339335,
+      "learning_rate": 1.5998186251482416e-08,
+      "loss": 0.7255,
+      "step": 15238
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8219512802260156,
+      "learning_rate": 1.5915445414613208e-08,
+      "loss": 0.7299,
+      "step": 15239
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8628714315551047,
+      "learning_rate": 1.5832918754958293e-08,
+      "loss": 0.7329,
+      "step": 15240
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.898257766199161,
+      "learning_rate": 1.5750606276062063e-08,
+      "loss": 0.6653,
+      "step": 15241
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7778496295050017,
+      "learning_rate": 1.5668507981463354e-08,
+      "loss": 0.6678,
+      "step": 15242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7127991225522978,
+      "learning_rate": 1.5586623874689345e-08,
+      "loss": 0.7901,
+      "step": 15243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8683766914009183,
+      "learning_rate": 1.550495395925944e-08,
+      "loss": 0.7239,
+      "step": 15244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8626866307088528,
+      "learning_rate": 1.54234982386825e-08,
+      "loss": 0.6694,
+      "step": 15245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6538241620488798,
+      "learning_rate": 1.5342256716459058e-08,
+      "loss": 0.6968,
+      "step": 15246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6544491181694858,
+      "learning_rate": 1.5261229396079657e-08,
+      "loss": 0.7562,
+      "step": 15247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3012982756498874,
+      "learning_rate": 1.518041628102762e-08,
+      "loss": 0.7152,
+      "step": 15248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2701634473923673,
+      "learning_rate": 1.5099817374774615e-08,
+      "loss": 0.7338,
+      "step": 15249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9924537023036006,
+      "learning_rate": 1.501943268078454e-08,
+      "loss": 0.7228,
+      "step": 15250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5828381638636164,
+      "learning_rate": 1.4939262202511297e-08,
+      "loss": 0.6804,
+      "step": 15251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7060426598431508,
+      "learning_rate": 1.4859305943400459e-08,
+      "loss": 0.7295,
+      "step": 15252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6748748379893321,
+      "learning_rate": 1.4779563906888172e-08,
+      "loss": 0.8208,
+      "step": 15253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.639368622648066,
+      "learning_rate": 1.4700036096400028e-08,
+      "loss": 0.7285,
+      "step": 15254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8411516195054292,
+      "learning_rate": 1.4620722515354957e-08,
+      "loss": 0.8702,
+      "step": 15255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.265607542485906,
+      "learning_rate": 1.454162316715968e-08,
+      "loss": 0.6252,
+      "step": 15256
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9126973082580585,
+      "learning_rate": 1.4462738055214254e-08,
+      "loss": 0.6581,
+      "step": 15257
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.768523599507953,
+      "learning_rate": 1.4384067182908746e-08,
+      "loss": 0.6172,
+      "step": 15258
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2933561897244494,
+      "learning_rate": 1.4305610553623228e-08,
+      "loss": 0.7276,
+      "step": 15259
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8340132291433477,
+      "learning_rate": 1.4227368170728894e-08,
+      "loss": 0.6791,
+      "step": 15260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4562767924520237,
+      "learning_rate": 1.4149340037589166e-08,
+      "loss": 0.7373,
+      "step": 15261
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8414897593113217,
+      "learning_rate": 1.4071526157556358e-08,
+      "loss": 0.848,
+      "step": 15262
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6351908284356338,
+      "learning_rate": 1.3993926533973912e-08,
+      "loss": 0.6037,
+      "step": 15263
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7288446647163274,
+      "learning_rate": 1.3916541170176934e-08,
+      "loss": 0.7562,
+      "step": 15264
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.159694796202555,
+      "learning_rate": 1.3839370069490543e-08,
+      "loss": 0.6387,
+      "step": 15265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7501541802438516,
+      "learning_rate": 1.3762413235232086e-08,
+      "loss": 0.7746,
+      "step": 15266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.276468598124349,
+      "learning_rate": 1.3685670670706697e-08,
+      "loss": 0.7961,
+      "step": 15267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8845184455624289,
+      "learning_rate": 1.3609142379213957e-08,
+      "loss": 0.7884,
+      "step": 15268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.576441396712989,
+      "learning_rate": 1.3532828364041239e-08,
+      "loss": 0.6137,
+      "step": 15269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8015124791162929,
+      "learning_rate": 1.3456728628468141e-08,
+      "loss": 0.7292,
+      "step": 15270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.010986498436129,
+      "learning_rate": 1.3380843175765379e-08,
+      "loss": 0.7958,
+      "step": 15271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.626663131595579,
+      "learning_rate": 1.3305172009193678e-08,
+      "loss": 0.7777,
+      "step": 15272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6761672708215611,
+      "learning_rate": 1.3229715132004328e-08,
+      "loss": 0.5463,
+      "step": 15273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.985200577589822,
+      "learning_rate": 1.3154472547440289e-08,
+      "loss": 0.7795,
+      "step": 15274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8097638939305012,
+      "learning_rate": 1.3079444258735086e-08,
+      "loss": 0.8148,
+      "step": 15275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6700197136244939,
+      "learning_rate": 1.3004630269112806e-08,
+      "loss": 0.7377,
+      "step": 15276
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.737894756484138,
+      "learning_rate": 1.29300305817881e-08,
+      "loss": 0.6946,
+      "step": 15277
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6425237132140917,
+      "learning_rate": 1.2855645199966738e-08,
+      "loss": 0.6898,
+      "step": 15278
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5909437506862874,
+      "learning_rate": 1.2781474126845051e-08,
+      "loss": 0.6254,
+      "step": 15279
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9431077312655618,
+      "learning_rate": 1.270751736561049e-08,
+      "loss": 0.8183,
+      "step": 15280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7019424276099044,
+      "learning_rate": 1.2633774919441622e-08,
+      "loss": 0.7192,
+      "step": 15281
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4625669023845183,
+      "learning_rate": 1.256024679150647e-08,
+      "loss": 0.5811,
+      "step": 15282
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7093400439282196,
+      "learning_rate": 1.2486932984965282e-08,
+      "loss": 0.6733,
+      "step": 15283
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0128518992832545,
+      "learning_rate": 1.241383350296832e-08,
+      "loss": 0.7712,
+      "step": 15284
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9691266147081923,
+      "learning_rate": 1.2340948348656955e-08,
+      "loss": 0.6717,
+      "step": 15285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.649056361520321,
+      "learning_rate": 1.2268277525163685e-08,
+      "loss": 0.7021,
+      "step": 15286
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7550246548139086,
+      "learning_rate": 1.21958210356099e-08,
+      "loss": 0.5551,
+      "step": 15287
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8532507880226283,
+      "learning_rate": 1.2123578883110887e-08,
+      "loss": 0.6192,
+      "step": 15288
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5427260377064693,
+      "learning_rate": 1.2051551070769719e-08,
+      "loss": 0.6745,
+      "step": 15289
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7850451009360797,
+      "learning_rate": 1.1979737601682252e-08,
+      "loss": 0.791,
+      "step": 15290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7826213798745607,
+      "learning_rate": 1.1908138478934905e-08,
+      "loss": 0.767,
+      "step": 15291
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7342168120040709,
+      "learning_rate": 1.1836753705602998e-08,
+      "loss": 0.7673,
+      "step": 15292
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.757155933569789,
+      "learning_rate": 1.1765583284755743e-08,
+      "loss": 0.6712,
+      "step": 15293
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.428217565214845,
+      "learning_rate": 1.1694627219450694e-08,
+      "loss": 0.6799,
+      "step": 15294
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.681013364662161,
+      "learning_rate": 1.1623885512737076e-08,
+      "loss": 0.7136,
+      "step": 15295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7020835715138374,
+      "learning_rate": 1.1553358167654683e-08,
+      "loss": 0.7058,
+      "step": 15296
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.514505647983024,
+      "learning_rate": 1.1483045187233866e-08,
+      "loss": 0.7597,
+      "step": 15297
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7192772451506226,
+      "learning_rate": 1.1412946574497208e-08,
+      "loss": 0.7475,
+      "step": 15298
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8815128090387385,
+      "learning_rate": 1.134306233245619e-08,
+      "loss": 0.6807,
+      "step": 15299
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9844027723148452,
+      "learning_rate": 1.1273392464114518e-08,
+      "loss": 0.645,
+      "step": 15300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9681176219246157,
+      "learning_rate": 1.1203936972465356e-08,
+      "loss": 0.6437,
+      "step": 15301
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.550047817565005,
+      "learning_rate": 1.1134695860493539e-08,
+      "loss": 0.6493,
+      "step": 15302
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0574669327846062,
+      "learning_rate": 1.1065669131175016e-08,
+      "loss": 0.6761,
+      "step": 15303
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6065664426563115,
+      "learning_rate": 1.0996856787475197e-08,
+      "loss": 0.704,
+      "step": 15304
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.572909731702415,
+      "learning_rate": 1.0928258832351712e-08,
+      "loss": 0.7397,
+      "step": 15305
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.061847623467374,
+      "learning_rate": 1.0859875268752206e-08,
+      "loss": 0.7784,
+      "step": 15306
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2030169397591406,
+      "learning_rate": 1.079170609961544e-08,
+      "loss": 0.6724,
+      "step": 15307
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7274135853599137,
+      "learning_rate": 1.072375132787018e-08,
+      "loss": 0.6897,
+      "step": 15308
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9924988209562944,
+      "learning_rate": 1.0656010956437979e-08,
+      "loss": 0.8462,
+      "step": 15309
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6065039666867955,
+      "learning_rate": 1.0588484988228175e-08,
+      "loss": 0.8209,
+      "step": 15310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.695236372529407,
+      "learning_rate": 1.0521173426142894e-08,
+      "loss": 0.7277,
+      "step": 15311
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0732899119431276,
+      "learning_rate": 1.0454076273075375e-08,
+      "loss": 0.7285,
+      "step": 15312
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2264025150025257,
+      "learning_rate": 1.0387193531908313e-08,
+      "loss": 0.6693,
+      "step": 15313
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6919495038126193,
+      "learning_rate": 1.0320525205516629e-08,
+      "loss": 0.719,
+      "step": 15314
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8739193683749804,
+      "learning_rate": 1.025407129676359e-08,
+      "loss": 0.6849,
+      "step": 15315
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.743715708726564,
+      "learning_rate": 1.018783180850691e-08,
+      "loss": 0.6776,
+      "step": 15316
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.922834542299846,
+      "learning_rate": 1.0121806743591534e-08,
+      "loss": 0.7538,
+      "step": 15317
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0733301157236266,
+      "learning_rate": 1.0055996104854637e-08,
+      "loss": 0.655,
+      "step": 15318
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6338286834757862,
+      "learning_rate": 9.990399895125624e-09,
+      "loss": 0.6692,
+      "step": 15319
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6742458728623477,
+      "learning_rate": 9.925018117221686e-09,
+      "loss": 0.7917,
+      "step": 15320
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5987872473287545,
+      "learning_rate": 9.859850773953905e-09,
+      "loss": 0.5753,
+      "step": 15321
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0598064445163335,
+      "learning_rate": 9.794897868121711e-09,
+      "loss": 0.5943,
+      "step": 15322
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.591891323451959,
+      "learning_rate": 9.73015940251676e-09,
+      "loss": 0.588,
+      "step": 15323
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7082342296562505,
+      "learning_rate": 9.665635379920157e-09,
+      "loss": 0.6285,
+      "step": 15324
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3591043638420564,
+      "learning_rate": 9.6013258031058e-09,
+      "loss": 0.6763,
+      "step": 15325
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8461979459040407,
+      "learning_rate": 9.537230674836473e-09,
+      "loss": 0.6532,
+      "step": 15326
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6297218682538708,
+      "learning_rate": 9.473349997867198e-09,
+      "loss": 0.7249,
+      "step": 15327
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0789254589517796,
+      "learning_rate": 9.409683774941891e-09,
+      "loss": 0.767,
+      "step": 15328
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5761885705689629,
+      "learning_rate": 9.346232008797252e-09,
+      "loss": 0.6624,
+      "step": 15329
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4850466313552126,
+      "learning_rate": 9.282994702159986e-09,
+      "loss": 0.7521,
+      "step": 15330
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7631977318895782,
+      "learning_rate": 9.219971857747368e-09,
+      "loss": 0.6982,
+      "step": 15331
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.500657613435917,
+      "learning_rate": 9.157163478267783e-09,
+      "loss": 0.7961,
+      "step": 15332
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.53285520930874,
+      "learning_rate": 9.094569566419632e-09,
+      "loss": 0.7788,
+      "step": 15333
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7956275116405789,
+      "learning_rate": 9.032190124893536e-09,
+      "loss": 0.7618,
+      "step": 15334
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7660108984818272,
+      "learning_rate": 8.970025156369578e-09,
+      "loss": 0.7129,
+      "step": 15335
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6010774969830641,
+      "learning_rate": 8.908074663519505e-09,
+      "loss": 0.6523,
+      "step": 15336
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6370358298114336,
+      "learning_rate": 8.846338649005082e-09,
+      "loss": 0.7081,
+      "step": 15337
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.833305171090279,
+      "learning_rate": 8.784817115479183e-09,
+      "loss": 0.7855,
+      "step": 15338
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8228297406696439,
+      "learning_rate": 8.723510065585806e-09,
+      "loss": 0.7972,
+      "step": 15339
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.688610249395675,
+      "learning_rate": 8.662417501959507e-09,
+      "loss": 0.8113,
+      "step": 15340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.67317540076708,
+      "learning_rate": 8.601539427225414e-09,
+      "loss": 0.7273,
+      "step": 15341
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.993323911143876,
+      "learning_rate": 8.540875843999208e-09,
+      "loss": 0.8025,
+      "step": 15342
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7025390739152853,
+      "learning_rate": 8.480426754888805e-09,
+      "loss": 0.6644,
+      "step": 15343
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7233819157389165,
+      "learning_rate": 8.42019216249046e-09,
+      "loss": 0.7524,
+      "step": 15344
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1326454108476245,
+      "learning_rate": 8.360172069393768e-09,
+      "loss": 0.6604,
+      "step": 15345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2876484048429093,
+      "learning_rate": 8.300366478177779e-09,
+      "loss": 0.7208,
+      "step": 15346
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2042521174987735,
+      "learning_rate": 8.240775391411548e-09,
+      "loss": 0.6427,
+      "step": 15347
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6482701521917478,
+      "learning_rate": 8.18139881165636e-09,
+      "loss": 0.806,
+      "step": 15348
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8152173701561771,
+      "learning_rate": 8.122236741464618e-09,
+      "loss": 0.6856,
+      "step": 15349
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.001299575151384,
+      "learning_rate": 8.063289183377065e-09,
+      "loss": 0.7477,
+      "step": 15350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.135466359873491,
+      "learning_rate": 8.004556139927788e-09,
+      "loss": 0.6824,
+      "step": 15351
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9353432307901806,
+      "learning_rate": 7.946037613640878e-09,
+      "loss": 0.6635,
+      "step": 15352
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.935804287858141,
+      "learning_rate": 7.887733607030435e-09,
+      "loss": 0.6424,
+      "step": 15353
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7369735826618125,
+      "learning_rate": 7.82964412260223e-09,
+      "loss": 0.814,
+      "step": 15354
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7719408263801424,
+      "learning_rate": 7.77176916285205e-09,
+      "loss": 0.676,
+      "step": 15355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8258042947330864,
+      "learning_rate": 7.714108730267345e-09,
+      "loss": 0.7256,
+      "step": 15356
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6493072608487191,
+      "learning_rate": 7.656662827326134e-09,
+      "loss": 0.6943,
+      "step": 15357
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.564971004274878,
+      "learning_rate": 7.599431456495888e-09,
+      "loss": 0.6186,
+      "step": 15358
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9933228886395604,
+      "learning_rate": 7.542414620237414e-09,
+      "loss": 0.6244,
+      "step": 15359
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0996453815823466,
+      "learning_rate": 7.485612321000424e-09,
+      "loss": 0.6925,
+      "step": 15360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5536090033839676,
+      "learning_rate": 7.4290245612246294e-09,
+      "loss": 0.7319,
+      "step": 15361
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1156670956588481,
+      "learning_rate": 7.372651343343639e-09,
+      "loss": 0.7523,
+      "step": 15362
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6252681235721478,
+      "learning_rate": 7.316492669778297e-09,
+      "loss": 0.596,
+      "step": 15363
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.598784061276021,
+      "learning_rate": 7.260548542943335e-09,
+      "loss": 0.6257,
+      "step": 15364
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2039831877426048,
+      "learning_rate": 7.2048189652412784e-09,
+      "loss": 0.6821,
+      "step": 15365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1199541990743977,
+      "learning_rate": 7.149303939067986e-09,
+      "loss": 0.7184,
+      "step": 15366
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8043557812398734,
+      "learning_rate": 7.094003466808774e-09,
+      "loss": 0.8084,
+      "step": 15367
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.659930408787882,
+      "learning_rate": 7.038917550840074e-09,
+      "loss": 0.7636,
+      "step": 15368
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.11662917986538,
+      "learning_rate": 6.984046193528881e-09,
+      "loss": 0.6149,
+      "step": 15369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4381913274600253,
+      "learning_rate": 6.9293893972338654e-09,
+      "loss": 0.6819,
+      "step": 15370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8411535931309473,
+      "learning_rate": 6.874947164302592e-09,
+      "loss": 0.7107,
+      "step": 15371
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5627016789709873,
+      "learning_rate": 6.820719497074857e-09,
+      "loss": 0.7256,
+      "step": 15372
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.536659241534014,
+      "learning_rate": 6.7667063978815724e-09,
+      "loss": 0.6211,
+      "step": 15373
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4972610388327598,
+      "learning_rate": 6.712907869043661e-09,
+      "loss": 0.7272,
+      "step": 15374
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0970761490436234,
+      "learning_rate": 6.659323912872606e-09,
+      "loss": 0.6875,
+      "step": 15375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7214972427776924,
+      "learning_rate": 6.60595453167101e-09,
+      "loss": 0.8049,
+      "step": 15376
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6089916654524397,
+      "learning_rate": 6.552799727732595e-09,
+      "loss": 0.7019,
+      "step": 15377
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8559966501689011,
+      "learning_rate": 6.4998595033410886e-09,
+      "loss": 0.7947,
+      "step": 15378
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0056890719531926,
+      "learning_rate": 6.447133860771893e-09,
+      "loss": 0.6127,
+      "step": 15379
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8391522149464594,
+      "learning_rate": 6.394622802290973e-09,
+      "loss": 0.7057,
+      "step": 15380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7706621686244397,
+      "learning_rate": 6.342326330153747e-09,
+      "loss": 0.6362,
+      "step": 15381
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6264150576764773,
+      "learning_rate": 6.290244446608973e-09,
+      "loss": 0.7206,
+      "step": 15382
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.733875220159213,
+      "learning_rate": 6.238377153893749e-09,
+      "loss": 0.7623,
+      "step": 15383
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9528000587471159,
+      "learning_rate": 6.186724454236847e-09,
+      "loss": 0.6679,
+      "step": 15384
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6977403853047268,
+      "learning_rate": 6.13528634985816e-09,
+      "loss": 0.6424,
+      "step": 15385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6175848157157358,
+      "learning_rate": 6.084062842968696e-09,
+      "loss": 0.7161,
+      "step": 15386
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9218972059672534,
+      "learning_rate": 6.0330539357689175e-09,
+      "loss": 0.7296,
+      "step": 15387
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.72686541053237,
+      "learning_rate": 5.98225963045096e-09,
+      "loss": 0.8288,
+      "step": 15388
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9569720323638629,
+      "learning_rate": 5.9316799291969654e-09,
+      "loss": 0.7679,
+      "step": 15389
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5601807050188723,
+      "learning_rate": 5.881314834181862e-09,
+      "loss": 0.7083,
+      "step": 15390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6318558786554656,
+      "learning_rate": 5.831164347568918e-09,
+      "loss": 0.8256,
+      "step": 15391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6629395985045872,
+      "learning_rate": 5.781228471513633e-09,
+      "loss": 0.6715,
+      "step": 15392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9486895698995084,
+      "learning_rate": 5.731507208160958e-09,
+      "loss": 0.7703,
+      "step": 15393
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.293121766767669,
+      "learning_rate": 5.682000559649181e-09,
+      "loss": 0.7397,
+      "step": 15394
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.527255502276372,
+      "learning_rate": 5.63270852810438e-09,
+      "loss": 0.7953,
+      "step": 15395
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6481742918302664,
+      "learning_rate": 5.583631115644861e-09,
+      "loss": 0.7662,
+      "step": 15396
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.582392148846149,
+      "learning_rate": 5.534768324380046e-09,
+      "loss": 0.6805,
+      "step": 15397
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5952856509282818,
+      "learning_rate": 5.486120156409924e-09,
+      "loss": 0.7179,
+      "step": 15398
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.655604318820946,
+      "learning_rate": 5.437686613823934e-09,
+      "loss": 0.6542,
+      "step": 15399
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7067761091336249,
+      "learning_rate": 5.389467698704298e-09,
+      "loss": 0.6668,
+      "step": 15400
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.705747971814269,
+      "learning_rate": 5.341463413123249e-09,
+      "loss": 0.8724,
+      "step": 15401
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9514623905745083,
+      "learning_rate": 5.29367375914247e-09,
+      "loss": 0.6218,
+      "step": 15402
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3072700431674362,
+      "learning_rate": 5.246098738816985e-09,
+      "loss": 0.8759,
+      "step": 15403
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6055926743737738,
+      "learning_rate": 5.198738354190158e-09,
+      "loss": 0.6495,
+      "step": 15404
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7226573443813502,
+      "learning_rate": 5.151592607298139e-09,
+      "loss": 0.6943,
+      "step": 15405
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5428678728585123,
+      "learning_rate": 5.104661500165975e-09,
+      "loss": 0.5755,
+      "step": 15406
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1176633943617287,
+      "learning_rate": 5.057945034810385e-09,
+      "loss": 0.6656,
+      "step": 15407
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2803568252987894,
+      "learning_rate": 5.0114432132397636e-09,
+      "loss": 0.6235,
+      "step": 15408
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5692708143888083,
+      "learning_rate": 4.9651560374514015e-09,
+      "loss": 0.7101,
+      "step": 15409
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1248010955568994,
+      "learning_rate": 4.919083509434819e-09,
+      "loss": 0.7445,
+      "step": 15410
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7791949300349967,
+      "learning_rate": 4.873225631170098e-09,
+      "loss": 0.5616,
+      "step": 15411
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9723841202938641,
+      "learning_rate": 4.827582404627884e-09,
+      "loss": 0.6454,
+      "step": 15412
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8028946809530648,
+      "learning_rate": 4.782153831768832e-09,
+      "loss": 0.6679,
+      "step": 15413
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9803137857488167,
+      "learning_rate": 4.736939914545824e-09,
+      "loss": 0.6772,
+      "step": 15414
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4914512338051327,
+      "learning_rate": 4.691940654901195e-09,
+      "loss": 0.6984,
+      "step": 15415
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.862944652165951,
+      "learning_rate": 4.647156054769508e-09,
+      "loss": 0.6539,
+      "step": 15416
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.69291864489584,
+      "learning_rate": 4.602586116074781e-09,
+      "loss": 0.7688,
+      "step": 15417
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.527035235450515,
+      "learning_rate": 4.558230840731592e-09,
+      "loss": 0.6106,
+      "step": 15418
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.721968307271597,
+      "learning_rate": 4.514090230647305e-09,
+      "loss": 0.7267,
+      "step": 15419
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8000505774819175,
+      "learning_rate": 4.470164287717627e-09,
+      "loss": 0.5293,
+      "step": 15420
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0437034760584452,
+      "learning_rate": 4.4264530138310445e-09,
+      "loss": 0.5951,
+      "step": 15421
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.080983672200297,
+      "learning_rate": 4.382956410865502e-09,
+      "loss": 0.8152,
+      "step": 15422
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4895974131468084,
+      "learning_rate": 4.339674480690059e-09,
+      "loss": 0.7013,
+      "step": 15423
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6196872636530184,
+      "learning_rate": 4.296607225164895e-09,
+      "loss": 0.6988,
+      "step": 15424
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.771176849567606,
+      "learning_rate": 4.253754646140196e-09,
+      "loss": 0.6731,
+      "step": 15425
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0375115237035197,
+      "learning_rate": 4.211116745458377e-09,
+      "loss": 0.7367,
+      "step": 15426
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6006958731408323,
+      "learning_rate": 4.168693524950196e-09,
+      "loss": 0.7019,
+      "step": 15427
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7069376715231654,
+      "learning_rate": 4.1264849864403044e-09,
+      "loss": 0.7218,
+      "step": 15428
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1166085537521502,
+      "learning_rate": 4.084491131741697e-09,
+      "loss": 0.6388,
+      "step": 15429
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5721763508287216,
+      "learning_rate": 4.042711962658486e-09,
+      "loss": 0.6715,
+      "step": 15430
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7543698514309005,
+      "learning_rate": 4.001147480987011e-09,
+      "loss": 0.6376,
+      "step": 15431
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4644905249056053,
+      "learning_rate": 3.959797688513067e-09,
+      "loss": 0.7009,
+      "step": 15432
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0373472429599078,
+      "learning_rate": 3.9186625870130115e-09,
+      "loss": 0.7786,
+      "step": 15433
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8523379131845166,
+      "learning_rate": 3.877742178254873e-09,
+      "loss": 0.7033,
+      "step": 15434
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1790773795438412,
+      "learning_rate": 3.837036463997246e-09,
+      "loss": 0.7481,
+      "step": 15435
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4470648219431816,
+      "learning_rate": 3.796545445989286e-09,
+      "loss": 0.6458,
+      "step": 15436
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3444435769489766,
+      "learning_rate": 3.756269125970713e-09,
+      "loss": 0.6966,
+      "step": 15437
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.554326724303324,
+      "learning_rate": 3.7162075056729196e-09,
+      "loss": 0.7281,
+      "step": 15438
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7871009441601144,
+      "learning_rate": 3.6763605868167516e-09,
+      "loss": 0.5966,
+      "step": 15439
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6787291585033102,
+      "learning_rate": 3.636728371114173e-09,
+      "loss": 0.7172,
+      "step": 15440
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7715214583676266,
+      "learning_rate": 3.5973108602699313e-09,
+      "loss": 0.7653,
+      "step": 15441
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.581211269777275,
+      "learning_rate": 3.558108055976006e-09,
+      "loss": 0.6595,
+      "step": 15442
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8904155102183244,
+      "learning_rate": 3.5191199599182712e-09,
+      "loss": 0.7939,
+      "step": 15443
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6804515461979126,
+      "learning_rate": 3.4803465737714983e-09,
+      "loss": 0.7813,
+      "step": 15444
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6291418334730636,
+      "learning_rate": 3.4417878992021315e-09,
+      "loss": 0.8996,
+      "step": 15445
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6495340877087088,
+      "learning_rate": 3.403443937867179e-09,
+      "loss": 0.6873,
+      "step": 15446
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6210772031555218,
+      "learning_rate": 3.365314691414212e-09,
+      "loss": 0.7517,
+      "step": 15447
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7072870423663185,
+      "learning_rate": 3.3274001614819195e-09,
+      "loss": 0.6155,
+      "step": 15448
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8046086480903865,
+      "learning_rate": 3.289700349698999e-09,
+      "loss": 0.6896,
+      "step": 15449
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6787262827640355,
+      "learning_rate": 3.252215257686375e-09,
+      "loss": 0.7725,
+      "step": 15450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4850720228594325,
+      "learning_rate": 3.2149448870544273e-09,
+      "loss": 0.7807,
+      "step": 15451
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.05608289046102,
+      "learning_rate": 3.1778892394046525e-09,
+      "loss": 0.618,
+      "step": 15452
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8972995132365165,
+      "learning_rate": 3.1410483163291093e-09,
+      "loss": 0.6663,
+      "step": 15453
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.189518701100943,
+      "learning_rate": 3.104422119411532e-09,
+      "loss": 0.7554,
+      "step": 15454
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7283990096598136,
+      "learning_rate": 3.0680106502256613e-09,
+      "loss": 0.6961,
+      "step": 15455
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8531310255389404,
+      "learning_rate": 3.0318139103363564e-09,
+      "loss": 0.67,
+      "step": 15456
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5398916740634077,
+      "learning_rate": 2.9958319012984847e-09,
+      "loss": 0.5625,
+      "step": 15457
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6432013131668384,
+      "learning_rate": 2.9600646246591425e-09,
+      "loss": 0.6246,
+      "step": 15458
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0889694616494041,
+      "learning_rate": 2.9245120819543226e-09,
+      "loss": 0.6397,
+      "step": 15459
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6520521604573497,
+      "learning_rate": 2.8891742747128025e-09,
+      "loss": 0.6409,
+      "step": 15460
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7073311174026968,
+      "learning_rate": 2.854051204452257e-09,
+      "loss": 0.8077,
+      "step": 15461
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0488640014772512,
+      "learning_rate": 2.8191428726820347e-09,
+      "loss": 0.6727,
+      "step": 15462
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5941826696920678,
+      "learning_rate": 2.7844492809031567e-09,
+      "loss": 0.7508,
+      "step": 15463
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9714588446881708,
+      "learning_rate": 2.749970430605542e-09,
+      "loss": 0.6454,
+      "step": 15464
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.711204054534621,
+      "learning_rate": 2.715706323271339e-09,
+      "loss": 0.6801,
+      "step": 15465
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5636749878595804,
+      "learning_rate": 2.681656960372703e-09,
+      "loss": 0.6711,
+      "step": 15466
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5556603297033103,
+      "learning_rate": 2.6478223433734627e-09,
+      "loss": 0.6673,
+      "step": 15467
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6178174260157454,
+      "learning_rate": 2.614202473726346e-09,
+      "loss": 0.7282,
+      "step": 15468
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6547635723517604,
+      "learning_rate": 2.5807973528768626e-09,
+      "loss": 0.706,
+      "step": 15469
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1920920241158537,
+      "learning_rate": 2.547606982260531e-09,
+      "loss": 0.6756,
+      "step": 15470
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1854059294249817,
+      "learning_rate": 2.5146313633028775e-09,
+      "loss": 0.6967,
+      "step": 15471
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2090365564028136,
+      "learning_rate": 2.481870497422212e-09,
+      "loss": 0.6285,
+      "step": 15472
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.643440726283558,
+      "learning_rate": 2.449324386025187e-09,
+      "loss": 0.7091,
+      "step": 15473
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.286092268887532,
+      "learning_rate": 2.416993030511239e-09,
+      "loss": 0.8042,
+      "step": 15474
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6208865546812838,
+      "learning_rate": 2.3848764322687014e-09,
+      "loss": 0.7487,
+      "step": 15475
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5191190768836234,
+      "learning_rate": 2.3529745926786916e-09,
+      "loss": 0.6849,
+      "step": 15476
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.815702905335521,
+      "learning_rate": 2.3212875131117805e-09,
+      "loss": 0.6956,
+      "step": 15477
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8231799471738608,
+      "learning_rate": 2.2898151949296566e-09,
+      "loss": 0.687,
+      "step": 15478
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3485137905532723,
+      "learning_rate": 2.258557639484571e-09,
+      "loss": 0.6977,
+      "step": 15479
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.686190499097588,
+      "learning_rate": 2.227514848119894e-09,
+      "loss": 0.6343,
+      "step": 15480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7244779953860148,
+      "learning_rate": 2.196686822169003e-09,
+      "loss": 0.7567,
+      "step": 15481
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.469076900547598,
+      "learning_rate": 2.1660735629580598e-09,
+      "loss": 0.6927,
+      "step": 15482
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.014766257953631,
+      "learning_rate": 2.1356750718010133e-09,
+      "loss": 0.6497,
+      "step": 15483
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.668815681067392,
+      "learning_rate": 2.1054913500051512e-09,
+      "loss": 0.6911,
+      "step": 15484
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7590794137612913,
+      "learning_rate": 2.0755223988672136e-09,
+      "loss": 0.6712,
+      "step": 15485
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1939074585408564,
+      "learning_rate": 2.0457682196750595e-09,
+      "loss": 0.6134,
+      "step": 15486
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0419239742122266,
+      "learning_rate": 2.0162288137071105e-09,
+      "loss": 0.8978,
+      "step": 15487
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6922229043090213,
+      "learning_rate": 1.9869041822329073e-09,
+      "loss": 0.7399,
+      "step": 15488
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6567566421748527,
+      "learning_rate": 1.957794326513107e-09,
+      "loss": 0.6689,
+      "step": 15489
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8599132374520977,
+      "learning_rate": 1.928899247797822e-09,
+      "loss": 0.8022,
+      "step": 15490
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6000325417267216,
+      "learning_rate": 1.9002189473288356e-09,
+      "loss": 0.778,
+      "step": 15491
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6228564184762548,
+      "learning_rate": 1.8717534263390513e-09,
+      "loss": 0.6718,
+      "step": 15492
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6973342239082527,
+      "learning_rate": 1.8435026860513794e-09,
+      "loss": 0.7184,
+      "step": 15493
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6244647677063107,
+      "learning_rate": 1.8154667276798488e-09,
+      "loss": 0.7022,
+      "step": 15494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6815554857947947,
+      "learning_rate": 1.7876455524290514e-09,
+      "loss": 0.7862,
+      "step": 15495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7411309691127956,
+      "learning_rate": 1.7600391614952529e-09,
+      "loss": 0.7195,
+      "step": 15496
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.178698967067727,
+      "learning_rate": 1.7326475560636157e-09,
+      "loss": 0.7179,
+      "step": 15497
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.714438247895917,
+      "learning_rate": 1.7054707373126423e-09,
+      "loss": 0.6921,
+      "step": 15498
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8994923566674546,
+      "learning_rate": 1.6785087064086213e-09,
+      "loss": 0.8016,
+      "step": 15499
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8129708359924364,
+      "learning_rate": 1.6517614645111812e-09,
+      "loss": 0.5823,
+      "step": 15500
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8390979738242212,
+      "learning_rate": 1.6252290127699577e-09,
+      "loss": 0.8228,
+      "step": 15501
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6608985017380185,
+      "learning_rate": 1.5989113523240396e-09,
+      "loss": 0.7602,
+      "step": 15502
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7860014640225228,
+      "learning_rate": 1.5728084843052992e-09,
+      "loss": 0.6984,
+      "step": 15503
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.035702994185283,
+      "learning_rate": 1.546920409834507e-09,
+      "loss": 0.6066,
+      "step": 15504
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8719101195825592,
+      "learning_rate": 1.5212471300252163e-09,
+      "loss": 0.6128,
+      "step": 15505
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4698021878676149,
+      "learning_rate": 1.4957886459798787e-09,
+      "loss": 0.7663,
+      "step": 15506
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6326473627514553,
+      "learning_rate": 1.4705449587931742e-09,
+      "loss": 0.8133,
+      "step": 15507
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6281814032112711,
+      "learning_rate": 1.4455160695492354e-09,
+      "loss": 0.7272,
+      "step": 15508
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8313253742706601,
+      "learning_rate": 1.4207019793238686e-09,
+      "loss": 0.7384,
+      "step": 15509
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5909765688798816,
+      "learning_rate": 1.3961026891828878e-09,
+      "loss": 0.706,
+      "step": 15510
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4552164327904193,
+      "learning_rate": 1.3717182001843355e-09,
+      "loss": 0.7182,
+      "step": 15511
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6433684179759307,
+      "learning_rate": 1.347548513375707e-09,
+      "loss": 0.636,
+      "step": 15512
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6768756001623393,
+      "learning_rate": 1.3235936297956165e-09,
+      "loss": 0.8365,
+      "step": 15513
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.396181885707452,
+      "learning_rate": 1.299853550472685e-09,
+      "loss": 0.6851,
+      "step": 15514
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8008479322104483,
+      "learning_rate": 1.276328276428318e-09,
+      "loss": 0.6584,
+      "step": 15515
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.618554106211368,
+      "learning_rate": 1.2530178086728183e-09,
+      "loss": 0.609,
+      "step": 15516
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6001471877872206,
+      "learning_rate": 1.2299221482076074e-09,
+      "loss": 0.634,
+      "step": 15517
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6024060039880954,
+      "learning_rate": 1.2070412960257793e-09,
+      "loss": 0.6526,
+      "step": 15518
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1332313682124293,
+      "learning_rate": 1.1843752531104368e-09,
+      "loss": 0.7263,
+      "step": 15519
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7034606674416828,
+      "learning_rate": 1.1619240204352455e-09,
+      "loss": 0.7719,
+      "step": 15520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.259781247685038,
+      "learning_rate": 1.1396875989655443e-09,
+      "loss": 0.6113,
+      "step": 15521
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.718758059462401,
+      "learning_rate": 1.1176659896561247e-09,
+      "loss": 0.8141,
+      "step": 15522
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8038779124184712,
+      "learning_rate": 1.0958591934534525e-09,
+      "loss": 0.6246,
+      "step": 15523
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.008997291628213,
+      "learning_rate": 1.0742672112951103e-09,
+      "loss": 0.7564,
+      "step": 15524
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9086899681073186,
+      "learning_rate": 1.052890044109245e-09,
+      "loss": 0.732,
+      "step": 15525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.713228902206267,
+      "learning_rate": 1.0317276928134557e-09,
+      "loss": 0.7813,
+      "step": 15526
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2081247701279343,
+      "learning_rate": 1.0107801583175702e-09,
+      "loss": 0.7046,
+      "step": 15527
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8202086691710624,
+      "learning_rate": 9.900474415219795e-10,
+      "loss": 0.7757,
+      "step": 15528
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.618687498931888,
+      "learning_rate": 9.695295433170826e-10,
+      "loss": 0.646,
+      "step": 15529
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6499391952319724,
+      "learning_rate": 9.492264645855065e-10,
+      "loss": 0.7052,
+      "step": 15530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.785673059033842,
+      "learning_rate": 9.291382061987763e-10,
+      "loss": 0.7864,
+      "step": 15531
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.615870175609887,
+      "learning_rate": 9.092647690206457e-10,
+      "loss": 0.5873,
+      "step": 15532
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2183805320479804,
+      "learning_rate": 8.896061539048762e-10,
+      "loss": 0.7115,
+      "step": 15533
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.109121625323622,
+      "learning_rate": 8.701623616963472e-10,
+      "loss": 0.7894,
+      "step": 15534
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.654682309939192,
+      "learning_rate": 8.509333932310571e-10,
+      "loss": 0.6951,
+      "step": 15535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.679069318832962,
+      "learning_rate": 8.319192493344563e-10,
+      "loss": 0.7384,
+      "step": 15536
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7639969505104203,
+      "learning_rate": 8.131199308236692e-10,
+      "loss": 0.6875,
+      "step": 15537
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6249133132271367,
+      "learning_rate": 7.945354385074933e-10,
+      "loss": 0.6751,
+      "step": 15538
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8734774099079787,
+      "learning_rate": 7.761657731836236e-10,
+      "loss": 0.6827,
+      "step": 15539
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9035474806584474,
+      "learning_rate": 7.580109356419841e-10,
+      "loss": 0.7132,
+      "step": 15540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5513839527992959,
+      "learning_rate": 7.400709266625062e-10,
+      "loss": 0.7069,
+      "step": 15541
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5865438459951797,
+      "learning_rate": 7.223457470156847e-10,
+      "loss": 0.6657,
+      "step": 15542
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7190360773548583,
+      "learning_rate": 7.04835397464243e-10,
+      "loss": 0.663,
+      "step": 15543
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8178486506057459,
+      "learning_rate": 6.87539878759802e-10,
+      "loss": 0.7303,
+      "step": 15544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8442218620967834,
+      "learning_rate": 6.704591916462111e-10,
+      "loss": 0.6134,
+      "step": 15545
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1021759152878057,
+      "learning_rate": 6.535933368567726e-10,
+      "loss": 0.7798,
+      "step": 15546
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6819827209706446,
+      "learning_rate": 6.369423151164622e-10,
+      "loss": 0.804,
+      "step": 15547
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5762283917688615,
+      "learning_rate": 6.205061271408186e-10,
+      "loss": 0.7424,
+      "step": 15548
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.842410050920086,
+      "learning_rate": 6.042847736364987e-10,
+      "loss": 0.6272,
+      "step": 15549
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.583373582207125,
+      "learning_rate": 5.882782553001676e-10,
+      "loss": 0.6921,
+      "step": 15550
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7536053653845454,
+      "learning_rate": 5.724865728201634e-10,
+      "loss": 0.7855,
+      "step": 15551
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7371093239534394,
+      "learning_rate": 5.569097268742773e-10,
+      "loss": 0.735,
+      "step": 15552
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.572237438402278,
+      "learning_rate": 5.415477181325291e-10,
+      "loss": 0.7073,
+      "step": 15553
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.827222707317945,
+      "learning_rate": 5.264005472549461e-10,
+      "loss": 0.7337,
+      "step": 15554
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7043633670094402,
+      "learning_rate": 5.114682148921191e-10,
+      "loss": 0.7945,
+      "step": 15555
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6201698673547087,
+      "learning_rate": 4.967507216863121e-10,
+      "loss": 0.6281,
+      "step": 15556
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7386874615513872,
+      "learning_rate": 4.822480682692421e-10,
+      "loss": 0.6422,
+      "step": 15557
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.850956005869514,
+      "learning_rate": 4.679602552648543e-10,
+      "loss": 0.735,
+      "step": 15558
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6727400902434058,
+      "learning_rate": 4.538872832865471e-10,
+      "loss": 0.7614,
+      "step": 15559
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.032807213277971,
+      "learning_rate": 4.4002915293939186e-10,
+      "loss": 0.6211,
+      "step": 15560
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0303747709314752,
+      "learning_rate": 4.2638586481846823e-10,
+      "loss": 0.6351,
+      "step": 15561
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6723089172618244,
+      "learning_rate": 4.1295741951108417e-10,
+      "loss": 0.6994,
+      "step": 15562
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4441979387498614,
+      "learning_rate": 3.9974381759289027e-10,
+      "loss": 0.7771,
+      "step": 15563
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.325472694323548,
+      "learning_rate": 3.867450596328759e-10,
+      "loss": 0.5265,
+      "step": 15564
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6331618553353593,
+      "learning_rate": 3.739611461889281e-10,
+      "loss": 0.6738,
+      "step": 15565
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5288407658303216,
+      "learning_rate": 3.6139207781060723e-10,
+      "loss": 0.5816,
+      "step": 15566
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5366242015908556,
+      "learning_rate": 3.490378550380369e-10,
+      "loss": 0.7516,
+      "step": 15567
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.36203853775454,
+      "learning_rate": 3.368984784024587e-10,
+      "loss": 0.6769,
+      "step": 15568
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8245390434848996,
+      "learning_rate": 3.2497394842512244e-10,
+      "loss": 0.7567,
+      "step": 15569
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6257818641355795,
+      "learning_rate": 3.1326426561839597e-10,
+      "loss": 0.7554,
+      "step": 15570
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.739711928951949,
+      "learning_rate": 3.017694304852103e-10,
+      "loss": 0.7389,
+      "step": 15571
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5628074817223725,
+      "learning_rate": 2.9048944352072505e-10,
+      "loss": 0.774,
+      "step": 15572
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.643511891911213,
+      "learning_rate": 2.794243052084422e-10,
+      "loss": 0.7647,
+      "step": 15573
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6521052638479472,
+      "learning_rate": 2.685740160240924e-10,
+      "loss": 0.746,
+      "step": 15574
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2459355520857636,
+      "learning_rate": 2.5793857643396924e-10,
+      "loss": 0.6823,
+      "step": 15575
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.836072699836164,
+      "learning_rate": 2.475179868960398e-10,
+      "loss": 0.7083,
+      "step": 15576
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7641969634391534,
+      "learning_rate": 2.3731224785661365e-10,
+      "loss": 0.7385,
+      "step": 15577
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.655039506459945,
+      "learning_rate": 2.273213597553392e-10,
+      "loss": 0.8675,
+      "step": 15578
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7594265665508733,
+      "learning_rate": 2.1754532302076247e-10,
+      "loss": 0.7393,
+      "step": 15579
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5496317529819905,
+      "learning_rate": 2.0798413807365802e-10,
+      "loss": 0.6451,
+      "step": 15580
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6181046528343321,
+      "learning_rate": 1.9863780532480836e-10,
+      "loss": 0.6288,
+      "step": 15581
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1357180599455075,
+      "learning_rate": 1.89506325175004e-10,
+      "loss": 0.6738,
+      "step": 15582
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.221707669829302,
+      "learning_rate": 1.80589698017819e-10,
+      "loss": 0.6966,
+      "step": 15583
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6219380675160686,
+      "learning_rate": 1.718879242357252e-10,
+      "loss": 0.6813,
+      "step": 15584
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5457883300702753,
+      "learning_rate": 1.6340100420286776e-10,
+      "loss": 0.5698,
+      "step": 15585
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2609625767180181,
+      "learning_rate": 1.5512893828395493e-10,
+      "loss": 0.6765,
+      "step": 15586
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6182409280689813,
+      "learning_rate": 1.470717268342581e-10,
+      "loss": 0.712,
+      "step": 15587
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6286323818506556,
+      "learning_rate": 1.392293702001668e-10,
+      "loss": 0.6633,
+      "step": 15588
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8528422404436786,
+      "learning_rate": 1.316018687191889e-10,
+      "loss": 0.8566,
+      "step": 15589
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8567826936531335,
+      "learning_rate": 1.2418922271772993e-10,
+      "loss": 0.7106,
+      "step": 15590
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8933189890449107,
+      "learning_rate": 1.1699143251553414e-10,
+      "loss": 0.6983,
+      "step": 15591
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9462684592610757,
+      "learning_rate": 1.1000849842179862e-10,
+      "loss": 0.8032,
+      "step": 15592
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1404553316549886,
+      "learning_rate": 1.032404207362836e-10,
+      "loss": 0.7318,
+      "step": 15593
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6531809734312513,
+      "learning_rate": 9.668719974986751e-11,
+      "loss": 0.705,
+      "step": 15594
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4385171675236768,
+      "learning_rate": 9.03488357439919e-11,
+      "loss": 0.7566,
+      "step": 15595
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.717913537055915,
+      "learning_rate": 8.422532899121649e-11,
+      "loss": 0.7017,
+      "step": 15596
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8140579113952746,
+      "learning_rate": 7.831667975466417e-11,
+      "loss": 0.7091,
+      "step": 15597
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6157123628871368,
+      "learning_rate": 7.2622888288576e-11,
+      "loss": 0.631,
+      "step": 15598
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8519962213379038,
+      "learning_rate": 6.714395483720105e-11,
+      "loss": 0.7261,
+      "step": 15599
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4832160915275885,
+      "learning_rate": 6.187987963590658e-11,
+      "loss": 0.6614,
+      "step": 15600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6468731810663617,
+      "learning_rate": 5.6830662911178115e-11,
+      "loss": 0.848,
+      "step": 15601
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0216882496494493,
+      "learning_rate": 5.199630488006424e-11,
+      "loss": 0.6905,
+      "step": 15602
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9880153485722383,
+      "learning_rate": 4.737680575017667e-11,
+      "loss": 0.6937,
+      "step": 15603
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7262776123117631,
+      "learning_rate": 4.297216571969021e-11,
+      "loss": 0.7014,
+      "step": 15604
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8369435971978536,
+      "learning_rate": 3.878238497845299e-11,
+      "loss": 0.75,
+      "step": 15605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6016223693841556,
+      "learning_rate": 3.4807463705766044e-11,
+      "loss": 0.7796,
+      "step": 15606
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0256475394598366,
+      "learning_rate": 3.104740207371393e-11,
+      "loss": 0.7815,
+      "step": 15607
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.169158545971192,
+      "learning_rate": 2.7502200242723874e-11,
+      "loss": 0.6172,
+      "step": 15608
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5072122683186275,
+      "learning_rate": 2.417185836545155e-11,
+      "loss": 0.7133,
+      "step": 15609
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0017685653774113,
+      "learning_rate": 2.1056376585115723e-11,
+      "loss": 0.6376,
+      "step": 15610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5930758967069176,
+      "learning_rate": 1.815575503549827e-11,
+      "loss": 0.6767,
+      "step": 15611
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0937629811693,
+      "learning_rate": 1.5469993840944163e-11,
+      "loss": 0.7774,
+      "step": 15612
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4771887753017219,
+      "learning_rate": 1.2999093117471717e-11,
+      "loss": 0.6775,
+      "step": 15613
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8114809784460506,
+      "learning_rate": 1.0743052971107225e-11,
+      "loss": 0.748,
+      "step": 15614
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8204219971275815,
+      "learning_rate": 8.701873498995205e-12,
+      "loss": 0.6342,
+      "step": 15615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5629472653237946,
+      "learning_rate": 6.875554787733052e-12,
+      "loss": 0.6991,
+      "step": 15616
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.717695069885182,
+      "learning_rate": 5.2640969172568225e-12,
+      "loss": 0.776,
+      "step": 15617
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5571153282871368,
+      "learning_rate": 3.867499955845233e-12,
+      "loss": 0.6967,
+      "step": 15618
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1233756854589823,
+      "learning_rate": 2.6857639640054387e-12,
+      "loss": 0.582,
+      "step": 15619
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8187601386287193,
+      "learning_rate": 1.7188889922525876e-12,
+      "loss": 0.7709,
+      "step": 15620
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2260351465029165,
+      "learning_rate": 9.668750822200423e-13,
+      "loss": 0.7189,
+      "step": 15621
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0613137538113637,
+      "learning_rate": 4.2972226665938255e-13,
+      "loss": 0.6465,
+      "step": 15622
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7407346905107464,
+      "learning_rate": 1.0743056777506866e-13,
+      "loss": 0.712,
+      "step": 15623
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.51028227774696,
+      "learning_rate": 0.0,
+      "loss": 0.6649,
+      "step": 15624
+    },
+    {
+      "epoch": 1.0,
+      "step": 15624,
+      "total_flos": 3270011191017472.0,
+      "train_loss": 0.7456011022958491,
+      "train_runtime": 61646.3002,
+      "train_samples_per_second": 16.221,
+      "train_steps_per_second": 0.253
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 15624,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 3270011191017472.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}