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": 0.9999679989759672,
+  "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": 73.22029480722578,
+      "learning_rate": 2.1321961620469085e-08,
+      "loss": 2.8463,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 60.89059177985938,
+      "learning_rate": 4.264392324093817e-08,
+      "loss": 2.8177,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.61284072376907,
+      "learning_rate": 6.396588486140725e-08,
+      "loss": 2.2864,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 60.063364744201685,
+      "learning_rate": 8.528784648187634e-08,
+      "loss": 2.7498,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 84.96413005685196,
+      "learning_rate": 1.0660980810234542e-07,
+      "loss": 2.8979,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 51.17412279315548,
+      "learning_rate": 1.279317697228145e-07,
+      "loss": 2.3705,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 73.28983025772678,
+      "learning_rate": 1.4925373134328358e-07,
+      "loss": 2.4624,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 63.49155082883118,
+      "learning_rate": 1.7057569296375268e-07,
+      "loss": 2.6212,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 278.15648985237993,
+      "learning_rate": 1.918976545842218e-07,
+      "loss": 2.6036,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 94.90328718676945,
+      "learning_rate": 2.1321961620469084e-07,
+      "loss": 2.6829,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 44.77966021209984,
+      "learning_rate": 2.3454157782515995e-07,
+      "loss": 2.2883,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 71.00024516870761,
+      "learning_rate": 2.55863539445629e-07,
+      "loss": 2.6157,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 57.4491157029394,
+      "learning_rate": 2.771855010660981e-07,
+      "loss": 2.6069,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 75.86776750197814,
+      "learning_rate": 2.9850746268656716e-07,
+      "loss": 2.5596,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 63.007345223673354,
+      "learning_rate": 3.1982942430703626e-07,
+      "loss": 2.5844,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 49.18458832905806,
+      "learning_rate": 3.4115138592750537e-07,
+      "loss": 2.4342,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 50.992983415134525,
+      "learning_rate": 3.624733475479744e-07,
+      "loss": 2.5765,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 72.6963926783354,
+      "learning_rate": 3.837953091684436e-07,
+      "loss": 2.9592,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 45.74331762737774,
+      "learning_rate": 4.0511727078891263e-07,
+      "loss": 2.5616,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.767026217592509,
+      "learning_rate": 4.264392324093817e-07,
+      "loss": 0.8812,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 62.31038019817941,
+      "learning_rate": 4.4776119402985074e-07,
+      "loss": 2.6034,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 56.103544410278396,
+      "learning_rate": 4.690831556503199e-07,
+      "loss": 2.387,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 62.42502907873537,
+      "learning_rate": 4.904051172707889e-07,
+      "loss": 2.3095,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 64.97639451474666,
+      "learning_rate": 5.11727078891258e-07,
+      "loss": 2.3995,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 62.32331964067838,
+      "learning_rate": 5.33049040511727e-07,
+      "loss": 2.5911,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 64.82206526120157,
+      "learning_rate": 5.543710021321962e-07,
+      "loss": 2.4495,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 63.159337081715364,
+      "learning_rate": 5.756929637526653e-07,
+      "loss": 2.5475,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 51.00077506356932,
+      "learning_rate": 5.970149253731343e-07,
+      "loss": 2.3218,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 58.30578551544221,
+      "learning_rate": 6.183368869936035e-07,
+      "loss": 2.232,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 64.18174184503285,
+      "learning_rate": 6.396588486140725e-07,
+      "loss": 2.2568,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 86.22269658330983,
+      "learning_rate": 6.609808102345417e-07,
+      "loss": 2.0562,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 87.29185295362402,
+      "learning_rate": 6.823027718550107e-07,
+      "loss": 2.1047,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 83.51574750669549,
+      "learning_rate": 7.036247334754798e-07,
+      "loss": 2.1052,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.70066938984924,
+      "learning_rate": 7.249466950959488e-07,
+      "loss": 0.9186,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 68.17579507584504,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 1.9685,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.6249801473523,
+      "learning_rate": 7.675906183368872e-07,
+      "loss": 2.1414,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.833821333276674,
+      "learning_rate": 7.889125799573562e-07,
+      "loss": 0.8636,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 248.11559604792785,
+      "learning_rate": 8.102345415778253e-07,
+      "loss": 2.0657,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.95795199444233,
+      "learning_rate": 8.315565031982943e-07,
+      "loss": 2.0564,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.897357289416306,
+      "learning_rate": 8.528784648187634e-07,
+      "loss": 1.9331,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 38.157679321314504,
+      "learning_rate": 8.742004264392324e-07,
+      "loss": 2.0134,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 66.97239312690319,
+      "learning_rate": 8.955223880597015e-07,
+      "loss": 1.8253,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.53607611848116,
+      "learning_rate": 9.168443496801707e-07,
+      "loss": 0.7007,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.204926714662463,
+      "learning_rate": 9.381663113006398e-07,
+      "loss": 1.6678,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.609866617276943,
+      "learning_rate": 9.594882729211088e-07,
+      "loss": 1.7156,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 55.90917253788717,
+      "learning_rate": 9.808102345415779e-07,
+      "loss": 1.5906,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.19385683644841,
+      "learning_rate": 1.002132196162047e-06,
+      "loss": 0.8057,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.689133367305164,
+      "learning_rate": 1.023454157782516e-06,
+      "loss": 1.7826,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.929640812642898,
+      "learning_rate": 1.044776119402985e-06,
+      "loss": 1.5411,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.776964757557103,
+      "learning_rate": 1.066098081023454e-06,
+      "loss": 1.6069,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.235034066294077,
+      "learning_rate": 1.0874200426439234e-06,
+      "loss": 1.5816,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.59296996138092,
+      "learning_rate": 1.1087420042643924e-06,
+      "loss": 1.6951,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 46.53019391104285,
+      "learning_rate": 1.1300639658848615e-06,
+      "loss": 1.4011,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.417391241399192,
+      "learning_rate": 1.1513859275053305e-06,
+      "loss": 1.4323,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.783386053397983,
+      "learning_rate": 1.1727078891257996e-06,
+      "loss": 1.4347,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.73848464399898,
+      "learning_rate": 1.1940298507462686e-06,
+      "loss": 1.3042,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.07510340352344,
+      "learning_rate": 1.2153518123667379e-06,
+      "loss": 1.3994,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.458102287047573,
+      "learning_rate": 1.236673773987207e-06,
+      "loss": 1.4245,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.557308594424608,
+      "learning_rate": 1.257995735607676e-06,
+      "loss": 1.342,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.2539287984031,
+      "learning_rate": 1.279317697228145e-06,
+      "loss": 1.2302,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.636266160691681,
+      "learning_rate": 1.300639658848614e-06,
+      "loss": 1.4054,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.980733008067061,
+      "learning_rate": 1.3219616204690834e-06,
+      "loss": 1.5028,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.523998695144336,
+      "learning_rate": 1.3432835820895524e-06,
+      "loss": 1.3473,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.455874195903272,
+      "learning_rate": 1.3646055437100215e-06,
+      "loss": 1.4501,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.786474591457521,
+      "learning_rate": 1.3859275053304905e-06,
+      "loss": 1.3345,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.713974116310471,
+      "learning_rate": 1.4072494669509596e-06,
+      "loss": 1.4953,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.771938715488602,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 1.3075,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.37203719623312,
+      "learning_rate": 1.4498933901918977e-06,
+      "loss": 1.3802,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.248665967343827,
+      "learning_rate": 1.4712153518123667e-06,
+      "loss": 1.4261,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.257935584131752,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 1.3276,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.432567404612079,
+      "learning_rate": 1.5138592750533053e-06,
+      "loss": 1.3671,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 37.1653468256582,
+      "learning_rate": 1.5351812366737743e-06,
+      "loss": 1.389,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.917274108513551,
+      "learning_rate": 1.5565031982942434e-06,
+      "loss": 1.368,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.446172420325409,
+      "learning_rate": 1.5778251599147124e-06,
+      "loss": 1.0826,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5597635686863358,
+      "learning_rate": 1.5991471215351815e-06,
+      "loss": 0.6199,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.42489654552413,
+      "learning_rate": 1.6204690831556505e-06,
+      "loss": 1.2677,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.94981921811569,
+      "learning_rate": 1.6417910447761196e-06,
+      "loss": 1.1958,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.47221563900225,
+      "learning_rate": 1.6631130063965886e-06,
+      "loss": 1.376,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.291163111073828,
+      "learning_rate": 1.6844349680170577e-06,
+      "loss": 1.2901,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.547232874613226,
+      "learning_rate": 1.7057569296375267e-06,
+      "loss": 1.366,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1063455988348685,
+      "learning_rate": 1.7270788912579958e-06,
+      "loss": 0.7183,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.18564074704744,
+      "learning_rate": 1.7484008528784648e-06,
+      "loss": 1.2324,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9850047383766287,
+      "learning_rate": 1.7697228144989339e-06,
+      "loss": 0.7561,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.189988218618845,
+      "learning_rate": 1.791044776119403e-06,
+      "loss": 1.2632,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.391867104922794,
+      "learning_rate": 1.812366737739872e-06,
+      "loss": 1.4113,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.95870135816185,
+      "learning_rate": 1.8336886993603415e-06,
+      "loss": 1.2626,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 22.318968733691865,
+      "learning_rate": 1.8550106609808105e-06,
+      "loss": 1.2364,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.17503586710914,
+      "learning_rate": 1.8763326226012796e-06,
+      "loss": 1.185,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.518481091226265,
+      "learning_rate": 1.8976545842217486e-06,
+      "loss": 1.1927,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.296227207927595,
+      "learning_rate": 1.9189765458422177e-06,
+      "loss": 1.2847,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.151930810652454,
+      "learning_rate": 1.9402985074626867e-06,
+      "loss": 1.163,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.624102769700826,
+      "learning_rate": 1.9616204690831558e-06,
+      "loss": 1.3188,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.850168706406778,
+      "learning_rate": 1.982942430703625e-06,
+      "loss": 1.2343,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.879157225857172,
+      "learning_rate": 2.004264392324094e-06,
+      "loss": 1.3046,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.603133041244414,
+      "learning_rate": 2.025586353944563e-06,
+      "loss": 1.2393,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.370889315559152,
+      "learning_rate": 2.046908315565032e-06,
+      "loss": 1.094,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.235153108164525,
+      "learning_rate": 2.068230277185501e-06,
+      "loss": 1.3825,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.444193368552368,
+      "learning_rate": 2.08955223880597e-06,
+      "loss": 1.0877,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.270719877217546,
+      "learning_rate": 2.110874200426439e-06,
+      "loss": 1.0861,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.750356321721096,
+      "learning_rate": 2.132196162046908e-06,
+      "loss": 1.1845,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.53265395978728,
+      "learning_rate": 2.1535181236673773e-06,
+      "loss": 1.267,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4135468140183365,
+      "learning_rate": 2.1748400852878467e-06,
+      "loss": 1.3806,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.922912988320972,
+      "learning_rate": 2.1961620469083158e-06,
+      "loss": 1.2718,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7231614621300566,
+      "learning_rate": 2.217484008528785e-06,
+      "loss": 0.7582,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.825756720936894,
+      "learning_rate": 2.238805970149254e-06,
+      "loss": 1.0997,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.805472915031668,
+      "learning_rate": 2.260127931769723e-06,
+      "loss": 1.1958,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.868993093674682,
+      "learning_rate": 2.281449893390192e-06,
+      "loss": 1.2951,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.01138456608424,
+      "learning_rate": 2.302771855010661e-06,
+      "loss": 1.3222,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.872974463363175,
+      "learning_rate": 2.32409381663113e-06,
+      "loss": 1.0756,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 41.88222065768421,
+      "learning_rate": 2.345415778251599e-06,
+      "loss": 1.1766,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3726442195191721,
+      "learning_rate": 2.366737739872068e-06,
+      "loss": 0.7524,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.751320719796473,
+      "learning_rate": 2.3880597014925373e-06,
+      "loss": 1.182,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.230261996886388,
+      "learning_rate": 2.4093816631130067e-06,
+      "loss": 1.3797,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.509230853935799,
+      "learning_rate": 2.4307036247334758e-06,
+      "loss": 1.2606,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4751134398060541,
+      "learning_rate": 2.452025586353945e-06,
+      "loss": 0.604,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.626786611639764,
+      "learning_rate": 2.473347547974414e-06,
+      "loss": 1.2106,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.057105558601848,
+      "learning_rate": 2.494669509594883e-06,
+      "loss": 1.1911,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.3808010643814,
+      "learning_rate": 2.515991471215352e-06,
+      "loss": 1.2886,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2760786853390904,
+      "learning_rate": 2.537313432835821e-06,
+      "loss": 1.1661,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.92935178165713,
+      "learning_rate": 2.55863539445629e-06,
+      "loss": 1.2769,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.008532039812362,
+      "learning_rate": 2.579957356076759e-06,
+      "loss": 1.2949,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.858603220114233,
+      "learning_rate": 2.601279317697228e-06,
+      "loss": 1.371,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.015782618063819,
+      "learning_rate": 2.6226012793176977e-06,
+      "loss": 1.1403,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.689193275311496,
+      "learning_rate": 2.6439232409381667e-06,
+      "loss": 1.0232,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.711698322820527,
+      "learning_rate": 2.6652452025586358e-06,
+      "loss": 1.165,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6371799383125527,
+      "learning_rate": 2.686567164179105e-06,
+      "loss": 0.6344,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.247122746178185,
+      "learning_rate": 2.707889125799574e-06,
+      "loss": 0.6825,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.239808922103216,
+      "learning_rate": 2.729211087420043e-06,
+      "loss": 1.0637,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.43735564538259,
+      "learning_rate": 2.750533049040512e-06,
+      "loss": 1.1254,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 54.3200390832031,
+      "learning_rate": 2.771855010660981e-06,
+      "loss": 1.1736,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.759141369172903,
+      "learning_rate": 2.79317697228145e-06,
+      "loss": 1.1698,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.514823350919514,
+      "learning_rate": 2.814498933901919e-06,
+      "loss": 1.314,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.894211171539588,
+      "learning_rate": 2.835820895522388e-06,
+      "loss": 1.3229,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.064609599040754,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.1999,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.847685634963284,
+      "learning_rate": 2.8784648187633263e-06,
+      "loss": 1.1949,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.770036878420736,
+      "learning_rate": 2.8997867803837954e-06,
+      "loss": 1.3408,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.6041778353657845,
+      "learning_rate": 2.9211087420042644e-06,
+      "loss": 1.1747,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.195703438035721,
+      "learning_rate": 2.9424307036247335e-06,
+      "loss": 1.2388,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.0378270604497954,
+      "learning_rate": 2.9637526652452025e-06,
+      "loss": 1.0635,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.734883418687147,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 1.1715,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.052090843676889,
+      "learning_rate": 3.006396588486141e-06,
+      "loss": 1.0749,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.128489429016435,
+      "learning_rate": 3.0277185501066105e-06,
+      "loss": 1.0228,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2925711242905227,
+      "learning_rate": 3.0490405117270796e-06,
+      "loss": 0.6488,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.796730837874806,
+      "learning_rate": 3.0703624733475486e-06,
+      "loss": 1.1848,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.850079364384,
+      "learning_rate": 3.0916844349680177e-06,
+      "loss": 1.1195,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0054539642170073,
+      "learning_rate": 3.1130063965884867e-06,
+      "loss": 1.0159,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.43503491783842,
+      "learning_rate": 3.1343283582089558e-06,
+      "loss": 1.2785,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.800232473185519,
+      "learning_rate": 3.155650319829425e-06,
+      "loss": 1.1458,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.530239499786015,
+      "learning_rate": 3.176972281449894e-06,
+      "loss": 1.063,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.104032642758299,
+      "learning_rate": 3.198294243070363e-06,
+      "loss": 1.1003,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4241872055946598,
+      "learning_rate": 3.219616204690832e-06,
+      "loss": 0.7389,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.63778977804872,
+      "learning_rate": 3.240938166311301e-06,
+      "loss": 1.2016,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.3108397974122115,
+      "learning_rate": 3.26226012793177e-06,
+      "loss": 1.1472,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.822690788930453,
+      "learning_rate": 3.283582089552239e-06,
+      "loss": 1.1458,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.736116650545294,
+      "learning_rate": 3.304904051172708e-06,
+      "loss": 1.2672,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.469373210701871,
+      "learning_rate": 3.3262260127931773e-06,
+      "loss": 1.1819,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 36.900386580470844,
+      "learning_rate": 3.3475479744136463e-06,
+      "loss": 1.0204,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5249003568390236,
+      "learning_rate": 3.3688699360341154e-06,
+      "loss": 0.7142,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.93383128259406,
+      "learning_rate": 3.3901918976545844e-06,
+      "loss": 1.1797,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.68436041342194,
+      "learning_rate": 3.4115138592750535e-06,
+      "loss": 1.1035,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 37.475142994491264,
+      "learning_rate": 3.4328358208955225e-06,
+      "loss": 1.3653,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.279920851542017,
+      "learning_rate": 3.4541577825159916e-06,
+      "loss": 1.0657,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.559661314071839,
+      "learning_rate": 3.4754797441364606e-06,
+      "loss": 1.0885,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.871890391323651,
+      "learning_rate": 3.4968017057569297e-06,
+      "loss": 1.2339,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.204555978521244,
+      "learning_rate": 3.5181236673773987e-06,
+      "loss": 1.2936,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 22.136110465288667,
+      "learning_rate": 3.5394456289978678e-06,
+      "loss": 1.0211,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.5164232079556,
+      "learning_rate": 3.560767590618337e-06,
+      "loss": 0.9979,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.225906014185934,
+      "learning_rate": 3.582089552238806e-06,
+      "loss": 1.2196,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.263811809198724,
+      "learning_rate": 3.603411513859275e-06,
+      "loss": 1.1136,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.636911610786065,
+      "learning_rate": 3.624733475479744e-06,
+      "loss": 1.0301,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.177494631448425,
+      "learning_rate": 3.6460554371002135e-06,
+      "loss": 1.219,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.674012570037817,
+      "learning_rate": 3.667377398720683e-06,
+      "loss": 1.129,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 28.081620453804835,
+      "learning_rate": 3.688699360341152e-06,
+      "loss": 1.2278,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.330804179578259,
+      "learning_rate": 3.710021321961621e-06,
+      "loss": 1.0443,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.36490214456325,
+      "learning_rate": 3.73134328358209e-06,
+      "loss": 1.0414,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.02571690419291,
+      "learning_rate": 3.752665245202559e-06,
+      "loss": 1.0479,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4757155894705205,
+      "learning_rate": 3.773987206823028e-06,
+      "loss": 1.159,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 28.222455015202797,
+      "learning_rate": 3.7953091684434973e-06,
+      "loss": 1.0982,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3257871918898254,
+      "learning_rate": 3.816631130063966e-06,
+      "loss": 1.1249,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.901512484009744,
+      "learning_rate": 3.837953091684435e-06,
+      "loss": 1.072,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.924008242092377,
+      "learning_rate": 3.859275053304904e-06,
+      "loss": 1.2503,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.984755577708651,
+      "learning_rate": 3.8805970149253735e-06,
+      "loss": 1.0784,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.36237479311164,
+      "learning_rate": 3.9019189765458425e-06,
+      "loss": 1.1152,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.002535002686066,
+      "learning_rate": 3.9232409381663116e-06,
+      "loss": 1.181,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.85152445051231,
+      "learning_rate": 3.944562899786781e-06,
+      "loss": 1.1179,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.700706072879604,
+      "learning_rate": 3.96588486140725e-06,
+      "loss": 1.2031,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.475278851894241,
+      "learning_rate": 3.987206823027719e-06,
+      "loss": 0.9383,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.312194064806178,
+      "learning_rate": 4.008528784648188e-06,
+      "loss": 1.2565,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.289931840943319,
+      "learning_rate": 4.029850746268657e-06,
+      "loss": 1.1712,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 28.474468666601798,
+      "learning_rate": 4.051172707889126e-06,
+      "loss": 1.0801,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.281322924794292,
+      "learning_rate": 4.072494669509595e-06,
+      "loss": 0.6092,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.255959325222555,
+      "learning_rate": 4.093816631130064e-06,
+      "loss": 1.1308,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1919228490030436,
+      "learning_rate": 4.115138592750533e-06,
+      "loss": 0.5999,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.607448834315255,
+      "learning_rate": 4.136460554371002e-06,
+      "loss": 1.1884,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2518450952960458,
+      "learning_rate": 4.157782515991471e-06,
+      "loss": 0.7134,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.850617023167418,
+      "learning_rate": 4.17910447761194e-06,
+      "loss": 1.2841,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8295741053639114,
+      "learning_rate": 4.200426439232409e-06,
+      "loss": 1.0228,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.734248777763311,
+      "learning_rate": 4.221748400852878e-06,
+      "loss": 1.1588,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.304209973871444,
+      "learning_rate": 4.243070362473347e-06,
+      "loss": 1.2057,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.865091294329579,
+      "learning_rate": 4.264392324093816e-06,
+      "loss": 1.1558,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.433276971288247,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 1.1113,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.169684255253151,
+      "learning_rate": 4.3070362473347545e-06,
+      "loss": 0.977,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.723291584484334,
+      "learning_rate": 4.3283582089552236e-06,
+      "loss": 1.047,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.714847364114116,
+      "learning_rate": 4.3496801705756935e-06,
+      "loss": 1.0008,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.443423089785615,
+      "learning_rate": 4.3710021321961625e-06,
+      "loss": 1.0453,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.239908617209469,
+      "learning_rate": 4.3923240938166316e-06,
+      "loss": 1.1481,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5091296787974704,
+      "learning_rate": 4.413646055437101e-06,
+      "loss": 0.6167,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.080832225437122,
+      "learning_rate": 4.43496801705757e-06,
+      "loss": 1.0921,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.025165769851117,
+      "learning_rate": 4.456289978678039e-06,
+      "loss": 1.0279,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.53307275171082,
+      "learning_rate": 4.477611940298508e-06,
+      "loss": 1.1122,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.9275570561327795,
+      "learning_rate": 4.498933901918977e-06,
+      "loss": 1.277,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 22.128476523474014,
+      "learning_rate": 4.520255863539446e-06,
+      "loss": 1.0808,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.055552621509658,
+      "learning_rate": 4.541577825159915e-06,
+      "loss": 1.0503,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.088079275595002,
+      "learning_rate": 4.562899786780384e-06,
+      "loss": 1.2632,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5158241819824305,
+      "learning_rate": 4.584221748400853e-06,
+      "loss": 0.7528,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.145034049266508,
+      "learning_rate": 4.605543710021322e-06,
+      "loss": 1.1792,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3144830616025778,
+      "learning_rate": 4.626865671641791e-06,
+      "loss": 1.0473,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.486763031490411,
+      "learning_rate": 4.64818763326226e-06,
+      "loss": 1.0338,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1698302668382292,
+      "learning_rate": 4.669509594882729e-06,
+      "loss": 0.606,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.899168650409468,
+      "learning_rate": 4.690831556503198e-06,
+      "loss": 0.9825,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.992495179807848,
+      "learning_rate": 4.712153518123667e-06,
+      "loss": 1.1619,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.095247666419215,
+      "learning_rate": 4.733475479744136e-06,
+      "loss": 1.0974,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.948058492563431,
+      "learning_rate": 4.7547974413646055e-06,
+      "loss": 1.1445,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.1813942631086976,
+      "learning_rate": 4.7761194029850745e-06,
+      "loss": 1.1667,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.06813276254129,
+      "learning_rate": 4.797441364605544e-06,
+      "loss": 1.1481,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.2158163343491015,
+      "learning_rate": 4.8187633262260135e-06,
+      "loss": 1.1759,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.089976805320044,
+      "learning_rate": 4.8400852878464825e-06,
+      "loss": 1.069,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.280591843956564,
+      "learning_rate": 4.8614072494669516e-06,
+      "loss": 1.1297,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7153889714161805,
+      "learning_rate": 4.882729211087421e-06,
+      "loss": 0.7103,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.052725964655383,
+      "learning_rate": 4.90405117270789e-06,
+      "loss": 1.0194,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2462777227727866,
+      "learning_rate": 4.925373134328359e-06,
+      "loss": 1.1659,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.66253593039352,
+      "learning_rate": 4.946695095948828e-06,
+      "loss": 1.1306,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.194934581011807,
+      "learning_rate": 4.968017057569297e-06,
+      "loss": 0.7321,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.787065708937597,
+      "learning_rate": 4.989339019189766e-06,
+      "loss": 1.108,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.694790075756236,
+      "learning_rate": 5.010660980810235e-06,
+      "loss": 1.2238,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.457198874321435,
+      "learning_rate": 5.031982942430704e-06,
+      "loss": 1.2012,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.4816791943594305,
+      "learning_rate": 5.053304904051173e-06,
+      "loss": 1.0705,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.460790580798683,
+      "learning_rate": 5.074626865671642e-06,
+      "loss": 1.0557,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 26.747035211945278,
+      "learning_rate": 5.095948827292111e-06,
+      "loss": 1.2116,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.3697849026485125,
+      "learning_rate": 5.11727078891258e-06,
+      "loss": 1.0281,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.434673886530643,
+      "learning_rate": 5.138592750533049e-06,
+      "loss": 1.1542,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.14874427385892,
+      "learning_rate": 5.159914712153518e-06,
+      "loss": 1.1371,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.008381558748736,
+      "learning_rate": 5.181236673773987e-06,
+      "loss": 0.8841,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.58103950144051,
+      "learning_rate": 5.202558635394456e-06,
+      "loss": 1.1705,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.84052415834236,
+      "learning_rate": 5.2238805970149255e-06,
+      "loss": 1.16,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8137370350716617,
+      "learning_rate": 5.245202558635395e-06,
+      "loss": 1.2385,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.708214033400164,
+      "learning_rate": 5.2665245202558636e-06,
+      "loss": 1.0766,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.848810326558273,
+      "learning_rate": 5.2878464818763335e-06,
+      "loss": 0.9188,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8009278040031096,
+      "learning_rate": 5.309168443496802e-06,
+      "loss": 1.172,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8098987849675265,
+      "learning_rate": 5.3304904051172716e-06,
+      "loss": 1.033,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.769606563614491,
+      "learning_rate": 5.351812366737741e-06,
+      "loss": 1.048,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.5182201372359465,
+      "learning_rate": 5.37313432835821e-06,
+      "loss": 1.0992,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.05827259037666,
+      "learning_rate": 5.394456289978679e-06,
+      "loss": 1.1577,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6318738594225766,
+      "learning_rate": 5.415778251599148e-06,
+      "loss": 0.7405,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.338986609535587,
+      "learning_rate": 5.437100213219617e-06,
+      "loss": 1.1058,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.626234166627977,
+      "learning_rate": 5.458422174840086e-06,
+      "loss": 1.1775,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.329249332493462,
+      "learning_rate": 5.479744136460555e-06,
+      "loss": 1.073,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.719079613009029,
+      "learning_rate": 5.501066098081024e-06,
+      "loss": 0.972,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.063227947439973,
+      "learning_rate": 5.522388059701493e-06,
+      "loss": 0.8083,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.373270365463836,
+      "learning_rate": 5.543710021321962e-06,
+      "loss": 1.1051,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.828167862152183,
+      "learning_rate": 5.565031982942431e-06,
+      "loss": 1.1969,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.223610368487151,
+      "learning_rate": 5.5863539445629e-06,
+      "loss": 1.0752,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9448140430784373,
+      "learning_rate": 5.607675906183369e-06,
+      "loss": 0.9791,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.242180101444069,
+      "learning_rate": 5.628997867803838e-06,
+      "loss": 1.0321,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.888003427988579,
+      "learning_rate": 5.650319829424308e-06,
+      "loss": 1.1441,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.825956039372433,
+      "learning_rate": 5.671641791044776e-06,
+      "loss": 1.0906,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.60547281289873,
+      "learning_rate": 5.692963752665246e-06,
+      "loss": 1.0267,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.520711569146826,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.1683,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.067485454113877,
+      "learning_rate": 5.735607675906184e-06,
+      "loss": 1.2489,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.25009587085642,
+      "learning_rate": 5.756929637526653e-06,
+      "loss": 1.2476,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.586401497333892,
+      "learning_rate": 5.7782515991471225e-06,
+      "loss": 1.1528,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.304267460615097,
+      "learning_rate": 5.799573560767591e-06,
+      "loss": 1.0464,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.42181704700264,
+      "learning_rate": 5.820895522388061e-06,
+      "loss": 1.2026,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 31.407472908174952,
+      "learning_rate": 5.842217484008529e-06,
+      "loss": 1.1586,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.22090306345048,
+      "learning_rate": 5.863539445628999e-06,
+      "loss": 1.0177,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.247749408920635,
+      "learning_rate": 5.884861407249467e-06,
+      "loss": 0.9803,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.9679186244614275,
+      "learning_rate": 5.906183368869937e-06,
+      "loss": 1.0894,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.144595495354819,
+      "learning_rate": 5.927505330490405e-06,
+      "loss": 0.5853,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.373288766044588,
+      "learning_rate": 5.948827292110875e-06,
+      "loss": 1.0847,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.751822460861834,
+      "learning_rate": 5.970149253731343e-06,
+      "loss": 1.1593,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.544318069671876,
+      "learning_rate": 5.991471215351813e-06,
+      "loss": 0.6857,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.444543246301944,
+      "learning_rate": 6.012793176972282e-06,
+      "loss": 1.0439,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.468212098184548,
+      "learning_rate": 6.034115138592751e-06,
+      "loss": 1.1183,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.798721975140008,
+      "learning_rate": 6.055437100213221e-06,
+      "loss": 1.0742,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.060457595855817,
+      "learning_rate": 6.076759061833689e-06,
+      "loss": 0.9128,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.315232911637052,
+      "learning_rate": 6.098081023454159e-06,
+      "loss": 1.1946,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.043796648046574,
+      "learning_rate": 6.119402985074627e-06,
+      "loss": 1.0797,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7110989185427687,
+      "learning_rate": 6.140724946695097e-06,
+      "loss": 1.0732,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.981635080698233,
+      "learning_rate": 6.1620469083155655e-06,
+      "loss": 1.0604,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.103230066323087,
+      "learning_rate": 6.183368869936035e-06,
+      "loss": 0.9414,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.377292268266425,
+      "learning_rate": 6.2046908315565036e-06,
+      "loss": 1.1031,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.381386518548588,
+      "learning_rate": 6.2260127931769735e-06,
+      "loss": 1.0711,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.787192288561611,
+      "learning_rate": 6.247334754797442e-06,
+      "loss": 0.9692,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.649779060480321,
+      "learning_rate": 6.2686567164179116e-06,
+      "loss": 1.1104,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.559771755411592,
+      "learning_rate": 6.28997867803838e-06,
+      "loss": 1.1763,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.696941759927689,
+      "learning_rate": 6.31130063965885e-06,
+      "loss": 1.0897,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.37040337857279,
+      "learning_rate": 6.332622601279318e-06,
+      "loss": 1.2192,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.708983982631115,
+      "learning_rate": 6.353944562899788e-06,
+      "loss": 1.1021,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.272437486518209,
+      "learning_rate": 6.375266524520256e-06,
+      "loss": 0.7377,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.071636809623884,
+      "learning_rate": 6.396588486140726e-06,
+      "loss": 1.0478,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.01014320728466,
+      "learning_rate": 6.417910447761194e-06,
+      "loss": 0.9587,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2004172137024813,
+      "learning_rate": 6.439232409381664e-06,
+      "loss": 1.1161,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.605870264570785,
+      "learning_rate": 6.460554371002132e-06,
+      "loss": 1.0749,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.467390593998029,
+      "learning_rate": 6.481876332622602e-06,
+      "loss": 1.0749,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.3133274944724445,
+      "learning_rate": 6.50319829424307e-06,
+      "loss": 1.1346,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.526931259905275,
+      "learning_rate": 6.52452025586354e-06,
+      "loss": 1.062,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.404563617668419,
+      "learning_rate": 6.545842217484008e-06,
+      "loss": 1.032,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.16395154742014,
+      "learning_rate": 6.567164179104478e-06,
+      "loss": 1.0502,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.835433494810792,
+      "learning_rate": 6.5884861407249465e-06,
+      "loss": 1.1819,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.393828813423281,
+      "learning_rate": 6.609808102345416e-06,
+      "loss": 0.9649,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.922272217592682,
+      "learning_rate": 6.631130063965885e-06,
+      "loss": 0.8667,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.379401500724926,
+      "learning_rate": 6.6524520255863545e-06,
+      "loss": 0.8196,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.735853861639495,
+      "learning_rate": 6.673773987206824e-06,
+      "loss": 1.037,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.752132388007364,
+      "learning_rate": 6.695095948827293e-06,
+      "loss": 0.9432,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.7645801346063505,
+      "learning_rate": 6.7164179104477625e-06,
+      "loss": 0.9685,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.987866134205658,
+      "learning_rate": 6.737739872068231e-06,
+      "loss": 0.9619,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.90387876874462,
+      "learning_rate": 6.759061833688701e-06,
+      "loss": 1.1932,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.381239327342861,
+      "learning_rate": 6.780383795309169e-06,
+      "loss": 1.1083,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.8744379605832915,
+      "learning_rate": 6.801705756929639e-06,
+      "loss": 1.1901,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5894530305186176,
+      "learning_rate": 6.823027718550107e-06,
+      "loss": 1.0368,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5320859379928233,
+      "learning_rate": 6.844349680170577e-06,
+      "loss": 0.6578,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.932950152896916,
+      "learning_rate": 6.865671641791045e-06,
+      "loss": 0.6229,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.277530214099487,
+      "learning_rate": 6.886993603411515e-06,
+      "loss": 0.6263,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.436889765876073,
+      "learning_rate": 6.908315565031983e-06,
+      "loss": 0.8593,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.12723914896694,
+      "learning_rate": 6.929637526652453e-06,
+      "loss": 0.9496,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.785004831045194,
+      "learning_rate": 6.950959488272921e-06,
+      "loss": 1.2569,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.441705940533284,
+      "learning_rate": 6.972281449893391e-06,
+      "loss": 0.6841,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.805215344426466,
+      "learning_rate": 6.993603411513859e-06,
+      "loss": 1.1206,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.321788246897633,
+      "learning_rate": 7.014925373134329e-06,
+      "loss": 1.3162,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.356302488357698,
+      "learning_rate": 7.0362473347547975e-06,
+      "loss": 1.1112,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.412717301032853,
+      "learning_rate": 7.057569296375267e-06,
+      "loss": 0.7374,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.947043500615336,
+      "learning_rate": 7.0788912579957356e-06,
+      "loss": 0.9481,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.12389622150325,
+      "learning_rate": 7.1002132196162055e-06,
+      "loss": 1.0651,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.474157741678489,
+      "learning_rate": 7.121535181236674e-06,
+      "loss": 1.1137,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2218410550929883,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.6447,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.545996887004964,
+      "learning_rate": 7.164179104477612e-06,
+      "loss": 1.065,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.045413185009348,
+      "learning_rate": 7.185501066098082e-06,
+      "loss": 0.9655,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2771223866285966,
+      "learning_rate": 7.20682302771855e-06,
+      "loss": 1.1061,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.978961482172894,
+      "learning_rate": 7.22814498933902e-06,
+      "loss": 1.1242,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.8091894573649085,
+      "learning_rate": 7.249466950959488e-06,
+      "loss": 1.0437,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.421327561861191,
+      "learning_rate": 7.270788912579958e-06,
+      "loss": 1.2107,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8537498242841433,
+      "learning_rate": 7.292110874200427e-06,
+      "loss": 0.9111,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.859924158143228,
+      "learning_rate": 7.313432835820896e-06,
+      "loss": 1.0182,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5432363173171615,
+      "learning_rate": 7.334754797441366e-06,
+      "loss": 0.6373,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4461447373466014,
+      "learning_rate": 7.356076759061834e-06,
+      "loss": 1.1192,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2632179247111748,
+      "learning_rate": 7.377398720682304e-06,
+      "loss": 0.9426,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.831412477225999,
+      "learning_rate": 7.398720682302772e-06,
+      "loss": 1.1482,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.753536661375966,
+      "learning_rate": 7.420042643923242e-06,
+      "loss": 1.1361,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.500701980472856,
+      "learning_rate": 7.44136460554371e-06,
+      "loss": 0.9328,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3012951865483307,
+      "learning_rate": 7.46268656716418e-06,
+      "loss": 0.7334,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.580476644912745,
+      "learning_rate": 7.484008528784648e-06,
+      "loss": 0.9768,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4408114345597496,
+      "learning_rate": 7.505330490405118e-06,
+      "loss": 1.095,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1559838777077256,
+      "learning_rate": 7.5266524520255865e-06,
+      "loss": 1.0506,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2354688283080026,
+      "learning_rate": 7.547974413646056e-06,
+      "loss": 0.942,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9635970812183423,
+      "learning_rate": 7.569296375266525e-06,
+      "loss": 1.0118,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.39347114182356,
+      "learning_rate": 7.5906183368869945e-06,
+      "loss": 1.0522,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.234515478459951,
+      "learning_rate": 7.611940298507463e-06,
+      "loss": 0.9658,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4846694761067833,
+      "learning_rate": 7.633262260127933e-06,
+      "loss": 0.7234,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.500680173189644,
+      "learning_rate": 7.654584221748402e-06,
+      "loss": 0.9882,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.754825351626341,
+      "learning_rate": 7.67590618336887e-06,
+      "loss": 1.0069,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0505874338738908,
+      "learning_rate": 7.69722814498934e-06,
+      "loss": 1.1631,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1681811576362664,
+      "learning_rate": 7.718550106609809e-06,
+      "loss": 0.6286,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.725319997018703,
+      "learning_rate": 7.739872068230278e-06,
+      "loss": 1.0289,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.823517337422678,
+      "learning_rate": 7.761194029850747e-06,
+      "loss": 1.1641,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.233526493661488,
+      "learning_rate": 7.782515991471216e-06,
+      "loss": 1.03,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.128036473905242,
+      "learning_rate": 7.803837953091685e-06,
+      "loss": 0.6658,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.869535149661054,
+      "learning_rate": 7.825159914712154e-06,
+      "loss": 1.0553,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.140433477010861,
+      "learning_rate": 7.846481876332623e-06,
+      "loss": 1.1612,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.7058010095707195,
+      "learning_rate": 7.867803837953092e-06,
+      "loss": 1.0049,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.094842939403695,
+      "learning_rate": 7.889125799573561e-06,
+      "loss": 1.0547,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.662709598749715,
+      "learning_rate": 7.91044776119403e-06,
+      "loss": 1.0764,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.642856927267935,
+      "learning_rate": 7.9317697228145e-06,
+      "loss": 0.999,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.42398085604514,
+      "learning_rate": 7.953091684434968e-06,
+      "loss": 1.1626,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.747177618128887,
+      "learning_rate": 7.974413646055437e-06,
+      "loss": 1.0283,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.651523788897506,
+      "learning_rate": 7.995735607675907e-06,
+      "loss": 1.1248,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6130478535074944,
+      "learning_rate": 8.017057569296376e-06,
+      "loss": 1.0028,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.897577889353655,
+      "learning_rate": 8.038379530916846e-06,
+      "loss": 1.0599,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.174908235791682,
+      "learning_rate": 8.059701492537314e-06,
+      "loss": 0.9255,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.197833802902745,
+      "learning_rate": 8.081023454157784e-06,
+      "loss": 0.9124,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.024809468360133,
+      "learning_rate": 8.102345415778252e-06,
+      "loss": 1.0361,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6147980601628635,
+      "learning_rate": 8.123667377398723e-06,
+      "loss": 1.0816,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6292779884622717,
+      "learning_rate": 8.14498933901919e-06,
+      "loss": 1.0644,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4778077077747422,
+      "learning_rate": 8.16631130063966e-06,
+      "loss": 1.1553,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8264241239921817,
+      "learning_rate": 8.187633262260128e-06,
+      "loss": 1.0301,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1417264324121046,
+      "learning_rate": 8.208955223880599e-06,
+      "loss": 0.9595,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.459005724274181,
+      "learning_rate": 8.230277185501066e-06,
+      "loss": 0.7392,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3535443655623147,
+      "learning_rate": 8.251599147121537e-06,
+      "loss": 0.65,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7298521795323534,
+      "learning_rate": 8.272921108742004e-06,
+      "loss": 0.668,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9501390592153007,
+      "learning_rate": 8.294243070362475e-06,
+      "loss": 1.2048,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4061472579178593,
+      "learning_rate": 8.315565031982942e-06,
+      "loss": 1.0942,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 30.865589776978588,
+      "learning_rate": 8.336886993603413e-06,
+      "loss": 1.1039,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4375995138392912,
+      "learning_rate": 8.35820895522388e-06,
+      "loss": 0.6785,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2524618029200254,
+      "learning_rate": 8.379530916844351e-06,
+      "loss": 1.003,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8234404187723205,
+      "learning_rate": 8.400852878464819e-06,
+      "loss": 1.0028,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1493208669091386,
+      "learning_rate": 8.42217484008529e-06,
+      "loss": 0.8861,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.275761929529847,
+      "learning_rate": 8.443496801705757e-06,
+      "loss": 0.9674,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.019838720542055,
+      "learning_rate": 8.464818763326227e-06,
+      "loss": 1.189,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.804202449551473,
+      "learning_rate": 8.486140724946695e-06,
+      "loss": 0.9034,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7584214401409786,
+      "learning_rate": 8.507462686567165e-06,
+      "loss": 1.0579,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2345187171976275,
+      "learning_rate": 8.528784648187633e-06,
+      "loss": 0.9837,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4492162752120725,
+      "learning_rate": 8.550106609808104e-06,
+      "loss": 0.6852,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.202051485481815,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.859,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.4209881652462215,
+      "learning_rate": 8.592750533049042e-06,
+      "loss": 1.0909,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.774042141230826,
+      "learning_rate": 8.614072494669509e-06,
+      "loss": 1.0886,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.828204154007439,
+      "learning_rate": 8.63539445628998e-06,
+      "loss": 1.2551,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3069014858956955,
+      "learning_rate": 8.656716417910447e-06,
+      "loss": 1.1458,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.604708455191188,
+      "learning_rate": 8.678038379530918e-06,
+      "loss": 1.0138,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.806148398249084,
+      "learning_rate": 8.699360341151387e-06,
+      "loss": 1.0171,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.378406614750712,
+      "learning_rate": 8.720682302771856e-06,
+      "loss": 0.6438,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.1511785549291735,
+      "learning_rate": 8.742004264392325e-06,
+      "loss": 0.9514,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5593066297827862,
+      "learning_rate": 8.763326226012794e-06,
+      "loss": 1.1645,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.74962983192744,
+      "learning_rate": 8.784648187633263e-06,
+      "loss": 1.1947,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.423600943163277,
+      "learning_rate": 8.805970149253732e-06,
+      "loss": 0.6643,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6104884157732933,
+      "learning_rate": 8.827292110874201e-06,
+      "loss": 1.0806,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.723684778745114,
+      "learning_rate": 8.84861407249467e-06,
+      "loss": 1.0539,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.632885436408049,
+      "learning_rate": 8.86993603411514e-06,
+      "loss": 0.6531,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.828695550664385,
+      "learning_rate": 8.891257995735608e-06,
+      "loss": 0.8432,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.911940948259311,
+      "learning_rate": 8.912579957356077e-06,
+      "loss": 0.8894,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.096941304921919,
+      "learning_rate": 8.933901918976547e-06,
+      "loss": 1.0429,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0496135461818765,
+      "learning_rate": 8.955223880597016e-06,
+      "loss": 1.0259,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7089254320894858,
+      "learning_rate": 8.976545842217485e-06,
+      "loss": 1.0865,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.045340911854935,
+      "learning_rate": 8.997867803837954e-06,
+      "loss": 0.9814,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4819642963104283,
+      "learning_rate": 9.019189765458423e-06,
+      "loss": 1.0374,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.295960601699573,
+      "learning_rate": 9.040511727078892e-06,
+      "loss": 1.0136,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.816118450866647,
+      "learning_rate": 9.06183368869936e-06,
+      "loss": 0.9766,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2517700286154367,
+      "learning_rate": 9.08315565031983e-06,
+      "loss": 1.0846,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.358859305757556,
+      "learning_rate": 9.104477611940299e-06,
+      "loss": 0.9316,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4094437539875666,
+      "learning_rate": 9.125799573560768e-06,
+      "loss": 1.0162,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5970200192585886,
+      "learning_rate": 9.147121535181237e-06,
+      "loss": 0.8538,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.156714916100051,
+      "learning_rate": 9.168443496801706e-06,
+      "loss": 1.1356,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6271508046353715,
+      "learning_rate": 9.189765458422175e-06,
+      "loss": 1.0685,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0215049318155374,
+      "learning_rate": 9.211087420042644e-06,
+      "loss": 1.0324,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6372401246654595,
+      "learning_rate": 9.232409381663113e-06,
+      "loss": 1.0958,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6719912929638987,
+      "learning_rate": 9.253731343283582e-06,
+      "loss": 0.9759,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5472610759416745,
+      "learning_rate": 9.275053304904051e-06,
+      "loss": 0.9758,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.555195296791212,
+      "learning_rate": 9.29637526652452e-06,
+      "loss": 1.1034,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.46567462684545,
+      "learning_rate": 9.31769722814499e-06,
+      "loss": 0.9148,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7048108188921307,
+      "learning_rate": 9.339019189765458e-06,
+      "loss": 0.9075,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2181534448782885,
+      "learning_rate": 9.36034115138593e-06,
+      "loss": 0.9499,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2034262645214882,
+      "learning_rate": 9.381663113006397e-06,
+      "loss": 0.7096,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.848540237697331,
+      "learning_rate": 9.402985074626867e-06,
+      "loss": 1.0423,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.818841510032756,
+      "learning_rate": 9.424307036247335e-06,
+      "loss": 0.8889,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6956604988170776,
+      "learning_rate": 9.445628997867805e-06,
+      "loss": 0.8804,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.05076649788095,
+      "learning_rate": 9.466950959488273e-06,
+      "loss": 0.9032,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.163674761790701,
+      "learning_rate": 9.488272921108744e-06,
+      "loss": 1.1382,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5874106997275954,
+      "learning_rate": 9.509594882729211e-06,
+      "loss": 0.6139,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.222714660246352,
+      "learning_rate": 9.530916844349682e-06,
+      "loss": 1.0726,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4508779870381017,
+      "learning_rate": 9.552238805970149e-06,
+      "loss": 1.0143,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.23943272004539,
+      "learning_rate": 9.57356076759062e-06,
+      "loss": 1.142,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.330717792565949,
+      "learning_rate": 9.594882729211089e-06,
+      "loss": 1.1577,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5978352165595098,
+      "learning_rate": 9.616204690831558e-06,
+      "loss": 0.9636,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9644932475976433,
+      "learning_rate": 9.637526652452027e-06,
+      "loss": 1.0015,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4056069007741328,
+      "learning_rate": 9.658848614072496e-06,
+      "loss": 0.7115,
+      "step": 453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.563315402817599,
+      "learning_rate": 9.680170575692965e-06,
+      "loss": 1.1453,
+      "step": 454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.36367689866523,
+      "learning_rate": 9.701492537313434e-06,
+      "loss": 1.0886,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1411591329047508,
+      "learning_rate": 9.722814498933903e-06,
+      "loss": 0.7474,
+      "step": 456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.842171668095765,
+      "learning_rate": 9.744136460554372e-06,
+      "loss": 1.092,
+      "step": 457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.558559005679244,
+      "learning_rate": 9.765458422174841e-06,
+      "loss": 1.1533,
+      "step": 458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.704924826709696,
+      "learning_rate": 9.78678038379531e-06,
+      "loss": 1.008,
+      "step": 459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5193207772647526,
+      "learning_rate": 9.80810234541578e-06,
+      "loss": 1.1483,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2947278995343106,
+      "learning_rate": 9.829424307036248e-06,
+      "loss": 0.6101,
+      "step": 461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4678284062170226,
+      "learning_rate": 9.850746268656717e-06,
+      "loss": 1.0147,
+      "step": 462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.904862725265896,
+      "learning_rate": 9.872068230277187e-06,
+      "loss": 1.0721,
+      "step": 463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8657740530485816,
+      "learning_rate": 9.893390191897656e-06,
+      "loss": 1.1378,
+      "step": 464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8934815027115177,
+      "learning_rate": 9.914712153518125e-06,
+      "loss": 1.0592,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1358984760652255,
+      "learning_rate": 9.936034115138594e-06,
+      "loss": 0.8871,
+      "step": 466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.816250883581334,
+      "learning_rate": 9.957356076759063e-06,
+      "loss": 0.9433,
+      "step": 467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.307503345847576,
+      "learning_rate": 9.978678038379532e-06,
+      "loss": 0.9778,
+      "step": 468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5868691854164083,
+      "learning_rate": 1e-05,
+      "loss": 0.707,
+      "step": 469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6121743512560895,
+      "learning_rate": 9.999999892569433e-06,
+      "loss": 1.0395,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5992302318002345,
+      "learning_rate": 9.999999570277734e-06,
+      "loss": 0.9524,
+      "step": 471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.242410976825683,
+      "learning_rate": 9.999999033124919e-06,
+      "loss": 0.7122,
+      "step": 472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1937742705082084,
+      "learning_rate": 9.999998281111008e-06,
+      "loss": 0.6335,
+      "step": 473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.696462251317117,
+      "learning_rate": 9.999997314236036e-06,
+      "loss": 1.1489,
+      "step": 474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.20359019984893,
+      "learning_rate": 9.999996132500046e-06,
+      "loss": 0.9814,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.058867864213973,
+      "learning_rate": 9.999994735903083e-06,
+      "loss": 0.9027,
+      "step": 476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.375388158875278,
+      "learning_rate": 9.999993124445213e-06,
+      "loss": 1.0867,
+      "step": 477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.887655601963684,
+      "learning_rate": 9.999991298126501e-06,
+      "loss": 1.0489,
+      "step": 478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.183178204416538,
+      "learning_rate": 9.999989256947029e-06,
+      "loss": 1.0941,
+      "step": 479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.247548674760072,
+      "learning_rate": 9.999987000906883e-06,
+      "loss": 1.0149,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.825204846071956,
+      "learning_rate": 9.99998453000616e-06,
+      "loss": 1.0284,
+      "step": 481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.592614125264408,
+      "learning_rate": 9.999981844244966e-06,
+      "loss": 1.1095,
+      "step": 482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7916839137955574,
+      "learning_rate": 9.999978943623417e-06,
+      "loss": 1.1428,
+      "step": 483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.926064720621627,
+      "learning_rate": 9.999975828141635e-06,
+      "loss": 0.9725,
+      "step": 484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.246039237859176,
+      "learning_rate": 9.999972497799758e-06,
+      "loss": 0.9594,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4225232535488623,
+      "learning_rate": 9.999968952597926e-06,
+      "loss": 0.6892,
+      "step": 486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.343840873464299,
+      "learning_rate": 9.999965192536295e-06,
+      "loss": 0.8876,
+      "step": 487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.527602197593656,
+      "learning_rate": 9.999961217615022e-06,
+      "loss": 0.9427,
+      "step": 488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4464884835642937,
+      "learning_rate": 9.999957027834282e-06,
+      "loss": 0.7218,
+      "step": 489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.212735340088184,
+      "learning_rate": 9.999952623194252e-06,
+      "loss": 1.1474,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8176082996329503,
+      "learning_rate": 9.99994800369512e-06,
+      "loss": 0.9329,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.193113012812268,
+      "learning_rate": 9.99994316933709e-06,
+      "loss": 0.9754,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.445293623938351,
+      "learning_rate": 9.999938120120364e-06,
+      "loss": 0.9554,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4682560742538464,
+      "learning_rate": 9.999932856045164e-06,
+      "loss": 0.8764,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0151696865594797,
+      "learning_rate": 9.999927377111712e-06,
+      "loss": 0.8996,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7209634813272325,
+      "learning_rate": 9.999921683320247e-06,
+      "loss": 0.9678,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6596263906392967,
+      "learning_rate": 9.999915774671009e-06,
+      "loss": 1.0898,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6465533259768135,
+      "learning_rate": 9.999909651164258e-06,
+      "loss": 1.1493,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0028570486509834,
+      "learning_rate": 9.99990331280025e-06,
+      "loss": 0.8139,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4401759535531875,
+      "learning_rate": 9.999896759579263e-06,
+      "loss": 0.9531,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.958936479014827,
+      "learning_rate": 9.99988999150158e-06,
+      "loss": 0.9314,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8565071453960007,
+      "learning_rate": 9.999883008567487e-06,
+      "loss": 0.9976,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.394112308076246,
+      "learning_rate": 9.999875810777283e-06,
+      "loss": 1.1682,
+      "step": 503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.240931190793152,
+      "learning_rate": 9.999868398131282e-06,
+      "loss": 1.03,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.239054124693827,
+      "learning_rate": 9.9998607706298e-06,
+      "loss": 1.0865,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2419351487519394,
+      "learning_rate": 9.999852928273166e-06,
+      "loss": 1.0832,
+      "step": 506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.426877397329804,
+      "learning_rate": 9.999844871061718e-06,
+      "loss": 1.0078,
+      "step": 507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.0331054961980595,
+      "learning_rate": 9.999836598995799e-06,
+      "loss": 1.0105,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.520240784592903,
+      "learning_rate": 9.999828112075764e-06,
+      "loss": 1.0251,
+      "step": 509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.43058161647151,
+      "learning_rate": 9.999819410301983e-06,
+      "loss": 0.9522,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2212333023639244,
+      "learning_rate": 9.999810493674826e-06,
+      "loss": 0.8441,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.09116504138898,
+      "learning_rate": 9.999801362194676e-06,
+      "loss": 1.0648,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.385323505466683,
+      "learning_rate": 9.999792015861927e-06,
+      "loss": 0.9201,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4912240032308295,
+      "learning_rate": 9.99978245467698e-06,
+      "loss": 0.7035,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.505249107655878,
+      "learning_rate": 9.999772678640245e-06,
+      "loss": 0.9831,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5292056926777464,
+      "learning_rate": 9.999762687752143e-06,
+      "loss": 0.986,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.176492018488129,
+      "learning_rate": 9.999752482013105e-06,
+      "loss": 0.967,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.63725848150001,
+      "learning_rate": 9.999742061423567e-06,
+      "loss": 0.859,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6201319646498944,
+      "learning_rate": 9.999731425983975e-06,
+      "loss": 0.764,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3909378663147987,
+      "learning_rate": 9.999720575694792e-06,
+      "loss": 0.9777,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.791222774767542,
+      "learning_rate": 9.99970951055648e-06,
+      "loss": 1.1529,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5196061454179977,
+      "learning_rate": 9.999698230569515e-06,
+      "loss": 1.0838,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.136891049417779,
+      "learning_rate": 9.999686735734382e-06,
+      "loss": 0.9496,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9445827517583285,
+      "learning_rate": 9.999675026051576e-06,
+      "loss": 1.0998,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6931297564462304,
+      "learning_rate": 9.999663101521599e-06,
+      "loss": 0.9686,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0953920668043384,
+      "learning_rate": 9.999650962144962e-06,
+      "loss": 0.9919,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5072036146184273,
+      "learning_rate": 9.999638607922191e-06,
+      "loss": 0.9894,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0637924379654473,
+      "learning_rate": 9.99962603885381e-06,
+      "loss": 1.0247,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3140620075275655,
+      "learning_rate": 9.999613254940368e-06,
+      "loss": 0.9292,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3156471699998018,
+      "learning_rate": 9.999600256182408e-06,
+      "loss": 0.6166,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.981387520891124,
+      "learning_rate": 9.99958704258049e-06,
+      "loss": 0.8733,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.087578852721646,
+      "learning_rate": 9.999573614135183e-06,
+      "loss": 0.7087,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.065070428688435,
+      "learning_rate": 9.999559970847061e-06,
+      "loss": 0.9244,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.59518165732204,
+      "learning_rate": 9.999546112716715e-06,
+      "loss": 0.8757,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8267289151770068,
+      "learning_rate": 9.999532039744736e-06,
+      "loss": 0.9949,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0808380968121107,
+      "learning_rate": 9.999517751931731e-06,
+      "loss": 0.9975,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5130710756242105,
+      "learning_rate": 9.999503249278315e-06,
+      "loss": 0.9768,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.399411718876729,
+      "learning_rate": 9.99948853178511e-06,
+      "loss": 0.9725,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.539267189315805,
+      "learning_rate": 9.999473599452746e-06,
+      "loss": 0.8859,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.01369201987145,
+      "learning_rate": 9.999458452281868e-06,
+      "loss": 0.9487,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6211524582848367,
+      "learning_rate": 9.999443090273126e-06,
+      "loss": 1.0357,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.102520977196429,
+      "learning_rate": 9.99942751342718e-06,
+      "loss": 0.9401,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.135411929317992,
+      "learning_rate": 9.9994117217447e-06,
+      "loss": 1.1248,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4395705283059907,
+      "learning_rate": 9.999395715226365e-06,
+      "loss": 1.0139,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.170926933661236,
+      "learning_rate": 9.99937949387286e-06,
+      "loss": 0.8634,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2674478830329325,
+      "learning_rate": 9.999363057684885e-06,
+      "loss": 0.7503,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.72027556872647,
+      "learning_rate": 9.999346406663144e-06,
+      "loss": 0.6788,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.342855325858615,
+      "learning_rate": 9.999329540808355e-06,
+      "loss": 0.994,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.732551214065963,
+      "learning_rate": 9.999312460121242e-06,
+      "loss": 0.9331,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5365202572045766,
+      "learning_rate": 9.999295164602536e-06,
+      "loss": 1.2067,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9717038634335493,
+      "learning_rate": 9.999277654252985e-06,
+      "loss": 1.0464,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.270583588229037,
+      "learning_rate": 9.999259929073339e-06,
+      "loss": 0.9168,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1779463748812624,
+      "learning_rate": 9.999241989064358e-06,
+      "loss": 1.1564,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5483277286069885,
+      "learning_rate": 9.999223834226817e-06,
+      "loss": 1.0693,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.128890758388742,
+      "learning_rate": 9.999205464561494e-06,
+      "loss": 0.7747,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.606442286253325,
+      "learning_rate": 9.999186880069178e-06,
+      "loss": 1.0907,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.951750281370718,
+      "learning_rate": 9.999168080750666e-06,
+      "loss": 1.0906,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1876695582742722,
+      "learning_rate": 9.99914906660677e-06,
+      "loss": 1.0572,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.462710254568241,
+      "learning_rate": 9.999129837638303e-06,
+      "loss": 1.0202,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1149918074407608,
+      "learning_rate": 9.999110393846097e-06,
+      "loss": 1.0116,
+      "step": 560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.057017979395074,
+      "learning_rate": 9.999090735230981e-06,
+      "loss": 1.1067,
+      "step": 561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.055455943747748,
+      "learning_rate": 9.999070861793803e-06,
+      "loss": 0.8721,
+      "step": 562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3853004769546988,
+      "learning_rate": 9.999050773535414e-06,
+      "loss": 0.6168,
+      "step": 563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2682472640391071,
+      "learning_rate": 9.999030470456684e-06,
+      "loss": 0.671,
+      "step": 564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.45757795131312,
+      "learning_rate": 9.999009952558478e-06,
+      "loss": 0.8546,
+      "step": 565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7322755445557534,
+      "learning_rate": 9.998989219841684e-06,
+      "loss": 1.0533,
+      "step": 566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7725378237262674,
+      "learning_rate": 9.998968272307187e-06,
+      "loss": 1.0723,
+      "step": 567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.301793147569162,
+      "learning_rate": 9.998947109955892e-06,
+      "loss": 0.6807,
+      "step": 568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4535104844759412,
+      "learning_rate": 9.998925732788706e-06,
+      "loss": 1.1031,
+      "step": 569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3609163848334858,
+      "learning_rate": 9.998904140806547e-06,
+      "loss": 0.7157,
+      "step": 570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2680207237393666,
+      "learning_rate": 9.998882334010345e-06,
+      "loss": 1.01,
+      "step": 571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7540470573806264,
+      "learning_rate": 9.998860312401035e-06,
+      "loss": 1.107,
+      "step": 572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8367755898124862,
+      "learning_rate": 9.998838075979566e-06,
+      "loss": 0.938,
+      "step": 573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1659805128406844,
+      "learning_rate": 9.99881562474689e-06,
+      "loss": 0.6925,
+      "step": 574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0274343262767953,
+      "learning_rate": 9.998792958703975e-06,
+      "loss": 1.2977,
+      "step": 575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.477290784039585,
+      "learning_rate": 9.998770077851794e-06,
+      "loss": 1.0534,
+      "step": 576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.979066294682934,
+      "learning_rate": 9.998746982191327e-06,
+      "loss": 0.8906,
+      "step": 577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0620874416749086,
+      "learning_rate": 9.998723671723572e-06,
+      "loss": 1.0663,
+      "step": 578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4315449770014337,
+      "learning_rate": 9.998700146449528e-06,
+      "loss": 1.2804,
+      "step": 579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.610275402229476,
+      "learning_rate": 9.998676406370205e-06,
+      "loss": 1.1177,
+      "step": 580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.420758512001045,
+      "learning_rate": 9.998652451486626e-06,
+      "loss": 0.8446,
+      "step": 581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.254445402424304,
+      "learning_rate": 9.998628281799816e-06,
+      "loss": 1.003,
+      "step": 582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.169776928073641,
+      "learning_rate": 9.998603897310818e-06,
+      "loss": 1.0646,
+      "step": 583
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.495121857085088,
+      "learning_rate": 9.998579298020676e-06,
+      "loss": 1.0755,
+      "step": 584
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2082891826456628,
+      "learning_rate": 9.998554483930452e-06,
+      "loss": 1.0939,
+      "step": 585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2003074775090345,
+      "learning_rate": 9.998529455041209e-06,
+      "loss": 1.0259,
+      "step": 586
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7380862709300318,
+      "learning_rate": 9.998504211354021e-06,
+      "loss": 0.9837,
+      "step": 587
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.093612836783646,
+      "learning_rate": 9.998478752869976e-06,
+      "loss": 0.9764,
+      "step": 588
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.661446243185641,
+      "learning_rate": 9.998453079590167e-06,
+      "loss": 0.895,
+      "step": 589
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5041915306498415,
+      "learning_rate": 9.998427191515696e-06,
+      "loss": 0.9018,
+      "step": 590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1831525060544625,
+      "learning_rate": 9.998401088647677e-06,
+      "loss": 1.1667,
+      "step": 591
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5464170949355265,
+      "learning_rate": 9.998374770987232e-06,
+      "loss": 1.0953,
+      "step": 592
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0264561568942945,
+      "learning_rate": 9.99834823853549e-06,
+      "loss": 0.9903,
+      "step": 593
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.414932245083654,
+      "learning_rate": 9.998321491293592e-06,
+      "loss": 0.6315,
+      "step": 594
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.80569175671623,
+      "learning_rate": 9.998294529262688e-06,
+      "loss": 0.8861,
+      "step": 595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.072919935772463,
+      "learning_rate": 9.998267352443938e-06,
+      "loss": 0.9369,
+      "step": 596
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5396092829592847,
+      "learning_rate": 9.998239960838505e-06,
+      "loss": 0.7342,
+      "step": 597
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2648448317539747,
+      "learning_rate": 9.998212354447572e-06,
+      "loss": 0.9441,
+      "step": 598
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3615315503294565,
+      "learning_rate": 9.998184533272321e-06,
+      "loss": 1.0122,
+      "step": 599
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5572577838770068,
+      "learning_rate": 9.99815649731395e-06,
+      "loss": 1.1404,
+      "step": 600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.949746280814809,
+      "learning_rate": 9.998128246573663e-06,
+      "loss": 1.0985,
+      "step": 601
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.191468233276588,
+      "learning_rate": 9.998099781052673e-06,
+      "loss": 0.9179,
+      "step": 602
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3278603543946879,
+      "learning_rate": 9.998071100752203e-06,
+      "loss": 0.6247,
+      "step": 603
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.084786856190264,
+      "learning_rate": 9.998042205673489e-06,
+      "loss": 0.9181,
+      "step": 604
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6749537015651343,
+      "learning_rate": 9.998013095817768e-06,
+      "loss": 0.8704,
+      "step": 605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4965951388821113,
+      "learning_rate": 9.997983771186293e-06,
+      "loss": 1.0579,
+      "step": 606
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9092568734427187,
+      "learning_rate": 9.997954231780326e-06,
+      "loss": 0.7881,
+      "step": 607
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.548464803657111,
+      "learning_rate": 9.997924477601134e-06,
+      "loss": 1.0262,
+      "step": 608
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.001881718430198,
+      "learning_rate": 9.997894508649995e-06,
+      "loss": 0.9935,
+      "step": 609
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4018106175988208,
+      "learning_rate": 9.9978643249282e-06,
+      "loss": 1.088,
+      "step": 610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.221790203808282,
+      "learning_rate": 9.997833926437043e-06,
+      "loss": 0.9236,
+      "step": 611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.129480340066958,
+      "learning_rate": 9.997803313177831e-06,
+      "loss": 0.6573,
+      "step": 612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3520794177852884,
+      "learning_rate": 9.997772485151881e-06,
+      "loss": 0.8992,
+      "step": 613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0410635960598675,
+      "learning_rate": 9.997741442360515e-06,
+      "loss": 0.958,
+      "step": 614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8668703233212938,
+      "learning_rate": 9.997710184805072e-06,
+      "loss": 0.8187,
+      "step": 615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.282989966018686,
+      "learning_rate": 9.997678712486889e-06,
+      "loss": 1.0014,
+      "step": 616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4008705233673173,
+      "learning_rate": 9.997647025407322e-06,
+      "loss": 0.768,
+      "step": 617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.850940009967221,
+      "learning_rate": 9.997615123567733e-06,
+      "loss": 1.0267,
+      "step": 618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.203630342580387,
+      "learning_rate": 9.99758300696949e-06,
+      "loss": 0.6838,
+      "step": 619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5826318805784894,
+      "learning_rate": 9.997550675613976e-06,
+      "loss": 1.1435,
+      "step": 620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.40780857502491,
+      "learning_rate": 9.997518129502579e-06,
+      "loss": 0.9035,
+      "step": 621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3496193651495907,
+      "learning_rate": 9.997485368636698e-06,
+      "loss": 1.0192,
+      "step": 622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.844636450153469,
+      "learning_rate": 9.99745239301774e-06,
+      "loss": 1.0218,
+      "step": 623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4455274951579486,
+      "learning_rate": 9.997419202647124e-06,
+      "loss": 0.7465,
+      "step": 624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.154502753920736,
+      "learning_rate": 9.997385797526275e-06,
+      "loss": 0.6623,
+      "step": 625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0819000960415712,
+      "learning_rate": 9.997352177656627e-06,
+      "loss": 1.0073,
+      "step": 626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6529275868823574,
+      "learning_rate": 9.997318343039629e-06,
+      "loss": 0.8574,
+      "step": 627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2495153805396035,
+      "learning_rate": 9.99728429367673e-06,
+      "loss": 0.9016,
+      "step": 628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.330237413191616,
+      "learning_rate": 9.997250029569395e-06,
+      "loss": 1.0196,
+      "step": 629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.563381275324097,
+      "learning_rate": 9.997215550719097e-06,
+      "loss": 0.9898,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0795605278581113,
+      "learning_rate": 9.997180857127318e-06,
+      "loss": 1.0497,
+      "step": 631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1551610437871775,
+      "learning_rate": 9.997145948795548e-06,
+      "loss": 0.9529,
+      "step": 632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.516962102773063,
+      "learning_rate": 9.997110825725288e-06,
+      "loss": 0.984,
+      "step": 633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5766582899818413,
+      "learning_rate": 9.997075487918047e-06,
+      "loss": 0.9879,
+      "step": 634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3014529880713592,
+      "learning_rate": 9.99703993537534e-06,
+      "loss": 1.0165,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4711518515485413,
+      "learning_rate": 9.997004168098703e-06,
+      "loss": 0.813,
+      "step": 636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3809473444166545,
+      "learning_rate": 9.996968186089664e-06,
+      "loss": 0.7337,
+      "step": 637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.118836813005589,
+      "learning_rate": 9.996931989349775e-06,
+      "loss": 0.9711,
+      "step": 638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6406398390887196,
+      "learning_rate": 9.99689557788059e-06,
+      "loss": 0.8922,
+      "step": 639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8407300068260204,
+      "learning_rate": 9.996858951683672e-06,
+      "loss": 0.8347,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.606967130792558,
+      "learning_rate": 9.996822110760595e-06,
+      "loss": 0.9739,
+      "step": 641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9818182627004193,
+      "learning_rate": 9.996785055112947e-06,
+      "loss": 1.1584,
+      "step": 642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2966972001325505,
+      "learning_rate": 9.996747784742314e-06,
+      "loss": 0.6296,
+      "step": 643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3347290590544083,
+      "learning_rate": 9.996710299650302e-06,
+      "loss": 1.097,
+      "step": 644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9925766339774196,
+      "learning_rate": 9.99667259983852e-06,
+      "loss": 1.0349,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.246550809780768,
+      "learning_rate": 9.996634685308586e-06,
+      "loss": 1.0152,
+      "step": 646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.619541781013028,
+      "learning_rate": 9.996596556062133e-06,
+      "loss": 1.0511,
+      "step": 647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5133004273450816,
+      "learning_rate": 9.9965582121008e-06,
+      "loss": 0.9247,
+      "step": 648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.948421147381752,
+      "learning_rate": 9.996519653426229e-06,
+      "loss": 0.9483,
+      "step": 649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3759946401103003,
+      "learning_rate": 9.996480880040083e-06,
+      "loss": 1.1432,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7249394584056534,
+      "learning_rate": 9.996441891944023e-06,
+      "loss": 1.1598,
+      "step": 651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0430452451897914,
+      "learning_rate": 9.99640268913973e-06,
+      "loss": 1.1625,
+      "step": 652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2779336682310993,
+      "learning_rate": 9.996363271628886e-06,
+      "loss": 1.0535,
+      "step": 653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4214045632165484,
+      "learning_rate": 9.996323639413185e-06,
+      "loss": 0.7882,
+      "step": 654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7998872576684044,
+      "learning_rate": 9.996283792494328e-06,
+      "loss": 0.9709,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.186933784132359,
+      "learning_rate": 9.99624373087403e-06,
+      "loss": 0.9641,
+      "step": 656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.46894859274919,
+      "learning_rate": 9.996203454554011e-06,
+      "loss": 1.09,
+      "step": 657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.415071217463598,
+      "learning_rate": 9.996162963536004e-06,
+      "loss": 0.8791,
+      "step": 658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2241138095746367,
+      "learning_rate": 9.996122257821746e-06,
+      "loss": 0.6461,
+      "step": 659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8815218982746185,
+      "learning_rate": 9.996081337412988e-06,
+      "loss": 0.9658,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8799403380539346,
+      "learning_rate": 9.996040202311488e-06,
+      "loss": 0.9978,
+      "step": 661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.375239202455436,
+      "learning_rate": 9.995998852519014e-06,
+      "loss": 0.8918,
+      "step": 662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3168980364908407,
+      "learning_rate": 9.995957288037343e-06,
+      "loss": 1.3134,
+      "step": 663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0271385531583324,
+      "learning_rate": 9.99591550886826e-06,
+      "loss": 0.5918,
+      "step": 664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.685618451327792,
+      "learning_rate": 9.995873515013562e-06,
+      "loss": 0.9753,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8897253934951626,
+      "learning_rate": 9.99583130647505e-06,
+      "loss": 1.0045,
+      "step": 666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.820716572433625,
+      "learning_rate": 9.995788883254543e-06,
+      "loss": 0.9435,
+      "step": 667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4060092306294623,
+      "learning_rate": 9.995746245353861e-06,
+      "loss": 1.0705,
+      "step": 668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.112361133569284,
+      "learning_rate": 9.995703392774836e-06,
+      "loss": 0.8651,
+      "step": 669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.483913049844126,
+      "learning_rate": 9.99566032551931e-06,
+      "loss": 0.9748,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.179721096749841,
+      "learning_rate": 9.995617043589135e-06,
+      "loss": 1.0766,
+      "step": 671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.642063218763466,
+      "learning_rate": 9.99557354698617e-06,
+      "loss": 1.0432,
+      "step": 672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3144664391572594,
+      "learning_rate": 9.995529835712283e-06,
+      "loss": 0.7062,
+      "step": 673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0447860983009654,
+      "learning_rate": 9.995485909769354e-06,
+      "loss": 0.6614,
+      "step": 674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9600724134554737,
+      "learning_rate": 9.995441769159269e-06,
+      "loss": 0.9208,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.096210849423815,
+      "learning_rate": 9.995397413883928e-06,
+      "loss": 1.13,
+      "step": 676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0761727577857276,
+      "learning_rate": 9.995352843945232e-06,
+      "loss": 0.8588,
+      "step": 677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0855640985094466,
+      "learning_rate": 9.9953080593451e-06,
+      "loss": 0.6859,
+      "step": 678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.160575590855045,
+      "learning_rate": 9.995263060085456e-06,
+      "loss": 1.0198,
+      "step": 679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3751110736672665,
+      "learning_rate": 9.995217846168232e-06,
+      "loss": 0.8455,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7458577583065464,
+      "learning_rate": 9.995172417595373e-06,
+      "loss": 1.0537,
+      "step": 681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9205512458820886,
+      "learning_rate": 9.995126774368831e-06,
+      "loss": 1.1581,
+      "step": 682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6026235595070806,
+      "learning_rate": 9.995080916490566e-06,
+      "loss": 0.967,
+      "step": 683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.247160877094602,
+      "learning_rate": 9.99503484396255e-06,
+      "loss": 1.0156,
+      "step": 684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7294236899484523,
+      "learning_rate": 9.994988556786763e-06,
+      "loss": 0.971,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4891576016744135,
+      "learning_rate": 9.99494205496519e-06,
+      "loss": 1.0191,
+      "step": 686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.705525121860567,
+      "learning_rate": 9.994895338499835e-06,
+      "loss": 0.7554,
+      "step": 687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2794167983531615,
+      "learning_rate": 9.994848407392703e-06,
+      "loss": 0.6579,
+      "step": 688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.567922024639386,
+      "learning_rate": 9.99480126164581e-06,
+      "loss": 0.967,
+      "step": 689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.153892949839661,
+      "learning_rate": 9.994753901261183e-06,
+      "loss": 0.6861,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.651324376803948,
+      "learning_rate": 9.994706326240858e-06,
+      "loss": 1.0624,
+      "step": 691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1522943865273483,
+      "learning_rate": 9.994658536586878e-06,
+      "loss": 1.0131,
+      "step": 692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5224766453366327,
+      "learning_rate": 9.994610532301296e-06,
+      "loss": 0.9733,
+      "step": 693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.881949692931813,
+      "learning_rate": 9.994562313386177e-06,
+      "loss": 0.9074,
+      "step": 694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9093658528585067,
+      "learning_rate": 9.99451387984359e-06,
+      "loss": 0.9016,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6501916102386285,
+      "learning_rate": 9.99446523167562e-06,
+      "loss": 0.9257,
+      "step": 696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.212135620198715,
+      "learning_rate": 9.994416368884357e-06,
+      "loss": 0.6378,
+      "step": 697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0235445758355843,
+      "learning_rate": 9.994367291471896e-06,
+      "loss": 0.8211,
+      "step": 698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.446371397914466,
+      "learning_rate": 9.994317999440351e-06,
+      "loss": 0.8857,
+      "step": 699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3062050516955206,
+      "learning_rate": 9.99426849279184e-06,
+      "loss": 1.0366,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.270850391140929,
+      "learning_rate": 9.994218771528487e-06,
+      "loss": 0.6849,
+      "step": 701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.749401168208638,
+      "learning_rate": 9.994168835652432e-06,
+      "loss": 1.0733,
+      "step": 702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8818394728014587,
+      "learning_rate": 9.994118685165819e-06,
+      "loss": 1.0051,
+      "step": 703
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3660358334821576,
+      "learning_rate": 9.994068320070805e-06,
+      "loss": 1.1532,
+      "step": 704
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4265503080885886,
+      "learning_rate": 9.99401774036955e-06,
+      "loss": 0.9864,
+      "step": 705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.414373507821136,
+      "learning_rate": 9.993966946064233e-06,
+      "loss": 0.8994,
+      "step": 706
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2394647634050555,
+      "learning_rate": 9.993915937157033e-06,
+      "loss": 0.8932,
+      "step": 707
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.86094619728617,
+      "learning_rate": 9.993864713650143e-06,
+      "loss": 1.0945,
+      "step": 708
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3278143178498483,
+      "learning_rate": 9.993813275545764e-06,
+      "loss": 0.7569,
+      "step": 709
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8325938219491484,
+      "learning_rate": 9.993761622846107e-06,
+      "loss": 0.9515,
+      "step": 710
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.81952822078733,
+      "learning_rate": 9.99370975555339e-06,
+      "loss": 0.9671,
+      "step": 711
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5602360974603984,
+      "learning_rate": 9.993657673669848e-06,
+      "loss": 0.7635,
+      "step": 712
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.310256471423168,
+      "learning_rate": 9.993605377197711e-06,
+      "loss": 0.8212,
+      "step": 713
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7142530010054724,
+      "learning_rate": 9.99355286613923e-06,
+      "loss": 0.7509,
+      "step": 714
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.326192175462013,
+      "learning_rate": 9.993500140496659e-06,
+      "loss": 1.0056,
+      "step": 715
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.657252766557116,
+      "learning_rate": 9.993447200272269e-06,
+      "loss": 0.9779,
+      "step": 716
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2271779486062024,
+      "learning_rate": 9.993394045468329e-06,
+      "loss": 0.7617,
+      "step": 717
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3857915970627883,
+      "learning_rate": 9.993340676087129e-06,
+      "loss": 1.061,
+      "step": 718
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5642322972473166,
+      "learning_rate": 9.993287092130956e-06,
+      "loss": 1.0529,
+      "step": 719
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3369473300001509,
+      "learning_rate": 9.993233293602119e-06,
+      "loss": 0.7097,
+      "step": 720
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2462875468759083,
+      "learning_rate": 9.993179280502926e-06,
+      "loss": 0.8425,
+      "step": 721
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.388041255588855,
+      "learning_rate": 9.993125052835699e-06,
+      "loss": 1.0131,
+      "step": 722
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2518879396508806,
+      "learning_rate": 9.993070610602766e-06,
+      "loss": 0.9765,
+      "step": 723
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2296477065613767,
+      "learning_rate": 9.993015953806472e-06,
+      "loss": 0.7938,
+      "step": 724
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2142541671853622,
+      "learning_rate": 9.992961082449161e-06,
+      "loss": 0.7811,
+      "step": 725
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4304338330862163,
+      "learning_rate": 9.992905996533193e-06,
+      "loss": 1.0904,
+      "step": 726
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.884380082970056,
+      "learning_rate": 9.992850696060933e-06,
+      "loss": 0.9714,
+      "step": 727
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.091126040033933,
+      "learning_rate": 9.99279518103476e-06,
+      "loss": 0.8688,
+      "step": 728
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.675064288866467,
+      "learning_rate": 9.992739451457058e-06,
+      "loss": 1.0032,
+      "step": 729
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7653811477025165,
+      "learning_rate": 9.992683507330223e-06,
+      "loss": 1.0565,
+      "step": 730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.159463832088998,
+      "learning_rate": 9.992627348656658e-06,
+      "loss": 0.9638,
+      "step": 731
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.31001077268034,
+      "learning_rate": 9.992570975438776e-06,
+      "loss": 0.8325,
+      "step": 732
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.146235203726364,
+      "learning_rate": 9.992514387679e-06,
+      "loss": 1.0726,
+      "step": 733
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.641418414443428,
+      "learning_rate": 9.992457585379764e-06,
+      "loss": 0.9666,
+      "step": 734
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.290101770774778,
+      "learning_rate": 9.992400568543506e-06,
+      "loss": 1.2662,
+      "step": 735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.471184520596426,
+      "learning_rate": 9.992343337172674e-06,
+      "loss": 1.0166,
+      "step": 736
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3025282469627169,
+      "learning_rate": 9.992285891269732e-06,
+      "loss": 0.6478,
+      "step": 737
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.909097852720877,
+      "learning_rate": 9.992228230837148e-06,
+      "loss": 1.14,
+      "step": 738
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.060107653297737,
+      "learning_rate": 9.992170355877398e-06,
+      "loss": 1.1664,
+      "step": 739
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4453104598635416,
+      "learning_rate": 9.99211226639297e-06,
+      "loss": 0.8341,
+      "step": 740
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2764562264501547,
+      "learning_rate": 9.99205396238636e-06,
+      "loss": 1.0836,
+      "step": 741
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4159734876139782,
+      "learning_rate": 9.991995443860074e-06,
+      "loss": 0.9362,
+      "step": 742
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2711109890219916,
+      "learning_rate": 9.991936710816624e-06,
+      "loss": 0.8823,
+      "step": 743
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 34.451789957523665,
+      "learning_rate": 9.991877763258538e-06,
+      "loss": 0.8909,
+      "step": 744
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.050754024325341,
+      "learning_rate": 9.991818601188345e-06,
+      "loss": 0.8971,
+      "step": 745
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5125173097754505,
+      "learning_rate": 9.99175922460859e-06,
+      "loss": 0.969,
+      "step": 746
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.723604469254173,
+      "learning_rate": 9.991699633521823e-06,
+      "loss": 0.9669,
+      "step": 747
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3153555759373297,
+      "learning_rate": 9.991639827930607e-06,
+      "loss": 1.0364,
+      "step": 748
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7222944376888685,
+      "learning_rate": 9.991579807837511e-06,
+      "loss": 1.1033,
+      "step": 749
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.810498284170168,
+      "learning_rate": 9.991519573245112e-06,
+      "loss": 1.0246,
+      "step": 750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.643482817155567,
+      "learning_rate": 9.991459124156003e-06,
+      "loss": 0.8906,
+      "step": 751
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1887965960178464,
+      "learning_rate": 9.991398460572776e-06,
+      "loss": 0.7987,
+      "step": 752
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7257795511145066,
+      "learning_rate": 9.991337582498042e-06,
+      "loss": 0.9061,
+      "step": 753
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1496901540964983,
+      "learning_rate": 9.991276489934416e-06,
+      "loss": 1.1042,
+      "step": 754
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3126161202833875,
+      "learning_rate": 9.991215182884521e-06,
+      "loss": 0.9522,
+      "step": 755
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4431973483952205,
+      "learning_rate": 9.991153661350996e-06,
+      "loss": 0.6953,
+      "step": 756
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3144635374601603,
+      "learning_rate": 9.991091925336481e-06,
+      "loss": 1.0677,
+      "step": 757
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5424279007489923,
+      "learning_rate": 9.991029974843632e-06,
+      "loss": 0.951,
+      "step": 758
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.107703904137036,
+      "learning_rate": 9.990967809875107e-06,
+      "loss": 0.9809,
+      "step": 759
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4938224151155186,
+      "learning_rate": 9.990905430433582e-06,
+      "loss": 0.9457,
+      "step": 760
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8970983053498265,
+      "learning_rate": 9.990842836521732e-06,
+      "loss": 0.9363,
+      "step": 761
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6387231485420983,
+      "learning_rate": 9.990780028142253e-06,
+      "loss": 1.0698,
+      "step": 762
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.558391496532495,
+      "learning_rate": 9.990717005297841e-06,
+      "loss": 0.9769,
+      "step": 763
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.20283650025806,
+      "learning_rate": 9.990653767991203e-06,
+      "loss": 0.9185,
+      "step": 764
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.320109845984629,
+      "learning_rate": 9.99059031622506e-06,
+      "loss": 1.0556,
+      "step": 765
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8938850165575114,
+      "learning_rate": 9.990526650002133e-06,
+      "loss": 1.0852,
+      "step": 766
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1672473320242656,
+      "learning_rate": 9.990462769325163e-06,
+      "loss": 0.6906,
+      "step": 767
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.773210681103188,
+      "learning_rate": 9.990398674196896e-06,
+      "loss": 1.0282,
+      "step": 768
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.807831970633632,
+      "learning_rate": 9.99033436462008e-06,
+      "loss": 1.0064,
+      "step": 769
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.928588816000217,
+      "learning_rate": 9.990269840597484e-06,
+      "loss": 1.0344,
+      "step": 770
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9984600887239146,
+      "learning_rate": 9.990205102131879e-06,
+      "loss": 0.8862,
+      "step": 771
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4859127776654177,
+      "learning_rate": 9.990140149226047e-06,
+      "loss": 1.0292,
+      "step": 772
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.615067971934033,
+      "learning_rate": 9.99007498188278e-06,
+      "loss": 0.9854,
+      "step": 773
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7566081201620425,
+      "learning_rate": 9.990009600104875e-06,
+      "loss": 0.8869,
+      "step": 774
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.539444313786837,
+      "learning_rate": 9.989944003895146e-06,
+      "loss": 1.0719,
+      "step": 775
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7107197056309227,
+      "learning_rate": 9.98987819325641e-06,
+      "loss": 0.994,
+      "step": 776
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9172904751938495,
+      "learning_rate": 9.989812168191495e-06,
+      "loss": 0.7355,
+      "step": 777
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3972269963284867,
+      "learning_rate": 9.989745928703237e-06,
+      "loss": 0.8858,
+      "step": 778
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0354879424326384,
+      "learning_rate": 9.989679474794484e-06,
+      "loss": 1.1532,
+      "step": 779
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.649483524189425,
+      "learning_rate": 9.989612806468092e-06,
+      "loss": 0.9515,
+      "step": 780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7474816129442416,
+      "learning_rate": 9.989545923726925e-06,
+      "loss": 1.0673,
+      "step": 781
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.2421838848006885,
+      "learning_rate": 9.989478826573857e-06,
+      "loss": 0.8266,
+      "step": 782
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1232130505607185,
+      "learning_rate": 9.989411515011773e-06,
+      "loss": 0.8449,
+      "step": 783
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2032283995188013,
+      "learning_rate": 9.989343989043563e-06,
+      "loss": 0.9028,
+      "step": 784
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7723749435031526,
+      "learning_rate": 9.989276248672131e-06,
+      "loss": 0.9017,
+      "step": 785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0965553597888253,
+      "learning_rate": 9.989208293900386e-06,
+      "loss": 0.6676,
+      "step": 786
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5779598021704206,
+      "learning_rate": 9.98914012473125e-06,
+      "loss": 1.0885,
+      "step": 787
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8271784661510457,
+      "learning_rate": 9.989071741167649e-06,
+      "loss": 0.9766,
+      "step": 788
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2153271092936695,
+      "learning_rate": 9.989003143212526e-06,
+      "loss": 0.745,
+      "step": 789
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3374671810897698,
+      "learning_rate": 9.988934330868826e-06,
+      "loss": 0.7473,
+      "step": 790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8441437019288969,
+      "learning_rate": 9.988865304139509e-06,
+      "loss": 0.9312,
+      "step": 791
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2232297258888005,
+      "learning_rate": 9.988796063027534e-06,
+      "loss": 0.9353,
+      "step": 792
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.208047098500028,
+      "learning_rate": 9.988726607535885e-06,
+      "loss": 0.6177,
+      "step": 793
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4312190089557726,
+      "learning_rate": 9.988656937667544e-06,
+      "loss": 0.9203,
+      "step": 794
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9262896296861243,
+      "learning_rate": 9.988587053425504e-06,
+      "loss": 0.9592,
+      "step": 795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6566267558730954,
+      "learning_rate": 9.988516954812767e-06,
+      "loss": 1.0421,
+      "step": 796
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6687244377188386,
+      "learning_rate": 9.988446641832346e-06,
+      "loss": 0.8557,
+      "step": 797
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.427292346418425,
+      "learning_rate": 9.988376114487264e-06,
+      "loss": 0.9255,
+      "step": 798
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.706735286254915,
+      "learning_rate": 9.98830537278055e-06,
+      "loss": 1.0327,
+      "step": 799
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6324424509464848,
+      "learning_rate": 9.988234416715246e-06,
+      "loss": 1.0427,
+      "step": 800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.293437039516723,
+      "learning_rate": 9.988163246294398e-06,
+      "loss": 0.9357,
+      "step": 801
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5641043850494833,
+      "learning_rate": 9.988091861521067e-06,
+      "loss": 0.941,
+      "step": 802
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6184277672121907,
+      "learning_rate": 9.98802026239832e-06,
+      "loss": 0.8469,
+      "step": 803
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.748765052723711,
+      "learning_rate": 9.987948448929232e-06,
+      "loss": 0.9056,
+      "step": 804
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9428537269772064,
+      "learning_rate": 9.98787642111689e-06,
+      "loss": 0.9427,
+      "step": 805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.720689658944206,
+      "learning_rate": 9.987804178964391e-06,
+      "loss": 0.8494,
+      "step": 806
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7854426275602915,
+      "learning_rate": 9.987731722474839e-06,
+      "loss": 1.0675,
+      "step": 807
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0717818570268536,
+      "learning_rate": 9.987659051651345e-06,
+      "loss": 0.9678,
+      "step": 808
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0646440618555713,
+      "learning_rate": 9.987586166497032e-06,
+      "loss": 0.8356,
+      "step": 809
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3562165555689742,
+      "learning_rate": 9.987513067015035e-06,
+      "loss": 1.014,
+      "step": 810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1038413161073717,
+      "learning_rate": 9.987439753208494e-06,
+      "loss": 0.6442,
+      "step": 811
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8937993357356575,
+      "learning_rate": 9.98736622508056e-06,
+      "loss": 1.0075,
+      "step": 812
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4479609488343423,
+      "learning_rate": 9.98729248263439e-06,
+      "loss": 0.9926,
+      "step": 813
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2226748877424427,
+      "learning_rate": 9.987218525873155e-06,
+      "loss": 0.8662,
+      "step": 814
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0335497244448373,
+      "learning_rate": 9.987144354800035e-06,
+      "loss": 1.023,
+      "step": 815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.628238673443518,
+      "learning_rate": 9.987069969418213e-06,
+      "loss": 1.0044,
+      "step": 816
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.339437574234407,
+      "learning_rate": 9.986995369730887e-06,
+      "loss": 0.9809,
+      "step": 817
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2525782855670613,
+      "learning_rate": 9.986920555741267e-06,
+      "loss": 0.7421,
+      "step": 818
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9487013457374487,
+      "learning_rate": 9.98684552745256e-06,
+      "loss": 0.804,
+      "step": 819
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2744055369205252,
+      "learning_rate": 9.986770284867997e-06,
+      "loss": 0.7018,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.120908675044228,
+      "learning_rate": 9.986694827990807e-06,
+      "loss": 0.7197,
+      "step": 821
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0410890668260753,
+      "learning_rate": 9.986619156824236e-06,
+      "loss": 0.8786,
+      "step": 822
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.418769826027286,
+      "learning_rate": 9.986543271371532e-06,
+      "loss": 0.8111,
+      "step": 823
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1833115470461713,
+      "learning_rate": 9.98646717163596e-06,
+      "loss": 0.8615,
+      "step": 824
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6455817026176582,
+      "learning_rate": 9.986390857620787e-06,
+      "loss": 0.9787,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.534536485743664,
+      "learning_rate": 9.986314329329294e-06,
+      "loss": 0.8602,
+      "step": 826
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.411732851962556,
+      "learning_rate": 9.986237586764768e-06,
+      "loss": 1.2201,
+      "step": 827
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.143908136471734,
+      "learning_rate": 9.98616062993051e-06,
+      "loss": 1.0787,
+      "step": 828
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9260732339085447,
+      "learning_rate": 9.986083458829824e-06,
+      "loss": 0.8778,
+      "step": 829
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8704352042580585,
+      "learning_rate": 9.986006073466027e-06,
+      "loss": 0.7694,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.600476158956252,
+      "learning_rate": 9.985928473842445e-06,
+      "loss": 0.9446,
+      "step": 831
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.560165461229762,
+      "learning_rate": 9.985850659962411e-06,
+      "loss": 0.8529,
+      "step": 832
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3396947141276843,
+      "learning_rate": 9.985772631829272e-06,
+      "loss": 0.8486,
+      "step": 833
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.502407238039646,
+      "learning_rate": 9.985694389446378e-06,
+      "loss": 0.9032,
+      "step": 834
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.499186038057688,
+      "learning_rate": 9.985615932817092e-06,
+      "loss": 0.9477,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.072361795391179,
+      "learning_rate": 9.985537261944787e-06,
+      "loss": 0.8768,
+      "step": 836
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.316638397316478,
+      "learning_rate": 9.985458376832842e-06,
+      "loss": 0.9853,
+      "step": 837
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.274751797470494,
+      "learning_rate": 9.985379277484646e-06,
+      "loss": 0.9457,
+      "step": 838
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5688005204893978,
+      "learning_rate": 9.9852999639036e-06,
+      "loss": 0.9769,
+      "step": 839
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5549421741579397,
+      "learning_rate": 9.985220436093112e-06,
+      "loss": 0.945,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6463457811041504,
+      "learning_rate": 9.985140694056601e-06,
+      "loss": 1.0427,
+      "step": 841
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2109865386467835,
+      "learning_rate": 9.985060737797489e-06,
+      "loss": 0.6517,
+      "step": 842
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8876932954376238,
+      "learning_rate": 9.984980567319216e-06,
+      "loss": 1.0178,
+      "step": 843
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7722512604902727,
+      "learning_rate": 9.984900182625226e-06,
+      "loss": 0.8277,
+      "step": 844
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6775996310466152,
+      "learning_rate": 9.984819583718973e-06,
+      "loss": 1.0343,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2965615575184968,
+      "learning_rate": 9.984738770603921e-06,
+      "loss": 1.201,
+      "step": 846
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.205035486660309,
+      "learning_rate": 9.984657743283543e-06,
+      "loss": 0.9633,
+      "step": 847
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4548463284809316,
+      "learning_rate": 9.984576501761319e-06,
+      "loss": 0.8789,
+      "step": 848
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5354523594797107,
+      "learning_rate": 9.98449504604074e-06,
+      "loss": 1.1298,
+      "step": 849
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8942607864204828,
+      "learning_rate": 9.984413376125312e-06,
+      "loss": 0.9437,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4065744020602025,
+      "learning_rate": 9.984331492018537e-06,
+      "loss": 1.0045,
+      "step": 851
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0921087594453494,
+      "learning_rate": 9.984249393723939e-06,
+      "loss": 0.9759,
+      "step": 852
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1561406673026764,
+      "learning_rate": 9.984167081245044e-06,
+      "loss": 0.7974,
+      "step": 853
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2486770144250943,
+      "learning_rate": 9.984084554585387e-06,
+      "loss": 1.0332,
+      "step": 854
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9423177009604227,
+      "learning_rate": 9.984001813748519e-06,
+      "loss": 1.0453,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.000242861200542,
+      "learning_rate": 9.98391885873799e-06,
+      "loss": 0.9932,
+      "step": 856
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3862754149603194,
+      "learning_rate": 9.98383568955737e-06,
+      "loss": 0.7034,
+      "step": 857
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.233765930282959,
+      "learning_rate": 9.983752306210233e-06,
+      "loss": 0.8895,
+      "step": 858
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9046289544046022,
+      "learning_rate": 9.983668708700156e-06,
+      "loss": 1.114,
+      "step": 859
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.0803737771428334,
+      "learning_rate": 9.983584897030737e-06,
+      "loss": 0.7656,
+      "step": 860
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8256530940933056,
+      "learning_rate": 9.983500871205577e-06,
+      "loss": 0.995,
+      "step": 861
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3932660612158725,
+      "learning_rate": 9.983416631228284e-06,
+      "loss": 0.8624,
+      "step": 862
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0878703747435634,
+      "learning_rate": 9.98333217710248e-06,
+      "loss": 0.7534,
+      "step": 863
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.833804439913853,
+      "learning_rate": 9.983247508831795e-06,
+      "loss": 0.9305,
+      "step": 864
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.546060372790528,
+      "learning_rate": 9.983162626419866e-06,
+      "loss": 0.9508,
+      "step": 865
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.517071668236195,
+      "learning_rate": 9.98307752987034e-06,
+      "loss": 1.0582,
+      "step": 866
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.627894690995411,
+      "learning_rate": 9.982992219186876e-06,
+      "loss": 0.9857,
+      "step": 867
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.223916832526487,
+      "learning_rate": 9.982906694373136e-06,
+      "loss": 0.7965,
+      "step": 868
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.127367979728812,
+      "learning_rate": 9.9828209554328e-06,
+      "loss": 0.6446,
+      "step": 869
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.492566612199719,
+      "learning_rate": 9.98273500236955e-06,
+      "loss": 1.0985,
+      "step": 870
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.87639868928895,
+      "learning_rate": 9.98264883518708e-06,
+      "loss": 1.1268,
+      "step": 871
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5212865566450917,
+      "learning_rate": 9.98256245388909e-06,
+      "loss": 0.9666,
+      "step": 872
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.097074436431164,
+      "learning_rate": 9.982475858479298e-06,
+      "loss": 0.9028,
+      "step": 873
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.102922657543181,
+      "learning_rate": 9.982389048961421e-06,
+      "loss": 1.0693,
+      "step": 874
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.550705683788794,
+      "learning_rate": 9.98230202533919e-06,
+      "loss": 0.9,
+      "step": 875
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.30793089467903,
+      "learning_rate": 9.982214787616345e-06,
+      "loss": 0.9465,
+      "step": 876
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.821418425975904,
+      "learning_rate": 9.982127335796631e-06,
+      "loss": 0.8458,
+      "step": 877
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2580447622714221,
+      "learning_rate": 9.982039669883815e-06,
+      "loss": 0.7299,
+      "step": 878
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.833497361732783,
+      "learning_rate": 9.981951789881657e-06,
+      "loss": 0.9555,
+      "step": 879
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3339231426261873,
+      "learning_rate": 9.981863695793934e-06,
+      "loss": 0.8987,
+      "step": 880
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.505323733244382,
+      "learning_rate": 9.981775387624433e-06,
+      "loss": 1.0334,
+      "step": 881
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.173593497039518,
+      "learning_rate": 9.98168686537695e-06,
+      "loss": 1.2086,
+      "step": 882
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2513505176932151,
+      "learning_rate": 9.981598129055286e-06,
+      "loss": 0.8356,
+      "step": 883
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0555690565506026,
+      "learning_rate": 9.981509178663256e-06,
+      "loss": 0.6227,
+      "step": 884
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5979276316566806,
+      "learning_rate": 9.981420014204682e-06,
+      "loss": 0.8819,
+      "step": 885
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7613680521749657,
+      "learning_rate": 9.981330635683396e-06,
+      "loss": 0.9902,
+      "step": 886
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.630313850447285,
+      "learning_rate": 9.981241043103239e-06,
+      "loss": 0.994,
+      "step": 887
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5834884232326565,
+      "learning_rate": 9.981151236468061e-06,
+      "loss": 0.9734,
+      "step": 888
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.079706710273589,
+      "learning_rate": 9.98106121578172e-06,
+      "loss": 1.0786,
+      "step": 889
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2694700913254686,
+      "learning_rate": 9.980970981048086e-06,
+      "loss": 1.0213,
+      "step": 890
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4803526925931614,
+      "learning_rate": 9.980880532271033e-06,
+      "loss": 1.004,
+      "step": 891
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.007960882077768,
+      "learning_rate": 9.980789869454454e-06,
+      "loss": 0.9757,
+      "step": 892
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9763282863037734,
+      "learning_rate": 9.980698992602241e-06,
+      "loss": 1.102,
+      "step": 893
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.337962983287217,
+      "learning_rate": 9.980607901718297e-06,
+      "loss": 0.7894,
+      "step": 894
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.07285741271841,
+      "learning_rate": 9.98051659680654e-06,
+      "loss": 0.9365,
+      "step": 895
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.918224358281339,
+      "learning_rate": 9.980425077870895e-06,
+      "loss": 0.8479,
+      "step": 896
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5280153972286463,
+      "learning_rate": 9.98033334491529e-06,
+      "loss": 0.87,
+      "step": 897
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4083643941909352,
+      "learning_rate": 9.98024139794367e-06,
+      "loss": 0.8095,
+      "step": 898
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2562303784384805,
+      "learning_rate": 9.980149236959986e-06,
+      "loss": 0.8769,
+      "step": 899
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.56419215194747,
+      "learning_rate": 9.980056861968197e-06,
+      "loss": 1.1413,
+      "step": 900
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.3268330920256055,
+      "learning_rate": 9.979964272972275e-06,
+      "loss": 1.0867,
+      "step": 901
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4636754395122646,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 0.8774,
+      "step": 902
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5096039845006763,
+      "learning_rate": 9.979778452983949e-06,
+      "loss": 1.0901,
+      "step": 903
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.651348594561017,
+      "learning_rate": 9.979685221999532e-06,
+      "loss": 0.9584,
+      "step": 904
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3931114117857186,
+      "learning_rate": 9.979591777026952e-06,
+      "loss": 0.9313,
+      "step": 905
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3277492457568623,
+      "learning_rate": 9.97949811807022e-06,
+      "loss": 0.8968,
+      "step": 906
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.133435181754004,
+      "learning_rate": 9.979404245133367e-06,
+      "loss": 0.9414,
+      "step": 907
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.803913405004498,
+      "learning_rate": 9.979310158220422e-06,
+      "loss": 0.7982,
+      "step": 908
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3667965945375546,
+      "learning_rate": 9.97921585733543e-06,
+      "loss": 0.9796,
+      "step": 909
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5502756773763542,
+      "learning_rate": 9.979121342482442e-06,
+      "loss": 0.719,
+      "step": 910
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.670893966900724,
+      "learning_rate": 9.979026613665523e-06,
+      "loss": 1.0207,
+      "step": 911
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7883873392395624,
+      "learning_rate": 9.978931670888741e-06,
+      "loss": 0.8712,
+      "step": 912
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9309711636527203,
+      "learning_rate": 9.978836514156177e-06,
+      "loss": 1.0057,
+      "step": 913
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.190992488347459,
+      "learning_rate": 9.97874114347192e-06,
+      "loss": 0.9709,
+      "step": 914
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.149454209188578,
+      "learning_rate": 9.978645558840066e-06,
+      "loss": 1.0955,
+      "step": 915
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.585966096666831,
+      "learning_rate": 9.978549760264724e-06,
+      "loss": 0.8267,
+      "step": 916
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.301016491680075,
+      "learning_rate": 9.978453747750012e-06,
+      "loss": 1.0346,
+      "step": 917
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.523473164960469,
+      "learning_rate": 9.978357521300054e-06,
+      "loss": 1.1684,
+      "step": 918
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.333832414201479,
+      "learning_rate": 9.978261080918988e-06,
+      "loss": 0.9126,
+      "step": 919
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6956693616460456,
+      "learning_rate": 9.978164426610956e-06,
+      "loss": 0.9297,
+      "step": 920
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.73895762219395,
+      "learning_rate": 9.97806755838011e-06,
+      "loss": 1.116,
+      "step": 921
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5132285720351146,
+      "learning_rate": 9.977970476230615e-06,
+      "loss": 1.1786,
+      "step": 922
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.435015334154656,
+      "learning_rate": 9.977873180166641e-06,
+      "loss": 0.9188,
+      "step": 923
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.985653877751031,
+      "learning_rate": 9.977775670192373e-06,
+      "loss": 1.1333,
+      "step": 924
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2183225356313305,
+      "learning_rate": 9.977677946311995e-06,
+      "loss": 0.8764,
+      "step": 925
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0863366615574344,
+      "learning_rate": 9.977580008529713e-06,
+      "loss": 0.698,
+      "step": 926
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2908048165978474,
+      "learning_rate": 9.977481856849729e-06,
+      "loss": 0.6384,
+      "step": 927
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1148981696973044,
+      "learning_rate": 9.977383491276267e-06,
+      "loss": 1.0259,
+      "step": 928
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1437054831993871,
+      "learning_rate": 9.977284911813549e-06,
+      "loss": 0.7905,
+      "step": 929
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0797864727038244,
+      "learning_rate": 9.977186118465813e-06,
+      "loss": 0.848,
+      "step": 930
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.797310593603924,
+      "learning_rate": 9.977087111237307e-06,
+      "loss": 1.0327,
+      "step": 931
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.459983059458689,
+      "learning_rate": 9.976987890132283e-06,
+      "loss": 0.9643,
+      "step": 932
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0495044383646754,
+      "learning_rate": 9.976888455155002e-06,
+      "loss": 0.6943,
+      "step": 933
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.900271012183429,
+      "learning_rate": 9.976788806309742e-06,
+      "loss": 0.9316,
+      "step": 934
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4866460712075047,
+      "learning_rate": 9.976688943600786e-06,
+      "loss": 1.1431,
+      "step": 935
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1780618518861932,
+      "learning_rate": 9.97658886703242e-06,
+      "loss": 0.5771,
+      "step": 936
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0991415843409253,
+      "learning_rate": 9.976488576608945e-06,
+      "loss": 1.0147,
+      "step": 937
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.669991891409403,
+      "learning_rate": 9.976388072334674e-06,
+      "loss": 0.9603,
+      "step": 938
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.594004043216879,
+      "learning_rate": 9.976287354213924e-06,
+      "loss": 0.8758,
+      "step": 939
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.306868244124468,
+      "learning_rate": 9.976186422251026e-06,
+      "loss": 0.8694,
+      "step": 940
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.927003601724445,
+      "learning_rate": 9.976085276450312e-06,
+      "loss": 0.8749,
+      "step": 941
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3362885789029666,
+      "learning_rate": 9.975983916816134e-06,
+      "loss": 0.9064,
+      "step": 942
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1654682191260903,
+      "learning_rate": 9.975882343352843e-06,
+      "loss": 0.9908,
+      "step": 943
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.530415077698997,
+      "learning_rate": 9.975780556064806e-06,
+      "loss": 0.953,
+      "step": 944
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.005732585869039,
+      "learning_rate": 9.975678554956397e-06,
+      "loss": 0.8882,
+      "step": 945
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1178966234007741,
+      "learning_rate": 9.975576340032e-06,
+      "loss": 0.7093,
+      "step": 946
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.183868319263339,
+      "learning_rate": 9.975473911296005e-06,
+      "loss": 0.7674,
+      "step": 947
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5109132700989227,
+      "learning_rate": 9.975371268752816e-06,
+      "loss": 1.0004,
+      "step": 948
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.496436643428661,
+      "learning_rate": 9.975268412406842e-06,
+      "loss": 0.9288,
+      "step": 949
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.053294491309358,
+      "learning_rate": 9.975165342262505e-06,
+      "loss": 1.0231,
+      "step": 950
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.607773634558227,
+      "learning_rate": 9.975062058324232e-06,
+      "loss": 0.8694,
+      "step": 951
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2929580726695191,
+      "learning_rate": 9.974958560596464e-06,
+      "loss": 0.7785,
+      "step": 952
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1794882851304278,
+      "learning_rate": 9.974854849083645e-06,
+      "loss": 0.7439,
+      "step": 953
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.043768204983579,
+      "learning_rate": 9.974750923790234e-06,
+      "loss": 0.662,
+      "step": 954
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6237020846373333,
+      "learning_rate": 9.974646784720696e-06,
+      "loss": 0.9914,
+      "step": 955
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9037553704636223,
+      "learning_rate": 9.974542431879507e-06,
+      "loss": 0.8434,
+      "step": 956
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.264333117353505,
+      "learning_rate": 9.974437865271152e-06,
+      "loss": 0.7964,
+      "step": 957
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8859278650675466,
+      "learning_rate": 9.97433308490012e-06,
+      "loss": 1.0552,
+      "step": 958
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.552373833838349,
+      "learning_rate": 9.97422809077092e-06,
+      "loss": 0.7868,
+      "step": 959
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7546222227107906,
+      "learning_rate": 9.97412288288806e-06,
+      "loss": 0.9772,
+      "step": 960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.099313649898104,
+      "learning_rate": 9.974017461256062e-06,
+      "loss": 0.8392,
+      "step": 961
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.480293291912968,
+      "learning_rate": 9.973911825879454e-06,
+      "loss": 0.9406,
+      "step": 962
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7254521752350827,
+      "learning_rate": 9.97380597676278e-06,
+      "loss": 0.9781,
+      "step": 963
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.298913315893285,
+      "learning_rate": 9.973699913910584e-06,
+      "loss": 0.9706,
+      "step": 964
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.198973015653919,
+      "learning_rate": 9.973593637327427e-06,
+      "loss": 0.7815,
+      "step": 965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2818452849120465,
+      "learning_rate": 9.973487147017874e-06,
+      "loss": 0.961,
+      "step": 966
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7567002214607568,
+      "learning_rate": 9.9733804429865e-06,
+      "loss": 1.0695,
+      "step": 967
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0295959378861372,
+      "learning_rate": 9.973273525237895e-06,
+      "loss": 1.1012,
+      "step": 968
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.894508806136415,
+      "learning_rate": 9.97316639377665e-06,
+      "loss": 0.9527,
+      "step": 969
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6697028951564317,
+      "learning_rate": 9.973059048607367e-06,
+      "loss": 0.8519,
+      "step": 970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.74571340678221,
+      "learning_rate": 9.972951489734664e-06,
+      "loss": 0.964,
+      "step": 971
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6273836032984774,
+      "learning_rate": 9.972843717163159e-06,
+      "loss": 0.948,
+      "step": 972
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4467186975810655,
+      "learning_rate": 9.972735730897484e-06,
+      "loss": 0.9112,
+      "step": 973
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1096916828170793,
+      "learning_rate": 9.97262753094228e-06,
+      "loss": 0.6395,
+      "step": 974
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.449422313807504,
+      "learning_rate": 9.972519117302195e-06,
+      "loss": 0.7533,
+      "step": 975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4387551386792223,
+      "learning_rate": 9.972410489981892e-06,
+      "loss": 0.9104,
+      "step": 976
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7474672390561614,
+      "learning_rate": 9.972301648986033e-06,
+      "loss": 0.9237,
+      "step": 977
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.48028757459085,
+      "learning_rate": 9.972192594319301e-06,
+      "loss": 0.9698,
+      "step": 978
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.739902569134069,
+      "learning_rate": 9.972083325986377e-06,
+      "loss": 1.0776,
+      "step": 979
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.791842017401302,
+      "learning_rate": 9.97197384399196e-06,
+      "loss": 0.8319,
+      "step": 980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.269175759917966,
+      "learning_rate": 9.971864148340755e-06,
+      "loss": 0.9946,
+      "step": 981
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.516908746316044,
+      "learning_rate": 9.971754239037474e-06,
+      "loss": 1.1539,
+      "step": 982
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.62611626800185,
+      "learning_rate": 9.971644116086841e-06,
+      "loss": 1.114,
+      "step": 983
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.83537691625511,
+      "learning_rate": 9.971533779493586e-06,
+      "loss": 0.9888,
+      "step": 984
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.777757813181085,
+      "learning_rate": 9.971423229262454e-06,
+      "loss": 1.0939,
+      "step": 985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2518442389329,
+      "learning_rate": 9.971312465398192e-06,
+      "loss": 0.7913,
+      "step": 986
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4210454360968887,
+      "learning_rate": 9.971201487905563e-06,
+      "loss": 0.7977,
+      "step": 987
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6796071019795025,
+      "learning_rate": 9.971090296789336e-06,
+      "loss": 0.829,
+      "step": 988
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.261499436219482,
+      "learning_rate": 9.970978892054286e-06,
+      "loss": 0.7931,
+      "step": 989
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.347354450784772,
+      "learning_rate": 9.970867273705203e-06,
+      "loss": 1.0475,
+      "step": 990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3012454523411408,
+      "learning_rate": 9.970755441746882e-06,
+      "loss": 0.6613,
+      "step": 991
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.540210030994667,
+      "learning_rate": 9.970643396184128e-06,
+      "loss": 0.9955,
+      "step": 992
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.036829655479102,
+      "learning_rate": 9.970531137021759e-06,
+      "loss": 0.8972,
+      "step": 993
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.1461601846671625,
+      "learning_rate": 9.970418664264596e-06,
+      "loss": 0.9412,
+      "step": 994
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3821027247323394,
+      "learning_rate": 9.970305977917473e-06,
+      "loss": 0.6566,
+      "step": 995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3065465242022194,
+      "learning_rate": 9.970193077985235e-06,
+      "loss": 0.7349,
+      "step": 996
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2923169732523823,
+      "learning_rate": 9.97007996447273e-06,
+      "loss": 0.8515,
+      "step": 997
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.180249777185171,
+      "learning_rate": 9.969966637384818e-06,
+      "loss": 1.1827,
+      "step": 998
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2514826559496453,
+      "learning_rate": 9.969853096726372e-06,
+      "loss": 0.6631,
+      "step": 999
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.998187914293974,
+      "learning_rate": 9.969739342502272e-06,
+      "loss": 0.947,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8866741813482324,
+      "learning_rate": 9.969625374717401e-06,
+      "loss": 0.877,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4743781087972432,
+      "learning_rate": 9.969511193376661e-06,
+      "loss": 0.6816,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3822149540406288,
+      "learning_rate": 9.969396798484958e-06,
+      "loss": 0.7827,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.41195975919726,
+      "learning_rate": 9.969282190047207e-06,
+      "loss": 1.1309,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1660673722899133,
+      "learning_rate": 9.969167368068334e-06,
+      "loss": 0.747,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9709285234251797,
+      "learning_rate": 9.96905233255327e-06,
+      "loss": 0.8363,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3981810700808093,
+      "learning_rate": 9.968937083506962e-06,
+      "loss": 0.7232,
+      "step": 1007
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.404134447811022,
+      "learning_rate": 9.96882162093436e-06,
+      "loss": 0.8073,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.175826147391627,
+      "learning_rate": 9.968705944840428e-06,
+      "loss": 0.6363,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1236977071775347,
+      "learning_rate": 9.968590055230135e-06,
+      "loss": 1.0206,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1239147102185783,
+      "learning_rate": 9.968473952108462e-06,
+      "loss": 1.1032,
+      "step": 1011
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.394913776727339,
+      "learning_rate": 9.968357635480397e-06,
+      "loss": 1.1036,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.383908567228465,
+      "learning_rate": 9.96824110535094e-06,
+      "loss": 0.921,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1585690417532422,
+      "learning_rate": 9.968124361725098e-06,
+      "loss": 0.6201,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.568906481531793,
+      "learning_rate": 9.968007404607887e-06,
+      "loss": 0.9518,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5965543829817814,
+      "learning_rate": 9.967890234004335e-06,
+      "loss": 0.994,
+      "step": 1016
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4900123882011824,
+      "learning_rate": 9.967772849919474e-06,
+      "loss": 0.8097,
+      "step": 1017
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.6198731961902375,
+      "learning_rate": 9.96765525235835e-06,
+      "loss": 0.9235,
+      "step": 1018
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5584037751347135,
+      "learning_rate": 9.967537441326018e-06,
+      "loss": 1.024,
+      "step": 1019
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2694040477004458,
+      "learning_rate": 9.967419416827537e-06,
+      "loss": 0.7008,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9451988582875082,
+      "learning_rate": 9.967301178867981e-06,
+      "loss": 0.7396,
+      "step": 1021
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.353628822552585,
+      "learning_rate": 9.96718272745243e-06,
+      "loss": 0.992,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9457985103096296,
+      "learning_rate": 9.967064062585977e-06,
+      "loss": 0.9564,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9322532766494347,
+      "learning_rate": 9.966945184273716e-06,
+      "loss": 0.8503,
+      "step": 1024
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4681780356592666,
+      "learning_rate": 9.96682609252076e-06,
+      "loss": 0.8414,
+      "step": 1025
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9049877390260264,
+      "learning_rate": 9.966706787332224e-06,
+      "loss": 0.883,
+      "step": 1026
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0426908608283507,
+      "learning_rate": 9.966587268713237e-06,
+      "loss": 1.0389,
+      "step": 1027
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.542913717255148,
+      "learning_rate": 9.966467536668933e-06,
+      "loss": 0.9023,
+      "step": 1028
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8057487183383465,
+      "learning_rate": 9.966347591204459e-06,
+      "loss": 0.8926,
+      "step": 1029
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2103666886664797,
+      "learning_rate": 9.96622743232497e-06,
+      "loss": 0.7263,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2550822622694724,
+      "learning_rate": 9.966107060035624e-06,
+      "loss": 0.7509,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8944739453025545,
+      "learning_rate": 9.9659864743416e-06,
+      "loss": 0.967,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9661684081669657,
+      "learning_rate": 9.965865675248077e-06,
+      "loss": 0.9793,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.62023890174226,
+      "learning_rate": 9.965744662760246e-06,
+      "loss": 0.9158,
+      "step": 1034
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4122466383105436,
+      "learning_rate": 9.965623436883307e-06,
+      "loss": 1.0805,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1861951582838355,
+      "learning_rate": 9.96550199762247e-06,
+      "loss": 1.1078,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1560353135950059,
+      "learning_rate": 9.965380344982955e-06,
+      "loss": 0.6925,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1907119954295955,
+      "learning_rate": 9.965258478969985e-06,
+      "loss": 0.9828,
+      "step": 1038
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.658271364410937,
+      "learning_rate": 9.965136399588803e-06,
+      "loss": 1.1138,
+      "step": 1039
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.161147027556223,
+      "learning_rate": 9.96501410684465e-06,
+      "loss": 0.9647,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0563586349831615,
+      "learning_rate": 9.964891600742785e-06,
+      "loss": 0.9395,
+      "step": 1041
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1195939809523145,
+      "learning_rate": 9.96476888128847e-06,
+      "loss": 0.8651,
+      "step": 1042
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.978677068210579,
+      "learning_rate": 9.964645948486978e-06,
+      "loss": 0.7831,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1901793616202636,
+      "learning_rate": 9.964522802343593e-06,
+      "loss": 0.6896,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.533523275790489,
+      "learning_rate": 9.964399442863609e-06,
+      "loss": 0.7096,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3705911746146886,
+      "learning_rate": 9.96427587005232e-06,
+      "loss": 0.786,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.028343141782053,
+      "learning_rate": 9.964152083915045e-06,
+      "loss": 0.6924,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1649822538113683,
+      "learning_rate": 9.964028084457098e-06,
+      "loss": 0.671,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3880167254991218,
+      "learning_rate": 9.963903871683806e-06,
+      "loss": 0.7791,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2478846334029043,
+      "learning_rate": 9.963779445600512e-06,
+      "loss": 0.9889,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3386608012868564,
+      "learning_rate": 9.963654806212561e-06,
+      "loss": 0.9864,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.193933529308126,
+      "learning_rate": 9.963529953525307e-06,
+      "loss": 0.6734,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.1319216770694265,
+      "learning_rate": 9.963404887544117e-06,
+      "loss": 0.9711,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.752129240161239,
+      "learning_rate": 9.963279608274364e-06,
+      "loss": 0.8812,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.094022113394416,
+      "learning_rate": 9.963154115721433e-06,
+      "loss": 0.903,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0839940194893787,
+      "learning_rate": 9.963028409890716e-06,
+      "loss": 0.8529,
+      "step": 1056
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0047113861822563,
+      "learning_rate": 9.962902490787616e-06,
+      "loss": 1.013,
+      "step": 1057
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7133436699856968,
+      "learning_rate": 9.962776358417543e-06,
+      "loss": 0.8574,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1078941821647599,
+      "learning_rate": 9.962650012785917e-06,
+      "loss": 0.5764,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.352386863758474,
+      "learning_rate": 9.962523453898166e-06,
+      "loss": 0.9362,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.940315047097218,
+      "learning_rate": 9.96239668175973e-06,
+      "loss": 0.9948,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8725052918203002,
+      "learning_rate": 9.962269696376059e-06,
+      "loss": 0.9801,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4543299354358643,
+      "learning_rate": 9.962142497752605e-06,
+      "loss": 1.1599,
+      "step": 1063
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.344377847940124,
+      "learning_rate": 9.962015085894838e-06,
+      "loss": 0.9488,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.471857175310381,
+      "learning_rate": 9.961887460808232e-06,
+      "loss": 0.8675,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.516656389896832,
+      "learning_rate": 9.96175962249827e-06,
+      "loss": 1.0944,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3715058988280302,
+      "learning_rate": 9.961631570970449e-06,
+      "loss": 0.8192,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2526331193690035,
+      "learning_rate": 9.961503306230267e-06,
+      "loss": 1.0462,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2297121384236185,
+      "learning_rate": 9.961374828283239e-06,
+      "loss": 0.8595,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.22051096559212,
+      "learning_rate": 9.961246137134883e-06,
+      "loss": 1.1246,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5177238620977485,
+      "learning_rate": 9.961117232790734e-06,
+      "loss": 0.9034,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.515502917352967,
+      "learning_rate": 9.960988115256328e-06,
+      "loss": 0.8657,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3925909144300666,
+      "learning_rate": 9.960858784537213e-06,
+      "loss": 0.8995,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9325760487843127,
+      "learning_rate": 9.960729240638947e-06,
+      "loss": 1.0535,
+      "step": 1074
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.7521025817539355,
+      "learning_rate": 9.9605994835671e-06,
+      "loss": 1.0402,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.415627825555859,
+      "learning_rate": 9.960469513327242e-06,
+      "loss": 1.148,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.601380767466556,
+      "learning_rate": 9.960339329924964e-06,
+      "loss": 0.9408,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7642867680592205,
+      "learning_rate": 9.960208933365857e-06,
+      "loss": 0.8075,
+      "step": 1078
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1186081369907237,
+      "learning_rate": 9.960078323655524e-06,
+      "loss": 1.0673,
+      "step": 1079
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.368869196592482,
+      "learning_rate": 9.95994750079958e-06,
+      "loss": 1.0479,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.923096682280585,
+      "learning_rate": 9.959816464803647e-06,
+      "loss": 0.9726,
+      "step": 1081
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.035864387470966,
+      "learning_rate": 9.959685215673351e-06,
+      "loss": 0.8923,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2633898607428824,
+      "learning_rate": 9.959553753414338e-06,
+      "loss": 0.7352,
+      "step": 1083
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.11442216911315,
+      "learning_rate": 9.959422078032253e-06,
+      "loss": 1.0797,
+      "step": 1084
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0919983675641753,
+      "learning_rate": 9.959290189532757e-06,
+      "loss": 0.9741,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.277911224472534,
+      "learning_rate": 9.959158087921516e-06,
+      "loss": 0.9232,
+      "step": 1086
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.159182819506481,
+      "learning_rate": 9.95902577320421e-06,
+      "loss": 0.9935,
+      "step": 1087
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.553471048568619,
+      "learning_rate": 9.95889324538652e-06,
+      "loss": 0.7931,
+      "step": 1088
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.389737008175372,
+      "learning_rate": 9.958760504474144e-06,
+      "loss": 0.7397,
+      "step": 1089
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.061516652458585,
+      "learning_rate": 9.958627550472784e-06,
+      "loss": 0.8635,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3552301764361505,
+      "learning_rate": 9.958494383388155e-06,
+      "loss": 0.6923,
+      "step": 1091
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.79060474727819,
+      "learning_rate": 9.958361003225979e-06,
+      "loss": 0.9777,
+      "step": 1092
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.896111701481863,
+      "learning_rate": 9.958227409991988e-06,
+      "loss": 0.8463,
+      "step": 1093
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.310848330181726,
+      "learning_rate": 9.958093603691923e-06,
+      "loss": 1.0338,
+      "step": 1094
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.117286940329302,
+      "learning_rate": 9.957959584331533e-06,
+      "loss": 0.9749,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.365023772044114,
+      "learning_rate": 9.957825351916579e-06,
+      "loss": 1.0155,
+      "step": 1096
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.164265458130438,
+      "learning_rate": 9.957690906452827e-06,
+      "loss": 1.0949,
+      "step": 1097
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4406576812062455,
+      "learning_rate": 9.957556247946057e-06,
+      "loss": 0.9164,
+      "step": 1098
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0790738091274386,
+      "learning_rate": 9.957421376402053e-06,
+      "loss": 0.6969,
+      "step": 1099
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.709836308167979,
+      "learning_rate": 9.95728629182661e-06,
+      "loss": 0.8147,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2215308712608883,
+      "learning_rate": 9.957150994225535e-06,
+      "loss": 0.8515,
+      "step": 1101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9560152372240875,
+      "learning_rate": 9.957015483604643e-06,
+      "loss": 0.8365,
+      "step": 1102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.677478481879237,
+      "learning_rate": 9.956879759969755e-06,
+      "loss": 0.961,
+      "step": 1103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.542651933219665,
+      "learning_rate": 9.956743823326704e-06,
+      "loss": 0.9264,
+      "step": 1104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3203655836796355,
+      "learning_rate": 9.956607673681334e-06,
+      "loss": 1.0524,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6329719794524897,
+      "learning_rate": 9.956471311039491e-06,
+      "loss": 1.0148,
+      "step": 1106
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5593846395675386,
+      "learning_rate": 9.956334735407037e-06,
+      "loss": 0.8492,
+      "step": 1107
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1934591858824244,
+      "learning_rate": 9.95619794678984e-06,
+      "loss": 0.6996,
+      "step": 1108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.71269598362435,
+      "learning_rate": 9.956060945193781e-06,
+      "loss": 0.7523,
+      "step": 1109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9023895169761227,
+      "learning_rate": 9.955923730624745e-06,
+      "loss": 0.8425,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.162824480453968,
+      "learning_rate": 9.95578630308863e-06,
+      "loss": 0.9439,
+      "step": 1111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0548090736855307,
+      "learning_rate": 9.95564866259134e-06,
+      "loss": 0.7882,
+      "step": 1112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2942479584339552,
+      "learning_rate": 9.95551080913879e-06,
+      "loss": 0.8578,
+      "step": 1113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8158254659560864,
+      "learning_rate": 9.955372742736903e-06,
+      "loss": 1.1024,
+      "step": 1114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.677641121917483,
+      "learning_rate": 9.955234463391612e-06,
+      "loss": 0.8889,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5651850618286973,
+      "learning_rate": 9.955095971108863e-06,
+      "loss": 0.7015,
+      "step": 1116
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6720857486102796,
+      "learning_rate": 9.954957265894602e-06,
+      "loss": 0.9696,
+      "step": 1117
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6350248407270156,
+      "learning_rate": 9.954818347754795e-06,
+      "loss": 0.8704,
+      "step": 1118
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3971874323538547,
+      "learning_rate": 9.954679216695406e-06,
+      "loss": 0.8821,
+      "step": 1119
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4495023422138482,
+      "learning_rate": 9.954539872722417e-06,
+      "loss": 0.6014,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7270604230650544,
+      "learning_rate": 9.954400315841817e-06,
+      "loss": 0.8861,
+      "step": 1121
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.14639074918527,
+      "learning_rate": 9.9542605460596e-06,
+      "loss": 0.9107,
+      "step": 1122
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3703521479210132,
+      "learning_rate": 9.954120563381772e-06,
+      "loss": 0.9355,
+      "step": 1123
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.865264103903431,
+      "learning_rate": 9.953980367814354e-06,
+      "loss": 0.961,
+      "step": 1124
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3595613771972714,
+      "learning_rate": 9.953839959363363e-06,
+      "loss": 0.8636,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1762151768578994,
+      "learning_rate": 9.953699338034838e-06,
+      "loss": 0.8211,
+      "step": 1126
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4205452322526164,
+      "learning_rate": 9.953558503834819e-06,
+      "loss": 1.0613,
+      "step": 1127
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4647013171389003,
+      "learning_rate": 9.95341745676936e-06,
+      "loss": 1.0412,
+      "step": 1128
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.745051186111235,
+      "learning_rate": 9.953276196844519e-06,
+      "loss": 1.0761,
+      "step": 1129
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4754385151795737,
+      "learning_rate": 9.95313472406637e-06,
+      "loss": 0.7979,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2525935723352264,
+      "learning_rate": 9.95299303844099e-06,
+      "loss": 0.6868,
+      "step": 1131
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9975329297507405,
+      "learning_rate": 9.952851139974468e-06,
+      "loss": 0.9676,
+      "step": 1132
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.338187440284261,
+      "learning_rate": 9.952709028672902e-06,
+      "loss": 0.8448,
+      "step": 1133
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4330161503286734,
+      "learning_rate": 9.9525667045424e-06,
+      "loss": 0.8171,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4508862624885754,
+      "learning_rate": 9.952424167589076e-06,
+      "loss": 1.0568,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.465770325750154,
+      "learning_rate": 9.952281417819057e-06,
+      "loss": 0.9468,
+      "step": 1136
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.187000740421444,
+      "learning_rate": 9.952138455238474e-06,
+      "loss": 0.7223,
+      "step": 1137
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4511444986967366,
+      "learning_rate": 9.951995279853472e-06,
+      "loss": 0.8073,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.425569542398238,
+      "learning_rate": 9.951851891670206e-06,
+      "loss": 0.9443,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.164643902155613,
+      "learning_rate": 9.951708290694837e-06,
+      "loss": 0.7383,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2285843612461265,
+      "learning_rate": 9.951564476933534e-06,
+      "loss": 0.9225,
+      "step": 1141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.73375172387479,
+      "learning_rate": 9.951420450392476e-06,
+      "loss": 1.0216,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3204614456134596,
+      "learning_rate": 9.951276211077854e-06,
+      "loss": 0.6794,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.27818525722311,
+      "learning_rate": 9.951131758995866e-06,
+      "loss": 1.2494,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4261690308145987,
+      "learning_rate": 9.950987094152721e-06,
+      "loss": 0.9912,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.59742570076728,
+      "learning_rate": 9.950842216554633e-06,
+      "loss": 0.7504,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0245058019210225,
+      "learning_rate": 9.95069712620783e-06,
+      "loss": 0.9943,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.205733068467976,
+      "learning_rate": 9.950551823118544e-06,
+      "loss": 1.1078,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.142117299881347,
+      "learning_rate": 9.950406307293023e-06,
+      "loss": 0.7077,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4049079509825133,
+      "learning_rate": 9.950260578737517e-06,
+      "loss": 0.8969,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.468518546671503,
+      "learning_rate": 9.950114637458288e-06,
+      "loss": 0.796,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0963597375678877,
+      "learning_rate": 9.949968483461609e-06,
+      "loss": 1.0323,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2722464244795546,
+      "learning_rate": 9.94982211675376e-06,
+      "loss": 0.9608,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4197104439466055,
+      "learning_rate": 9.949675537341031e-06,
+      "loss": 0.7941,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.573271424551844,
+      "learning_rate": 9.949528745229721e-06,
+      "loss": 0.8542,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.161972689071315,
+      "learning_rate": 9.949381740426139e-06,
+      "loss": 0.8105,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.169282453352002,
+      "learning_rate": 9.949234522936597e-06,
+      "loss": 0.9653,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1480503702300722,
+      "learning_rate": 9.949087092767428e-06,
+      "loss": 0.7332,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.327129258212959,
+      "learning_rate": 9.948939449924964e-06,
+      "loss": 0.9774,
+      "step": 1159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5103206563173917,
+      "learning_rate": 9.94879159441555e-06,
+      "loss": 0.864,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9010823409582134,
+      "learning_rate": 9.948643526245538e-06,
+      "loss": 0.7716,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.704371987633381,
+      "learning_rate": 9.948495245421294e-06,
+      "loss": 0.9139,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8125319022854884,
+      "learning_rate": 9.948346751949188e-06,
+      "loss": 1.0941,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2601987103551227,
+      "learning_rate": 9.948198045835601e-06,
+      "loss": 0.6852,
+      "step": 1164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0587641351621326,
+      "learning_rate": 9.948049127086925e-06,
+      "loss": 0.9148,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2315881150773123,
+      "learning_rate": 9.947899995709556e-06,
+      "loss": 0.7132,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8823030021206466,
+      "learning_rate": 9.947750651709907e-06,
+      "loss": 0.8413,
+      "step": 1167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.169198949706182,
+      "learning_rate": 9.947601095094392e-06,
+      "loss": 0.848,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1476772687126189,
+      "learning_rate": 9.94745132586944e-06,
+      "loss": 0.5988,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6851455590618674,
+      "learning_rate": 9.947301344041485e-06,
+      "loss": 0.8572,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1735114671956532,
+      "learning_rate": 9.947151149616972e-06,
+      "loss": 0.9028,
+      "step": 1171
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6249575971904373,
+      "learning_rate": 9.947000742602358e-06,
+      "loss": 0.9165,
+      "step": 1172
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2931005242112903,
+      "learning_rate": 9.946850123004103e-06,
+      "loss": 0.7679,
+      "step": 1173
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3440958299568546,
+      "learning_rate": 9.946699290828683e-06,
+      "loss": 0.9971,
+      "step": 1174
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.310106793084202,
+      "learning_rate": 9.946548246082577e-06,
+      "loss": 0.9907,
+      "step": 1175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.17818117361062,
+      "learning_rate": 9.946396988772275e-06,
+      "loss": 0.8402,
+      "step": 1176
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6352087430988362,
+      "learning_rate": 9.94624551890428e-06,
+      "loss": 0.924,
+      "step": 1177
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3586406130915436,
+      "learning_rate": 9.946093836485097e-06,
+      "loss": 1.204,
+      "step": 1178
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0746180257270923,
+      "learning_rate": 9.94594194152125e-06,
+      "loss": 0.6498,
+      "step": 1179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2095346418729167,
+      "learning_rate": 9.945789834019259e-06,
+      "loss": 0.6539,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.92805857749779,
+      "learning_rate": 9.945637513985664e-06,
+      "loss": 0.8507,
+      "step": 1181
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9620266738701808,
+      "learning_rate": 9.945484981427011e-06,
+      "loss": 0.9217,
+      "step": 1182
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4664496787569195,
+      "learning_rate": 9.945332236349857e-06,
+      "loss": 0.9835,
+      "step": 1183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2517867122679884,
+      "learning_rate": 9.945179278760759e-06,
+      "loss": 0.6795,
+      "step": 1184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7256523093106966,
+      "learning_rate": 9.945026108666296e-06,
+      "loss": 0.8436,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7774876050104167,
+      "learning_rate": 9.944872726073048e-06,
+      "loss": 0.9037,
+      "step": 1186
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.186436358873795,
+      "learning_rate": 9.944719130987606e-06,
+      "loss": 0.8137,
+      "step": 1187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.879100064345231,
+      "learning_rate": 9.94456532341657e-06,
+      "loss": 0.8885,
+      "step": 1188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6902202758590987,
+      "learning_rate": 9.94441130336655e-06,
+      "loss": 1.0003,
+      "step": 1189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.609976945511062,
+      "learning_rate": 9.944257070844165e-06,
+      "loss": 0.9494,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.372846446752374,
+      "learning_rate": 9.944102625856042e-06,
+      "loss": 1.0804,
+      "step": 1191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3914208973745446,
+      "learning_rate": 9.943947968408816e-06,
+      "loss": 0.96,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.764282233105401,
+      "learning_rate": 9.943793098509138e-06,
+      "loss": 1.1658,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0815522060782166,
+      "learning_rate": 9.943638016163658e-06,
+      "loss": 0.8518,
+      "step": 1194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5906668505927035,
+      "learning_rate": 9.943482721379045e-06,
+      "loss": 0.9238,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.567098976230415,
+      "learning_rate": 9.943327214161968e-06,
+      "loss": 0.9559,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.335961564489395,
+      "learning_rate": 9.943171494519111e-06,
+      "loss": 0.9475,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.139095962587013,
+      "learning_rate": 9.943015562457166e-06,
+      "loss": 0.8308,
+      "step": 1198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9156069038130146,
+      "learning_rate": 9.942859417982833e-06,
+      "loss": 0.7671,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1298034837800415,
+      "learning_rate": 9.942703061102823e-06,
+      "loss": 0.6591,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7225691115157,
+      "learning_rate": 9.942546491823856e-06,
+      "loss": 1.1145,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.448538573406191,
+      "learning_rate": 9.942389710152657e-06,
+      "loss": 0.9995,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2231437803306038,
+      "learning_rate": 9.942232716095965e-06,
+      "loss": 0.8135,
+      "step": 1203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.129897004370326,
+      "learning_rate": 9.942075509660527e-06,
+      "loss": 0.9234,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.945015480751331,
+      "learning_rate": 9.941918090853096e-06,
+      "loss": 0.9849,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.190631519465854,
+      "learning_rate": 9.941760459680441e-06,
+      "loss": 0.922,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9326600265772482,
+      "learning_rate": 9.94160261614933e-06,
+      "loss": 0.9246,
+      "step": 1207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.971693405569152,
+      "learning_rate": 9.941444560266551e-06,
+      "loss": 0.8064,
+      "step": 1208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1842321669698816,
+      "learning_rate": 9.941286292038894e-06,
+      "loss": 0.715,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5423864567355223,
+      "learning_rate": 9.941127811473162e-06,
+      "loss": 0.7148,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1445312884076304,
+      "learning_rate": 9.94096911857616e-06,
+      "loss": 0.7325,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0695060391921287,
+      "learning_rate": 9.940810213354713e-06,
+      "loss": 0.9605,
+      "step": 1212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7040419588430984,
+      "learning_rate": 9.940651095815646e-06,
+      "loss": 0.9784,
+      "step": 1213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1921711328169007,
+      "learning_rate": 9.940491765965798e-06,
+      "loss": 1.0577,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.234615095184679,
+      "learning_rate": 9.940332223812017e-06,
+      "loss": 0.9338,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.864634991977229,
+      "learning_rate": 9.940172469361158e-06,
+      "loss": 0.9419,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2400770689523182,
+      "learning_rate": 9.940012502620085e-06,
+      "loss": 0.6084,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0204066690669857,
+      "learning_rate": 9.939852323595671e-06,
+      "loss": 0.7585,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0423186164197153,
+      "learning_rate": 9.939691932294804e-06,
+      "loss": 0.6064,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.176788352510787,
+      "learning_rate": 9.939531328724371e-06,
+      "loss": 0.9045,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4136827960862237,
+      "learning_rate": 9.939370512891277e-06,
+      "loss": 0.7509,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1153078075909626,
+      "learning_rate": 9.93920948480243e-06,
+      "loss": 0.7295,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.66435950734608,
+      "learning_rate": 9.939048244464753e-06,
+      "loss": 1.0263,
+      "step": 1223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.138399278451951,
+      "learning_rate": 9.938886791885172e-06,
+      "loss": 0.947,
+      "step": 1224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5144693435507244,
+      "learning_rate": 9.938725127070628e-06,
+      "loss": 0.9598,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3067702490057527,
+      "learning_rate": 9.938563250028063e-06,
+      "loss": 0.7414,
+      "step": 1226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4092322516194424,
+      "learning_rate": 9.93840116076444e-06,
+      "loss": 0.9855,
+      "step": 1227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.031243727370259,
+      "learning_rate": 9.938238859286719e-06,
+      "loss": 0.9067,
+      "step": 1228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.065054344461064,
+      "learning_rate": 9.938076345601875e-06,
+      "loss": 1.0753,
+      "step": 1229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6400522024705224,
+      "learning_rate": 9.937913619716895e-06,
+      "loss": 1.0776,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1584098322864804,
+      "learning_rate": 9.937750681638766e-06,
+      "loss": 0.9515,
+      "step": 1231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2756714066688684,
+      "learning_rate": 9.937587531374497e-06,
+      "loss": 0.998,
+      "step": 1232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.494868073912578,
+      "learning_rate": 9.937424168931094e-06,
+      "loss": 1.0943,
+      "step": 1233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.642328674587481,
+      "learning_rate": 9.937260594315578e-06,
+      "loss": 1.0648,
+      "step": 1234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7957843914348564,
+      "learning_rate": 9.937096807534978e-06,
+      "loss": 1.005,
+      "step": 1235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4002038836172073,
+      "learning_rate": 9.936932808596332e-06,
+      "loss": 0.6553,
+      "step": 1236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.137154485861161,
+      "learning_rate": 9.936768597506689e-06,
+      "loss": 0.6757,
+      "step": 1237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.940567479948612,
+      "learning_rate": 9.936604174273105e-06,
+      "loss": 0.8956,
+      "step": 1238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1228083466075915,
+      "learning_rate": 9.936439538902644e-06,
+      "loss": 1.0579,
+      "step": 1239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.95621076328198,
+      "learning_rate": 9.936274691402384e-06,
+      "loss": 0.9317,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2199292760700473,
+      "learning_rate": 9.936109631779404e-06,
+      "loss": 0.8915,
+      "step": 1241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7706804811059937,
+      "learning_rate": 9.935944360040803e-06,
+      "loss": 1.0145,
+      "step": 1242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7251664784315897,
+      "learning_rate": 9.935778876193678e-06,
+      "loss": 1.0277,
+      "step": 1243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0319013111357416,
+      "learning_rate": 9.935613180245143e-06,
+      "loss": 0.7279,
+      "step": 1244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.63440047275755,
+      "learning_rate": 9.935447272202317e-06,
+      "loss": 0.9556,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0358510916340014,
+      "learning_rate": 9.935281152072329e-06,
+      "loss": 1.0002,
+      "step": 1246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.270721542078174,
+      "learning_rate": 9.93511481986232e-06,
+      "loss": 0.8784,
+      "step": 1247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2607710325489183,
+      "learning_rate": 9.934948275579435e-06,
+      "loss": 0.7614,
+      "step": 1248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.284457554707012,
+      "learning_rate": 9.934781519230832e-06,
+      "loss": 0.6795,
+      "step": 1249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8871305850931694,
+      "learning_rate": 9.934614550823679e-06,
+      "loss": 0.9674,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7211955018657665,
+      "learning_rate": 9.934447370365146e-06,
+      "loss": 0.9522,
+      "step": 1251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7877290640413013,
+      "learning_rate": 9.93427997786242e-06,
+      "loss": 1.0439,
+      "step": 1252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6890045682380657,
+      "learning_rate": 9.934112373322695e-06,
+      "loss": 0.8216,
+      "step": 1253
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.205689198372902,
+      "learning_rate": 9.933944556753173e-06,
+      "loss": 1.0605,
+      "step": 1254
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.675035967944229,
+      "learning_rate": 9.933776528161063e-06,
+      "loss": 0.9302,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1006165826694898,
+      "learning_rate": 9.933608287553589e-06,
+      "loss": 0.6148,
+      "step": 1256
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3443950507843367,
+      "learning_rate": 9.93343983493798e-06,
+      "loss": 1.0094,
+      "step": 1257
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.267368120900792,
+      "learning_rate": 9.93327117032147e-06,
+      "loss": 0.9229,
+      "step": 1258
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5576913767864093,
+      "learning_rate": 9.933102293711314e-06,
+      "loss": 1.0975,
+      "step": 1259
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3251886486885387,
+      "learning_rate": 9.932933205114766e-06,
+      "loss": 0.9783,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3386660971144844,
+      "learning_rate": 9.932763904539091e-06,
+      "loss": 1.0656,
+      "step": 1261
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.426737027899981,
+      "learning_rate": 9.932594391991566e-06,
+      "loss": 0.8993,
+      "step": 1262
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.83942683220251,
+      "learning_rate": 9.932424667479475e-06,
+      "loss": 0.8042,
+      "step": 1263
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.2239661078761035,
+      "learning_rate": 9.932254731010108e-06,
+      "loss": 0.9249,
+      "step": 1264
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1633803448795264,
+      "learning_rate": 9.932084582590773e-06,
+      "loss": 0.8334,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2117309717906273,
+      "learning_rate": 9.93191422222878e-06,
+      "loss": 1.0803,
+      "step": 1266
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2438065388356938,
+      "learning_rate": 9.931743649931446e-06,
+      "loss": 0.6578,
+      "step": 1267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.079623686964702,
+      "learning_rate": 9.931572865706105e-06,
+      "loss": 1.1211,
+      "step": 1268
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.25604593389086,
+      "learning_rate": 9.931401869560096e-06,
+      "loss": 0.9957,
+      "step": 1269
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.289145451828619,
+      "learning_rate": 9.931230661500764e-06,
+      "loss": 0.9088,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1502218585790613,
+      "learning_rate": 9.93105924153547e-06,
+      "loss": 0.9055,
+      "step": 1271
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1655146583707412,
+      "learning_rate": 9.930887609671577e-06,
+      "loss": 0.6542,
+      "step": 1272
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.59029679118931,
+      "learning_rate": 9.930715765916462e-06,
+      "loss": 1.0573,
+      "step": 1273
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0336068996204815,
+      "learning_rate": 9.93054371027751e-06,
+      "loss": 0.7818,
+      "step": 1274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.480689324391336,
+      "learning_rate": 9.930371442762112e-06,
+      "loss": 0.8434,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3942096750859214,
+      "learning_rate": 9.930198963377672e-06,
+      "loss": 0.8636,
+      "step": 1276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0135084386007036,
+      "learning_rate": 9.930026272131605e-06,
+      "loss": 0.8721,
+      "step": 1277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.10457051038635,
+      "learning_rate": 9.929853369031328e-06,
+      "loss": 1.0015,
+      "step": 1278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.556208420087465,
+      "learning_rate": 9.929680254084273e-06,
+      "loss": 0.7854,
+      "step": 1279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.103620748152359,
+      "learning_rate": 9.929506927297878e-06,
+      "loss": 0.8148,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0646113826765244,
+      "learning_rate": 9.929333388679593e-06,
+      "loss": 0.8115,
+      "step": 1281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1477744229091655,
+      "learning_rate": 9.929159638236872e-06,
+      "loss": 0.9338,
+      "step": 1282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.047868223129076,
+      "learning_rate": 9.928985675977185e-06,
+      "loss": 0.8418,
+      "step": 1283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.907321547391761,
+      "learning_rate": 9.928811501908006e-06,
+      "loss": 1.0318,
+      "step": 1284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.554473832340601,
+      "learning_rate": 9.928637116036818e-06,
+      "loss": 0.8176,
+      "step": 1285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.334684678235707,
+      "learning_rate": 9.928462518371119e-06,
+      "loss": 0.9406,
+      "step": 1286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.544788014494609,
+      "learning_rate": 9.928287708918408e-06,
+      "loss": 0.9531,
+      "step": 1287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1926963185433157,
+      "learning_rate": 9.928112687686197e-06,
+      "loss": 0.8111,
+      "step": 1288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0070165411134813,
+      "learning_rate": 9.92793745468201e-06,
+      "loss": 1.0317,
+      "step": 1289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.395534611842615,
+      "learning_rate": 9.927762009913377e-06,
+      "loss": 1.0948,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2680084217346037,
+      "learning_rate": 9.927586353387832e-06,
+      "loss": 0.8961,
+      "step": 1291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1308511431453985,
+      "learning_rate": 9.92741048511293e-06,
+      "loss": 0.8893,
+      "step": 1292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.446781075247447,
+      "learning_rate": 9.927234405096224e-06,
+      "loss": 0.8533,
+      "step": 1293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.469410754478872,
+      "learning_rate": 9.927058113345282e-06,
+      "loss": 1.1344,
+      "step": 1294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.455233523096212,
+      "learning_rate": 9.92688160986768e-06,
+      "loss": 0.6768,
+      "step": 1295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5450891569359206,
+      "learning_rate": 9.926704894671004e-06,
+      "loss": 0.8825,
+      "step": 1296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.008617981613745,
+      "learning_rate": 9.926527967762845e-06,
+      "loss": 0.9005,
+      "step": 1297
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3233386296203318,
+      "learning_rate": 9.926350829150806e-06,
+      "loss": 0.725,
+      "step": 1298
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5387960974568413,
+      "learning_rate": 9.926173478842502e-06,
+      "loss": 0.9392,
+      "step": 1299
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.058736216116884,
+      "learning_rate": 9.925995916845553e-06,
+      "loss": 0.9947,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3724103198121982,
+      "learning_rate": 9.925818143167587e-06,
+      "loss": 0.6201,
+      "step": 1301
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1003363443656613,
+      "learning_rate": 9.925640157816246e-06,
+      "loss": 0.9133,
+      "step": 1302
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.286490456390942,
+      "learning_rate": 9.925461960799177e-06,
+      "loss": 0.9267,
+      "step": 1303
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6141244259776606,
+      "learning_rate": 9.925283552124039e-06,
+      "loss": 1.0877,
+      "step": 1304
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8668171495419204,
+      "learning_rate": 9.925104931798496e-06,
+      "loss": 1.0181,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8860176329175942,
+      "learning_rate": 9.924926099830226e-06,
+      "loss": 0.8739,
+      "step": 1306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.379330326147918,
+      "learning_rate": 9.924747056226913e-06,
+      "loss": 0.9246,
+      "step": 1307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0756126589514463,
+      "learning_rate": 9.92456780099625e-06,
+      "loss": 0.9216,
+      "step": 1308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4248527590689464,
+      "learning_rate": 9.924388334145943e-06,
+      "loss": 0.9304,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9060108337843062,
+      "learning_rate": 9.9242086556837e-06,
+      "loss": 0.8323,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3639620963747445,
+      "learning_rate": 9.924028765617247e-06,
+      "loss": 1.0815,
+      "step": 1311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.054962285961922,
+      "learning_rate": 9.92384866395431e-06,
+      "loss": 0.8341,
+      "step": 1312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.405852739271322,
+      "learning_rate": 9.923668350702628e-06,
+      "loss": 0.9332,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7079955592736615,
+      "learning_rate": 9.923487825869955e-06,
+      "loss": 1.1538,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.863993520502624,
+      "learning_rate": 9.923307089464043e-06,
+      "loss": 0.9323,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3413363428110738,
+      "learning_rate": 9.92312614149266e-06,
+      "loss": 0.7943,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9995868226479008,
+      "learning_rate": 9.922944981963583e-06,
+      "loss": 0.9357,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.036937143848583,
+      "learning_rate": 9.922763610884596e-06,
+      "loss": 0.7131,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.710864496460764,
+      "learning_rate": 9.922582028263495e-06,
+      "loss": 0.9195,
+      "step": 1319
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3406312448303077,
+      "learning_rate": 9.922400234108079e-06,
+      "loss": 0.9177,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0642824565525806,
+      "learning_rate": 9.922218228426162e-06,
+      "loss": 1.0345,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2694676139520584,
+      "learning_rate": 9.922036011225564e-06,
+      "loss": 0.691,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.053704294060293,
+      "learning_rate": 9.92185358251412e-06,
+      "loss": 0.9348,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.162336047885978,
+      "learning_rate": 9.921670942299664e-06,
+      "loss": 0.7436,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1460576203013881,
+      "learning_rate": 9.921488090590047e-06,
+      "loss": 0.6542,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1661512694291125,
+      "learning_rate": 9.921305027393125e-06,
+      "loss": 1.022,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6808578514758574,
+      "learning_rate": 9.921121752716767e-06,
+      "loss": 0.9739,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7152555500979108,
+      "learning_rate": 9.920938266568847e-06,
+      "loss": 0.8114,
+      "step": 1328
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1595544120052563,
+      "learning_rate": 9.92075456895725e-06,
+      "loss": 0.8214,
+      "step": 1329
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4553829715880864,
+      "learning_rate": 9.92057065988987e-06,
+      "loss": 0.8429,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2218729353803488,
+      "learning_rate": 9.92038653937461e-06,
+      "loss": 0.6407,
+      "step": 1331
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.008156510544126,
+      "learning_rate": 9.920202207419382e-06,
+      "loss": 0.6686,
+      "step": 1332
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.289238307536233,
+      "learning_rate": 9.920017664032109e-06,
+      "loss": 0.9271,
+      "step": 1333
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3455503721012074,
+      "learning_rate": 9.919832909220717e-06,
+      "loss": 0.9702,
+      "step": 1334
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0349256458998526,
+      "learning_rate": 9.91964794299315e-06,
+      "loss": 0.8837,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1533532899389307,
+      "learning_rate": 9.919462765357353e-06,
+      "loss": 0.8874,
+      "step": 1336
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2689936788478315,
+      "learning_rate": 9.919277376321284e-06,
+      "loss": 0.6785,
+      "step": 1337
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.166800842748292,
+      "learning_rate": 9.91909177589291e-06,
+      "loss": 0.6808,
+      "step": 1338
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0235642667406877,
+      "learning_rate": 9.91890596408021e-06,
+      "loss": 1.0093,
+      "step": 1339
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2568688058602,
+      "learning_rate": 9.918719940891165e-06,
+      "loss": 0.8896,
+      "step": 1340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3062628240133507,
+      "learning_rate": 9.918533706333767e-06,
+      "loss": 0.9502,
+      "step": 1341
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7367228694916466,
+      "learning_rate": 9.918347260416024e-06,
+      "loss": 0.8933,
+      "step": 1342
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8512857023626696,
+      "learning_rate": 9.918160603145943e-06,
+      "loss": 1.0806,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.301878219002174,
+      "learning_rate": 9.917973734531549e-06,
+      "loss": 0.8016,
+      "step": 1344
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.527034636668726,
+      "learning_rate": 9.917786654580872e-06,
+      "loss": 0.9017,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2152869095802266,
+      "learning_rate": 9.917599363301947e-06,
+      "loss": 0.8525,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2930396034137093,
+      "learning_rate": 9.917411860702828e-06,
+      "loss": 0.7057,
+      "step": 1347
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2094886462010517,
+      "learning_rate": 9.917224146791568e-06,
+      "loss": 0.9641,
+      "step": 1348
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8535396679448033,
+      "learning_rate": 9.917036221576235e-06,
+      "loss": 0.8514,
+      "step": 1349
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3585933344302097,
+      "learning_rate": 9.916848085064906e-06,
+      "loss": 0.6776,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5249406415313134,
+      "learning_rate": 9.916659737265664e-06,
+      "loss": 1.202,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.387610684103858,
+      "learning_rate": 9.916471178186602e-06,
+      "loss": 1.1353,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2914412959554982,
+      "learning_rate": 9.916282407835826e-06,
+      "loss": 0.572,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1262196167141143,
+      "learning_rate": 9.916093426221445e-06,
+      "loss": 0.8174,
+      "step": 1354
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5431110873100438,
+      "learning_rate": 9.91590423335158e-06,
+      "loss": 0.8319,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.136053048567738,
+      "learning_rate": 9.915714829234361e-06,
+      "loss": 0.794,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0578274678415664,
+      "learning_rate": 9.91552521387793e-06,
+      "loss": 0.682,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0940911272756908,
+      "learning_rate": 9.915335387290432e-06,
+      "loss": 0.5913,
+      "step": 1358
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7700414258639694,
+      "learning_rate": 9.915145349480027e-06,
+      "loss": 0.864,
+      "step": 1359
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3157297254351272,
+      "learning_rate": 9.914955100454878e-06,
+      "loss": 0.9196,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.153600846672862,
+      "learning_rate": 9.914764640223162e-06,
+      "loss": 0.8157,
+      "step": 1361
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0940582532581207,
+      "learning_rate": 9.914573968793064e-06,
+      "loss": 0.8031,
+      "step": 1362
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5604210073935723,
+      "learning_rate": 9.914383086172778e-06,
+      "loss": 0.737,
+      "step": 1363
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.192589419056847,
+      "learning_rate": 9.914191992370504e-06,
+      "loss": 1.0165,
+      "step": 1364
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1270285854972992,
+      "learning_rate": 9.914000687394457e-06,
+      "loss": 0.8945,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4391550018583894,
+      "learning_rate": 9.913809171252856e-06,
+      "loss": 0.8707,
+      "step": 1366
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6446493478209874,
+      "learning_rate": 9.913617443953931e-06,
+      "loss": 1.0031,
+      "step": 1367
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.995363676563497,
+      "learning_rate": 9.913425505505923e-06,
+      "loss": 0.8523,
+      "step": 1368
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3965669949326593,
+      "learning_rate": 9.913233355917075e-06,
+      "loss": 0.9806,
+      "step": 1369
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.652791946205701,
+      "learning_rate": 9.913040995195651e-06,
+      "loss": 1.0548,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.46747004648607,
+      "learning_rate": 9.91284842334991e-06,
+      "loss": 1.038,
+      "step": 1371
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.841965738225694,
+      "learning_rate": 9.912655640388134e-06,
+      "loss": 1.0189,
+      "step": 1372
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9106025838824792,
+      "learning_rate": 9.912462646318602e-06,
+      "loss": 1.0395,
+      "step": 1373
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3766549609124485,
+      "learning_rate": 9.91226944114961e-06,
+      "loss": 0.9533,
+      "step": 1374
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.191970178920618,
+      "learning_rate": 9.91207602488946e-06,
+      "loss": 0.9638,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.0428106559063,
+      "learning_rate": 9.911882397546464e-06,
+      "loss": 0.896,
+      "step": 1376
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6591828226763257,
+      "learning_rate": 9.91168855912894e-06,
+      "loss": 0.6653,
+      "step": 1377
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.576471266361443,
+      "learning_rate": 9.911494509645222e-06,
+      "loss": 0.9193,
+      "step": 1378
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.369717421858428,
+      "learning_rate": 9.911300249103646e-06,
+      "loss": 0.8007,
+      "step": 1379
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.547977059785569,
+      "learning_rate": 9.91110577751256e-06,
+      "loss": 1.112,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0272481915234017,
+      "learning_rate": 9.91091109488032e-06,
+      "loss": 0.8358,
+      "step": 1381
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3658194046498178,
+      "learning_rate": 9.910716201215294e-06,
+      "loss": 0.8076,
+      "step": 1382
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9092533886534806,
+      "learning_rate": 9.910521096525856e-06,
+      "loss": 0.8437,
+      "step": 1383
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.398064185096333,
+      "learning_rate": 9.910325780820391e-06,
+      "loss": 0.7521,
+      "step": 1384
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9596288586252315,
+      "learning_rate": 9.910130254107288e-06,
+      "loss": 0.9512,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1420771863661527,
+      "learning_rate": 9.909934516394957e-06,
+      "loss": 0.6782,
+      "step": 1386
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.011896187076245,
+      "learning_rate": 9.909738567691803e-06,
+      "loss": 0.7522,
+      "step": 1387
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4047726616574896,
+      "learning_rate": 9.909542408006247e-06,
+      "loss": 0.8288,
+      "step": 1388
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1299403128902523,
+      "learning_rate": 9.90934603734672e-06,
+      "loss": 0.8193,
+      "step": 1389
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.234983041388055,
+      "learning_rate": 9.909149455721661e-06,
+      "loss": 0.7467,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.273420781399921,
+      "learning_rate": 9.908952663139518e-06,
+      "loss": 0.9249,
+      "step": 1391
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9320292477568954,
+      "learning_rate": 9.908755659608743e-06,
+      "loss": 0.9041,
+      "step": 1392
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3433997009769816,
+      "learning_rate": 9.908558445137807e-06,
+      "loss": 1.1351,
+      "step": 1393
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9968888220048322,
+      "learning_rate": 9.908361019735181e-06,
+      "loss": 0.9014,
+      "step": 1394
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1725688833957335,
+      "learning_rate": 9.908163383409353e-06,
+      "loss": 1.0259,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.767073804109383,
+      "learning_rate": 9.907965536168811e-06,
+      "loss": 0.8688,
+      "step": 1396
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.004958760111661,
+      "learning_rate": 9.907767478022061e-06,
+      "loss": 1.0956,
+      "step": 1397
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9461412093101598,
+      "learning_rate": 9.90756920897761e-06,
+      "loss": 1.1888,
+      "step": 1398
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3589552689172955,
+      "learning_rate": 9.907370729043984e-06,
+      "loss": 0.6927,
+      "step": 1399
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4763802472383794,
+      "learning_rate": 9.907172038229706e-06,
+      "loss": 1.0817,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2260268527562606,
+      "learning_rate": 9.906973136543317e-06,
+      "loss": 0.7794,
+      "step": 1401
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4017923592570924,
+      "learning_rate": 9.906774023993362e-06,
+      "loss": 0.9079,
+      "step": 1402
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4157320180421613,
+      "learning_rate": 9.906574700588403e-06,
+      "loss": 0.8509,
+      "step": 1403
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1493411984569435,
+      "learning_rate": 9.906375166336998e-06,
+      "loss": 0.8369,
+      "step": 1404
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.238842073366781,
+      "learning_rate": 9.906175421247728e-06,
+      "loss": 0.9086,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7091549042695178,
+      "learning_rate": 9.905975465329172e-06,
+      "loss": 0.8074,
+      "step": 1406
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.570836463224604,
+      "learning_rate": 9.905775298589923e-06,
+      "loss": 0.8618,
+      "step": 1407
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.129613229951485,
+      "learning_rate": 9.905574921038586e-06,
+      "loss": 1.0473,
+      "step": 1408
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2915285993721324,
+      "learning_rate": 9.905374332683768e-06,
+      "loss": 0.8433,
+      "step": 1409
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.311458586017917,
+      "learning_rate": 9.90517353353409e-06,
+      "loss": 0.9196,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5442616037944554,
+      "learning_rate": 9.904972523598183e-06,
+      "loss": 0.9619,
+      "step": 1411
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.896994823235115,
+      "learning_rate": 9.90477130288468e-06,
+      "loss": 1.0102,
+      "step": 1412
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6578313699515412,
+      "learning_rate": 9.904569871402233e-06,
+      "loss": 0.8064,
+      "step": 1413
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.276685534039032,
+      "learning_rate": 9.904368229159494e-06,
+      "loss": 0.9575,
+      "step": 1414
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9419152783785805,
+      "learning_rate": 9.904166376165129e-06,
+      "loss": 0.7942,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1737359394117184,
+      "learning_rate": 9.903964312427813e-06,
+      "loss": 1.029,
+      "step": 1416
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.32593226770928,
+      "learning_rate": 9.903762037956228e-06,
+      "loss": 0.8849,
+      "step": 1417
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3726381827284704,
+      "learning_rate": 9.903559552759069e-06,
+      "loss": 0.8154,
+      "step": 1418
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.829467816313093,
+      "learning_rate": 9.903356856845035e-06,
+      "loss": 0.7665,
+      "step": 1419
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.124343893486279,
+      "learning_rate": 9.903153950222834e-06,
+      "loss": 0.8431,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.146229587036498,
+      "learning_rate": 9.90295083290119e-06,
+      "loss": 0.6562,
+      "step": 1421
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.260918827992838,
+      "learning_rate": 9.90274750488883e-06,
+      "loss": 0.8175,
+      "step": 1422
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.002098937182269,
+      "learning_rate": 9.902543966194489e-06,
+      "loss": 0.8965,
+      "step": 1423
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1446625098007355,
+      "learning_rate": 9.902340216826915e-06,
+      "loss": 0.6099,
+      "step": 1424
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.16563812179084,
+      "learning_rate": 9.902136256794865e-06,
+      "loss": 0.9204,
+      "step": 1425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1374962452875113,
+      "learning_rate": 9.901932086107103e-06,
+      "loss": 0.8406,
+      "step": 1426
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0963472558760765,
+      "learning_rate": 9.901727704772401e-06,
+      "loss": 0.9774,
+      "step": 1427
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.398951665706738,
+      "learning_rate": 9.901523112799543e-06,
+      "loss": 1.0724,
+      "step": 1428
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.489950865656102,
+      "learning_rate": 9.90131831019732e-06,
+      "loss": 0.7379,
+      "step": 1429
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.763839305111693,
+      "learning_rate": 9.901113296974533e-06,
+      "loss": 0.9743,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2243753604052525,
+      "learning_rate": 9.900908073139995e-06,
+      "loss": 1.0074,
+      "step": 1431
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7355222193647708,
+      "learning_rate": 9.90070263870252e-06,
+      "loss": 0.6741,
+      "step": 1432
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7268786569750265,
+      "learning_rate": 9.900496993670937e-06,
+      "loss": 1.0708,
+      "step": 1433
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4147086966501548,
+      "learning_rate": 9.900291138054086e-06,
+      "loss": 0.7549,
+      "step": 1434
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.221784836293975,
+      "learning_rate": 9.90008507186081e-06,
+      "loss": 0.8911,
+      "step": 1435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.502015974487028,
+      "learning_rate": 9.899878795099966e-06,
+      "loss": 0.8068,
+      "step": 1436
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0685968188145551,
+      "learning_rate": 9.899672307780419e-06,
+      "loss": 0.6898,
+      "step": 1437
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.040445389698276,
+      "learning_rate": 9.89946560991104e-06,
+      "loss": 0.7349,
+      "step": 1438
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.431730260899017,
+      "learning_rate": 9.899258701500712e-06,
+      "loss": 0.7177,
+      "step": 1439
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.999635127272343,
+      "learning_rate": 9.899051582558325e-06,
+      "loss": 1.0037,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6022667314450882,
+      "learning_rate": 9.898844253092782e-06,
+      "loss": 0.9796,
+      "step": 1441
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.125998732314607,
+      "learning_rate": 9.898636713112992e-06,
+      "loss": 0.6823,
+      "step": 1442
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6108494346905844,
+      "learning_rate": 9.89842896262787e-06,
+      "loss": 0.7568,
+      "step": 1443
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.08617592886913,
+      "learning_rate": 9.89822100164635e-06,
+      "loss": 1.0915,
+      "step": 1444
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.577050610381808,
+      "learning_rate": 9.898012830177361e-06,
+      "loss": 1.0241,
+      "step": 1445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4348864456564456,
+      "learning_rate": 9.897804448229854e-06,
+      "loss": 0.9805,
+      "step": 1446
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.158601768579909,
+      "learning_rate": 9.89759585581278e-06,
+      "loss": 1.0288,
+      "step": 1447
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4320849228892714,
+      "learning_rate": 9.897387052935106e-06,
+      "loss": 0.8682,
+      "step": 1448
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3507231370555113,
+      "learning_rate": 9.897178039605803e-06,
+      "loss": 1.0467,
+      "step": 1449
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6310809605383167,
+      "learning_rate": 9.896968815833853e-06,
+      "loss": 0.9135,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3722368798217093,
+      "learning_rate": 9.896759381628247e-06,
+      "loss": 0.881,
+      "step": 1451
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.216833416003348,
+      "learning_rate": 9.896549736997985e-06,
+      "loss": 0.7168,
+      "step": 1452
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6488321650982116,
+      "learning_rate": 9.896339881952076e-06,
+      "loss": 0.9269,
+      "step": 1453
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2378706267467907,
+      "learning_rate": 9.896129816499535e-06,
+      "loss": 0.6898,
+      "step": 1454
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.304037808836499,
+      "learning_rate": 9.895919540649395e-06,
+      "loss": 0.6601,
+      "step": 1455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.096479719357336,
+      "learning_rate": 9.895709054410686e-06,
+      "loss": 0.9211,
+      "step": 1456
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5775151891087953,
+      "learning_rate": 9.895498357792456e-06,
+      "loss": 0.9269,
+      "step": 1457
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3500720959240917,
+      "learning_rate": 9.89528745080376e-06,
+      "loss": 0.848,
+      "step": 1458
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6031849922953967,
+      "learning_rate": 9.89507633345366e-06,
+      "loss": 0.9229,
+      "step": 1459
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.240635230920266,
+      "learning_rate": 9.894865005751227e-06,
+      "loss": 0.9115,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8448070799298484,
+      "learning_rate": 9.894653467705542e-06,
+      "loss": 0.987,
+      "step": 1461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.33233100709725,
+      "learning_rate": 9.894441719325698e-06,
+      "loss": 1.0548,
+      "step": 1462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.376689066370778,
+      "learning_rate": 9.894229760620793e-06,
+      "loss": 0.9264,
+      "step": 1463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.15465746799706,
+      "learning_rate": 9.894017591599934e-06,
+      "loss": 1.0527,
+      "step": 1464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.121550233768819,
+      "learning_rate": 9.89380521227224e-06,
+      "loss": 0.9555,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3730426380246892,
+      "learning_rate": 9.893592622646838e-06,
+      "loss": 0.7394,
+      "step": 1466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1739973410691524,
+      "learning_rate": 9.893379822732863e-06,
+      "loss": 0.9412,
+      "step": 1467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1149524609695307,
+      "learning_rate": 9.893166812539456e-06,
+      "loss": 1.0107,
+      "step": 1468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1645522200842175,
+      "learning_rate": 9.892953592075776e-06,
+      "loss": 0.9114,
+      "step": 1469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.605317993128238,
+      "learning_rate": 9.892740161350981e-06,
+      "loss": 0.8858,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3360142560086454,
+      "learning_rate": 9.892526520374246e-06,
+      "loss": 0.9273,
+      "step": 1471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0791920316251646,
+      "learning_rate": 9.892312669154749e-06,
+      "loss": 0.8733,
+      "step": 1472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8217815464722504,
+      "learning_rate": 9.892098607701682e-06,
+      "loss": 0.8901,
+      "step": 1473
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.320023312470209,
+      "learning_rate": 9.891884336024242e-06,
+      "loss": 0.9189,
+      "step": 1474
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6100224096137445,
+      "learning_rate": 9.891669854131636e-06,
+      "loss": 1.0103,
+      "step": 1475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.079991439101215,
+      "learning_rate": 9.891455162033085e-06,
+      "loss": 0.7908,
+      "step": 1476
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7298977758259402,
+      "learning_rate": 9.891240259737809e-06,
+      "loss": 0.9153,
+      "step": 1477
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6687259855025562,
+      "learning_rate": 9.891025147255047e-06,
+      "loss": 0.8942,
+      "step": 1478
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4469443245665197,
+      "learning_rate": 9.890809824594041e-06,
+      "loss": 1.0636,
+      "step": 1479
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.661768300608605,
+      "learning_rate": 9.890594291764044e-06,
+      "loss": 1.0483,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4252391136921787,
+      "learning_rate": 9.89037854877432e-06,
+      "loss": 1.1263,
+      "step": 1481
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1162615848385449,
+      "learning_rate": 9.890162595634138e-06,
+      "loss": 0.5785,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.199045388108663,
+      "learning_rate": 9.889946432352776e-06,
+      "loss": 0.7594,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4935241900460126,
+      "learning_rate": 9.889730058939529e-06,
+      "loss": 0.8048,
+      "step": 1484
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3355110254927296,
+      "learning_rate": 9.889513475403689e-06,
+      "loss": 0.872,
+      "step": 1485
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.514333320539631,
+      "learning_rate": 9.889296681754565e-06,
+      "loss": 0.8362,
+      "step": 1486
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.565319174918041,
+      "learning_rate": 9.889079678001474e-06,
+      "loss": 0.8991,
+      "step": 1487
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0216575208842413,
+      "learning_rate": 9.88886246415374e-06,
+      "loss": 0.8274,
+      "step": 1488
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7323430970789264,
+      "learning_rate": 9.8886450402207e-06,
+      "loss": 0.8945,
+      "step": 1489
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.579920576840676,
+      "learning_rate": 9.888427406211694e-06,
+      "loss": 1.0789,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8575717237109193,
+      "learning_rate": 9.888209562136074e-06,
+      "loss": 0.9093,
+      "step": 1491
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.050511172107358,
+      "learning_rate": 9.887991508003205e-06,
+      "loss": 0.7642,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.133188511680867,
+      "learning_rate": 9.887773243822454e-06,
+      "loss": 1.0207,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1689817286840802,
+      "learning_rate": 9.8875547696032e-06,
+      "loss": 1.1702,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.031027472495129,
+      "learning_rate": 9.887336085354834e-06,
+      "loss": 0.8877,
+      "step": 1495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.286613439318881,
+      "learning_rate": 9.887117191086749e-06,
+      "loss": 1.0079,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2047581346332052,
+      "learning_rate": 9.886898086808356e-06,
+      "loss": 0.715,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5513729082267504,
+      "learning_rate": 9.886678772529069e-06,
+      "loss": 0.8693,
+      "step": 1498
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1631560647088777,
+      "learning_rate": 9.88645924825831e-06,
+      "loss": 1.0053,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6491059105393022,
+      "learning_rate": 9.886239514005515e-06,
+      "loss": 1.0721,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.972094343755942,
+      "learning_rate": 9.886019569780128e-06,
+      "loss": 0.9092,
+      "step": 1501
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.458498925553126,
+      "learning_rate": 9.885799415591594e-06,
+      "loss": 1.0554,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9222909731015005,
+      "learning_rate": 9.88557905144938e-06,
+      "loss": 0.7508,
+      "step": 1503
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.614783967362693,
+      "learning_rate": 9.885358477362956e-06,
+      "loss": 0.9242,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7627997546623604,
+      "learning_rate": 9.885137693341795e-06,
+      "loss": 0.9983,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.126981482056467,
+      "learning_rate": 9.884916699395389e-06,
+      "loss": 0.7597,
+      "step": 1506
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2634279799901793,
+      "learning_rate": 9.884695495533232e-06,
+      "loss": 0.6496,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.277097801279786,
+      "learning_rate": 9.884474081764833e-06,
+      "loss": 0.9118,
+      "step": 1508
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9457694360999653,
+      "learning_rate": 9.8842524580997e-06,
+      "loss": 0.8201,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6911774816924017,
+      "learning_rate": 9.884030624547364e-06,
+      "loss": 0.9235,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9077550760266855,
+      "learning_rate": 9.883808581117355e-06,
+      "loss": 0.7895,
+      "step": 1511
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.77711808361038,
+      "learning_rate": 9.883586327819214e-06,
+      "loss": 0.8482,
+      "step": 1512
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7563736736852005,
+      "learning_rate": 9.883363864662491e-06,
+      "loss": 0.8938,
+      "step": 1513
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2237129780210425,
+      "learning_rate": 9.883141191656748e-06,
+      "loss": 0.722,
+      "step": 1514
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.198255220404255,
+      "learning_rate": 9.882918308811554e-06,
+      "loss": 0.9801,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.455611192761718,
+      "learning_rate": 9.882695216136485e-06,
+      "loss": 0.8855,
+      "step": 1516
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3488035132242135,
+      "learning_rate": 9.882471913641126e-06,
+      "loss": 0.9795,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1732727175919733,
+      "learning_rate": 9.882248401335076e-06,
+      "loss": 1.0503,
+      "step": 1518
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.881638024083401,
+      "learning_rate": 9.88202467922794e-06,
+      "loss": 0.8057,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.428926017767456,
+      "learning_rate": 9.881800747329328e-06,
+      "loss": 0.8727,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0549254637163252,
+      "learning_rate": 9.881576605648869e-06,
+      "loss": 0.6919,
+      "step": 1521
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4556515342256096,
+      "learning_rate": 9.881352254196189e-06,
+      "loss": 0.8535,
+      "step": 1522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0562032864861677,
+      "learning_rate": 9.881127692980932e-06,
+      "loss": 0.8188,
+      "step": 1523
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0265076901025068,
+      "learning_rate": 9.880902922012747e-06,
+      "loss": 0.9344,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.317700419206205,
+      "learning_rate": 9.880677941301293e-06,
+      "loss": 0.9333,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.309483558174134,
+      "learning_rate": 9.880452750856239e-06,
+      "loss": 1.0744,
+      "step": 1526
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1958080070318102,
+      "learning_rate": 9.880227350687259e-06,
+      "loss": 1.0175,
+      "step": 1527
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2928804321499747,
+      "learning_rate": 9.880001740804043e-06,
+      "loss": 1.0432,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8294351928832966,
+      "learning_rate": 9.879775921216284e-06,
+      "loss": 1.1286,
+      "step": 1529
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2056092233330284,
+      "learning_rate": 9.879549891933684e-06,
+      "loss": 1.0717,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4407966465247972,
+      "learning_rate": 9.879323652965957e-06,
+      "loss": 0.8353,
+      "step": 1531
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.447230307465269,
+      "learning_rate": 9.879097204322828e-06,
+      "loss": 0.9803,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.199504115758272,
+      "learning_rate": 9.878870546014025e-06,
+      "loss": 0.721,
+      "step": 1533
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1838844777783675,
+      "learning_rate": 9.87864367804929e-06,
+      "loss": 0.9466,
+      "step": 1534
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.824542519004607,
+      "learning_rate": 9.87841660043837e-06,
+      "loss": 0.79,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8359926803476263,
+      "learning_rate": 9.878189313191023e-06,
+      "loss": 0.987,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2169500306352465,
+      "learning_rate": 9.877961816317018e-06,
+      "loss": 0.9243,
+      "step": 1537
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0086241257644715,
+      "learning_rate": 9.87773410982613e-06,
+      "loss": 0.6699,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6819754968600655,
+      "learning_rate": 9.877506193728144e-06,
+      "loss": 0.7891,
+      "step": 1539
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9009694767543146,
+      "learning_rate": 9.877278068032852e-06,
+      "loss": 1.0215,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.343089441110634,
+      "learning_rate": 9.877049732750062e-06,
+      "loss": 0.9049,
+      "step": 1541
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1680194929100944,
+      "learning_rate": 9.876821187889581e-06,
+      "loss": 0.7269,
+      "step": 1542
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2449124045085065,
+      "learning_rate": 9.876592433461234e-06,
+      "loss": 0.9846,
+      "step": 1543
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.448611508176253,
+      "learning_rate": 9.876363469474848e-06,
+      "loss": 0.8519,
+      "step": 1544
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9493021646916966,
+      "learning_rate": 9.876134295940264e-06,
+      "loss": 0.8896,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.286353898715372,
+      "learning_rate": 9.875904912867328e-06,
+      "loss": 0.8775,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2179346022715662,
+      "learning_rate": 9.875675320265903e-06,
+      "loss": 0.6325,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0678507536516135,
+      "learning_rate": 9.875445518145848e-06,
+      "loss": 0.8665,
+      "step": 1548
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.635891482208342,
+      "learning_rate": 9.87521550651704e-06,
+      "loss": 0.679,
+      "step": 1549
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8479237549032286,
+      "learning_rate": 9.874985285389364e-06,
+      "loss": 0.8331,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.216068163990032,
+      "learning_rate": 9.874754854772714e-06,
+      "loss": 0.9157,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.326225489707214,
+      "learning_rate": 9.874524214676993e-06,
+      "loss": 0.8891,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.041404753485218,
+      "learning_rate": 9.874293365112109e-06,
+      "loss": 1.0121,
+      "step": 1553
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1594841677184755,
+      "learning_rate": 9.874062306087983e-06,
+      "loss": 0.6255,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4823001664223816,
+      "learning_rate": 9.873831037614544e-06,
+      "loss": 0.8921,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.760667160334258,
+      "learning_rate": 9.873599559701732e-06,
+      "loss": 0.717,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.43453940309937,
+      "learning_rate": 9.873367872359492e-06,
+      "loss": 0.9847,
+      "step": 1557
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.942928884704879,
+      "learning_rate": 9.873135975597781e-06,
+      "loss": 0.9731,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.229206388059393,
+      "learning_rate": 9.872903869426564e-06,
+      "loss": 0.8711,
+      "step": 1559
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8547360900106968,
+      "learning_rate": 9.872671553855817e-06,
+      "loss": 0.804,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.333644294558282,
+      "learning_rate": 9.872439028895518e-06,
+      "loss": 0.8506,
+      "step": 1561
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.004849118295581,
+      "learning_rate": 9.872206294555665e-06,
+      "loss": 0.9086,
+      "step": 1562
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.321062607229485,
+      "learning_rate": 9.871973350846255e-06,
+      "loss": 0.6704,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.791758695720121,
+      "learning_rate": 9.8717401977773e-06,
+      "loss": 0.7929,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5959564033476386,
+      "learning_rate": 9.87150683535882e-06,
+      "loss": 0.9855,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0249382601378776,
+      "learning_rate": 9.871273263600842e-06,
+      "loss": 0.7191,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8675196805311876,
+      "learning_rate": 9.871039482513402e-06,
+      "loss": 0.7182,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.129406105618574,
+      "learning_rate": 9.870805492106546e-06,
+      "loss": 0.7123,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.670937613599996,
+      "learning_rate": 9.870571292390331e-06,
+      "loss": 0.773,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1888304873842492,
+      "learning_rate": 9.87033688337482e-06,
+      "loss": 0.6839,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8601525282432243,
+      "learning_rate": 9.870102265070087e-06,
+      "loss": 0.7655,
+      "step": 1571
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.067708137042178,
+      "learning_rate": 9.869867437486214e-06,
+      "loss": 0.9015,
+      "step": 1572
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.657276908874598,
+      "learning_rate": 9.86963240063329e-06,
+      "loss": 0.9358,
+      "step": 1573
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.708269471923306,
+      "learning_rate": 9.869397154521418e-06,
+      "loss": 0.6997,
+      "step": 1574
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2351125888695336,
+      "learning_rate": 9.869161699160704e-06,
+      "loss": 0.8799,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3279536666047165,
+      "learning_rate": 9.868926034561268e-06,
+      "loss": 0.7078,
+      "step": 1576
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.010383553422733,
+      "learning_rate": 9.868690160733237e-06,
+      "loss": 0.7733,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1350567546876829,
+      "learning_rate": 9.868454077686746e-06,
+      "loss": 0.6961,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.663562117803317,
+      "learning_rate": 9.868217785431942e-06,
+      "loss": 0.8628,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.223692450532544,
+      "learning_rate": 9.867981283978975e-06,
+      "loss": 0.9099,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8916966299844675,
+      "learning_rate": 9.867744573338012e-06,
+      "loss": 0.763,
+      "step": 1581
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.527408799844875,
+      "learning_rate": 9.867507653519225e-06,
+      "loss": 0.8479,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.574268011374496,
+      "learning_rate": 9.867270524532792e-06,
+      "loss": 0.7954,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4287694238301736,
+      "learning_rate": 9.867033186388906e-06,
+      "loss": 0.8141,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9037628274961802,
+      "learning_rate": 9.866795639097763e-06,
+      "loss": 0.8681,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0373546703030092,
+      "learning_rate": 9.866557882669574e-06,
+      "loss": 0.7134,
+      "step": 1586
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0743273443712829,
+      "learning_rate": 9.866319917114554e-06,
+      "loss": 0.5265,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1224427167231044,
+      "learning_rate": 9.866081742442928e-06,
+      "loss": 1.0584,
+      "step": 1588
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9882132742792624,
+      "learning_rate": 9.865843358664933e-06,
+      "loss": 1.0513,
+      "step": 1589
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2364603150158544,
+      "learning_rate": 9.865604765790812e-06,
+      "loss": 0.7092,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.214705850257007,
+      "learning_rate": 9.865365963830818e-06,
+      "loss": 1.0153,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0124164613051536,
+      "learning_rate": 9.865126952795212e-06,
+      "loss": 1.0898,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3737015851510495,
+      "learning_rate": 9.864887732694266e-06,
+      "loss": 0.8669,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.035098170883858,
+      "learning_rate": 9.86464830353826e-06,
+      "loss": 0.8773,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.855314361541872,
+      "learning_rate": 9.864408665337482e-06,
+      "loss": 1.1114,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5915051922338055,
+      "learning_rate": 9.86416881810223e-06,
+      "loss": 0.9571,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3178080707572,
+      "learning_rate": 9.86392876184281e-06,
+      "loss": 0.8172,
+      "step": 1597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4301137021436507,
+      "learning_rate": 9.86368849656954e-06,
+      "loss": 0.9104,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0043156571993266,
+      "learning_rate": 9.863448022292742e-06,
+      "loss": 0.8557,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.348967713764559,
+      "learning_rate": 9.863207339022753e-06,
+      "loss": 0.9113,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9672944640447312,
+      "learning_rate": 9.862966446769912e-06,
+      "loss": 0.8656,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2602727708013182,
+      "learning_rate": 9.862725345544573e-06,
+      "loss": 0.716,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.126599849690192,
+      "learning_rate": 9.862484035357095e-06,
+      "loss": 1.0047,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.430770236437741,
+      "learning_rate": 9.86224251621785e-06,
+      "loss": 0.9088,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2456179048438494,
+      "learning_rate": 9.862000788137215e-06,
+      "loss": 1.0014,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5220595712416856,
+      "learning_rate": 9.861758851125578e-06,
+      "loss": 0.9864,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.204726412456283,
+      "learning_rate": 9.861516705193334e-06,
+      "loss": 0.9438,
+      "step": 1607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.106913529754811,
+      "learning_rate": 9.861274350350891e-06,
+      "loss": 0.9405,
+      "step": 1608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7999397589514716,
+      "learning_rate": 9.861031786608663e-06,
+      "loss": 0.7604,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5223687814253353,
+      "learning_rate": 9.860789013977074e-06,
+      "loss": 1.0108,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.886457304773291,
+      "learning_rate": 9.860546032466553e-06,
+      "loss": 1.0201,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2539212008121363,
+      "learning_rate": 9.860302842087545e-06,
+      "loss": 0.9347,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.843948443452135,
+      "learning_rate": 9.8600594428505e-06,
+      "loss": 0.8411,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7441438104080706,
+      "learning_rate": 9.859815834765875e-06,
+      "loss": 0.9533,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5547988496528,
+      "learning_rate": 9.859572017844143e-06,
+      "loss": 1.0399,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8905000898299478,
+      "learning_rate": 9.859327992095777e-06,
+      "loss": 0.862,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4280887706214522,
+      "learning_rate": 9.859083757531265e-06,
+      "loss": 0.8752,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0952301546190557,
+      "learning_rate": 9.858839314161101e-06,
+      "loss": 0.8559,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0144694139761077,
+      "learning_rate": 9.858594661995792e-06,
+      "loss": 0.9865,
+      "step": 1619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.249777598154265,
+      "learning_rate": 9.85834980104585e-06,
+      "loss": 0.6383,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3471689954960056,
+      "learning_rate": 9.858104731321796e-06,
+      "loss": 0.9254,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2190674606830976,
+      "learning_rate": 9.857859452834161e-06,
+      "loss": 0.5768,
+      "step": 1622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.587576741936584,
+      "learning_rate": 9.857613965593488e-06,
+      "loss": 0.8003,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.623393454329269,
+      "learning_rate": 9.857368269610325e-06,
+      "loss": 0.9275,
+      "step": 1624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.46102215626018,
+      "learning_rate": 9.857122364895229e-06,
+      "loss": 0.9009,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.201298210098228,
+      "learning_rate": 9.856876251458767e-06,
+      "loss": 0.7923,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.622537638961632,
+      "learning_rate": 9.856629929311516e-06,
+      "loss": 0.9277,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3621179431835273,
+      "learning_rate": 9.85638339846406e-06,
+      "loss": 0.9585,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.317286318670959,
+      "learning_rate": 9.856136658926993e-06,
+      "loss": 0.8355,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7282695068949874,
+      "learning_rate": 9.855889710710919e-06,
+      "loss": 0.8355,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.766148047521128,
+      "learning_rate": 9.85564255382645e-06,
+      "loss": 1.1871,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.011207176830425,
+      "learning_rate": 9.855395188284206e-06,
+      "loss": 1.0569,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2989199298886547,
+      "learning_rate": 9.85514761409482e-06,
+      "loss": 0.8385,
+      "step": 1633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1673759933432546,
+      "learning_rate": 9.854899831268926e-06,
+      "loss": 1.0491,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1931714570459224,
+      "learning_rate": 9.854651839817174e-06,
+      "loss": 0.9125,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8751118827628375,
+      "learning_rate": 9.854403639750219e-06,
+      "loss": 0.8124,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1012833086904572,
+      "learning_rate": 9.854155231078728e-06,
+      "loss": 0.7114,
+      "step": 1637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.926202391669015,
+      "learning_rate": 9.853906613813378e-06,
+      "loss": 0.972,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.027246491844954,
+      "learning_rate": 9.85365778796485e-06,
+      "loss": 0.9364,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.635185905605238,
+      "learning_rate": 9.853408753543835e-06,
+      "loss": 0.842,
+      "step": 1640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.371097233367124,
+      "learning_rate": 9.85315951056104e-06,
+      "loss": 0.7206,
+      "step": 1641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5162326056944426,
+      "learning_rate": 9.852910059027172e-06,
+      "loss": 0.9053,
+      "step": 1642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.518376650279791,
+      "learning_rate": 9.852660398952949e-06,
+      "loss": 0.9871,
+      "step": 1643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2563020550823802,
+      "learning_rate": 9.852410530349102e-06,
+      "loss": 0.9388,
+      "step": 1644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1338161411021472,
+      "learning_rate": 9.852160453226367e-06,
+      "loss": 0.9116,
+      "step": 1645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3196285892678157,
+      "learning_rate": 9.851910167595493e-06,
+      "loss": 0.8997,
+      "step": 1646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5918759643976013,
+      "learning_rate": 9.851659673467232e-06,
+      "loss": 0.8943,
+      "step": 1647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.768843129038996,
+      "learning_rate": 9.85140897085235e-06,
+      "loss": 0.6799,
+      "step": 1648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2344380002299795,
+      "learning_rate": 9.851158059761617e-06,
+      "loss": 0.9471,
+      "step": 1649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5038856223324966,
+      "learning_rate": 9.850906940205822e-06,
+      "loss": 0.9158,
+      "step": 1650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3632006947361477,
+      "learning_rate": 9.85065561219575e-06,
+      "loss": 1.0043,
+      "step": 1651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0449469997303975,
+      "learning_rate": 9.850404075742204e-06,
+      "loss": 1.1407,
+      "step": 1652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2372676132872615,
+      "learning_rate": 9.850152330855992e-06,
+      "loss": 0.8493,
+      "step": 1653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.445110639565022,
+      "learning_rate": 9.849900377547933e-06,
+      "loss": 1.0341,
+      "step": 1654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.759537935901521,
+      "learning_rate": 9.849648215828853e-06,
+      "loss": 0.8415,
+      "step": 1655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.70417764010756,
+      "learning_rate": 9.849395845709588e-06,
+      "loss": 1.0166,
+      "step": 1656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.270521916467868,
+      "learning_rate": 9.849143267200985e-06,
+      "loss": 1.2016,
+      "step": 1657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8356579646489997,
+      "learning_rate": 9.848890480313894e-06,
+      "loss": 0.9287,
+      "step": 1658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3513698659379187,
+      "learning_rate": 9.848637485059183e-06,
+      "loss": 1.0795,
+      "step": 1659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3887513122208097,
+      "learning_rate": 9.848384281447718e-06,
+      "loss": 1.0588,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.542965231545289,
+      "learning_rate": 9.848130869490384e-06,
+      "loss": 0.8776,
+      "step": 1661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4089572315893686,
+      "learning_rate": 9.847877249198069e-06,
+      "loss": 0.6423,
+      "step": 1662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4305702788761643,
+      "learning_rate": 9.84762342058167e-06,
+      "loss": 0.9189,
+      "step": 1663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.093735383223394,
+      "learning_rate": 9.8473693836521e-06,
+      "loss": 0.6863,
+      "step": 1664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.755345875950476,
+      "learning_rate": 9.84711513842027e-06,
+      "loss": 0.89,
+      "step": 1665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.159749450163094,
+      "learning_rate": 9.846860684897107e-06,
+      "loss": 0.9849,
+      "step": 1666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7179791914349813,
+      "learning_rate": 9.846606023093545e-06,
+      "loss": 0.8876,
+      "step": 1667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.375803978218793,
+      "learning_rate": 9.84635115302053e-06,
+      "loss": 1.0138,
+      "step": 1668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6045268328557145,
+      "learning_rate": 9.846096074689012e-06,
+      "loss": 0.9008,
+      "step": 1669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.104460902515852,
+      "learning_rate": 9.845840788109953e-06,
+      "loss": 0.8768,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8198289977818574,
+      "learning_rate": 9.845585293294321e-06,
+      "loss": 1.0088,
+      "step": 1671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8089486410664262,
+      "learning_rate": 9.845329590253098e-06,
+      "loss": 0.7804,
+      "step": 1672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.915206718378722,
+      "learning_rate": 9.845073678997275e-06,
+      "loss": 1.0159,
+      "step": 1673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5073517115265878,
+      "learning_rate": 9.844817559537841e-06,
+      "loss": 0.9077,
+      "step": 1674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2397186904979203,
+      "learning_rate": 9.84456123188581e-06,
+      "loss": 0.7346,
+      "step": 1675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3252022401112042,
+      "learning_rate": 9.84430469605219e-06,
+      "loss": 0.5529,
+      "step": 1676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.799872679299565,
+      "learning_rate": 9.84404795204801e-06,
+      "loss": 0.8902,
+      "step": 1677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.242737773041542,
+      "learning_rate": 9.843790999884301e-06,
+      "loss": 0.9922,
+      "step": 1678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6319117048014227,
+      "learning_rate": 9.843533839572105e-06,
+      "loss": 0.7259,
+      "step": 1679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.605166511206116,
+      "learning_rate": 9.843276471122473e-06,
+      "loss": 0.8642,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9897907203089833,
+      "learning_rate": 9.843018894546466e-06,
+      "loss": 1.0308,
+      "step": 1681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9789934448746234,
+      "learning_rate": 9.84276110985515e-06,
+      "loss": 0.6411,
+      "step": 1682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5503251788517303,
+      "learning_rate": 9.842503117059604e-06,
+      "loss": 0.6788,
+      "step": 1683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6810137469292847,
+      "learning_rate": 9.842244916170913e-06,
+      "loss": 1.0546,
+      "step": 1684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.420896363470687,
+      "learning_rate": 9.841986507200174e-06,
+      "loss": 1.0174,
+      "step": 1685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0676161637308317,
+      "learning_rate": 9.841727890158493e-06,
+      "loss": 0.9734,
+      "step": 1686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.719876997923859,
+      "learning_rate": 9.84146906505698e-06,
+      "loss": 0.8794,
+      "step": 1687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2959683900819923,
+      "learning_rate": 9.841210031906758e-06,
+      "loss": 0.8635,
+      "step": 1688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.070663865037792,
+      "learning_rate": 9.840950790718959e-06,
+      "loss": 0.8265,
+      "step": 1689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1934543238431172,
+      "learning_rate": 9.840691341504725e-06,
+      "loss": 1.0615,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7070059608837322,
+      "learning_rate": 9.8404316842752e-06,
+      "loss": 1.0202,
+      "step": 1691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.749902027600447,
+      "learning_rate": 9.84017181904155e-06,
+      "loss": 0.8056,
+      "step": 1692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.279826103334833,
+      "learning_rate": 9.839911745814933e-06,
+      "loss": 1.0234,
+      "step": 1693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5846445252950496,
+      "learning_rate": 9.83965146460653e-06,
+      "loss": 0.8807,
+      "step": 1694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7681632173154145,
+      "learning_rate": 9.839390975427526e-06,
+      "loss": 1.0737,
+      "step": 1695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7813629521526884,
+      "learning_rate": 9.839130278289115e-06,
+      "loss": 0.8796,
+      "step": 1696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0092352363674757,
+      "learning_rate": 9.838869373202498e-06,
+      "loss": 0.7388,
+      "step": 1697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8182785016901046,
+      "learning_rate": 9.838608260178885e-06,
+      "loss": 0.8767,
+      "step": 1698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1439670224917762,
+      "learning_rate": 9.838346939229501e-06,
+      "loss": 0.6574,
+      "step": 1699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.446299291820476,
+      "learning_rate": 9.838085410365573e-06,
+      "loss": 0.8592,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9777749927845065,
+      "learning_rate": 9.83782367359834e-06,
+      "loss": 0.9357,
+      "step": 1701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.342742690284923,
+      "learning_rate": 9.837561728939048e-06,
+      "loss": 0.6822,
+      "step": 1702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0626441687516253,
+      "learning_rate": 9.837299576398958e-06,
+      "loss": 0.7733,
+      "step": 1703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8484042781076233,
+      "learning_rate": 9.83703721598933e-06,
+      "loss": 0.9641,
+      "step": 1704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4895470202107535,
+      "learning_rate": 9.83677464772144e-06,
+      "loss": 0.9062,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.237183613854772,
+      "learning_rate": 9.83651187160657e-06,
+      "loss": 0.8557,
+      "step": 1706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.330095272317145,
+      "learning_rate": 9.836248887656016e-06,
+      "loss": 0.8467,
+      "step": 1707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1041844949404025,
+      "learning_rate": 9.835985695881076e-06,
+      "loss": 0.9115,
+      "step": 1708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6083500891548006,
+      "learning_rate": 9.835722296293058e-06,
+      "loss": 0.9297,
+      "step": 1709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.988248926421366,
+      "learning_rate": 9.835458688903286e-06,
+      "loss": 0.8724,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3584040173074743,
+      "learning_rate": 9.835194873723085e-06,
+      "loss": 0.8532,
+      "step": 1711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.201367460477395,
+      "learning_rate": 9.834930850763789e-06,
+      "loss": 0.8595,
+      "step": 1712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.260741264479531,
+      "learning_rate": 9.834666620036748e-06,
+      "loss": 0.9581,
+      "step": 1713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4031274507719824,
+      "learning_rate": 9.834402181553314e-06,
+      "loss": 1.0555,
+      "step": 1714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.545701472762068,
+      "learning_rate": 9.834137535324852e-06,
+      "loss": 0.9245,
+      "step": 1715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0670005722076854,
+      "learning_rate": 9.833872681362735e-06,
+      "loss": 0.9304,
+      "step": 1716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.031786387106085,
+      "learning_rate": 9.833607619678343e-06,
+      "loss": 0.9478,
+      "step": 1717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7826579745684703,
+      "learning_rate": 9.833342350283067e-06,
+      "loss": 1.0062,
+      "step": 1718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5413174690735245,
+      "learning_rate": 9.833076873188303e-06,
+      "loss": 0.7109,
+      "step": 1719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1717396517729957,
+      "learning_rate": 9.832811188405464e-06,
+      "loss": 0.8611,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2057402770950627,
+      "learning_rate": 9.832545295945964e-06,
+      "loss": 0.6476,
+      "step": 1721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.08329902000696,
+      "learning_rate": 9.83227919582123e-06,
+      "loss": 0.7121,
+      "step": 1722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.706527854555998,
+      "learning_rate": 9.832012888042699e-06,
+      "loss": 0.9523,
+      "step": 1723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5571911290942078,
+      "learning_rate": 9.831746372621811e-06,
+      "loss": 0.8993,
+      "step": 1724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.157302875474052,
+      "learning_rate": 9.831479649570021e-06,
+      "loss": 1.0489,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6106064129130893,
+      "learning_rate": 9.83121271889879e-06,
+      "loss": 0.9375,
+      "step": 1726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.385210301581203,
+      "learning_rate": 9.830945580619588e-06,
+      "loss": 0.6439,
+      "step": 1727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5057488851675944,
+      "learning_rate": 9.830678234743896e-06,
+      "loss": 0.7711,
+      "step": 1728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1482357091730813,
+      "learning_rate": 9.830410681283203e-06,
+      "loss": 1.0683,
+      "step": 1729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0512470840009436,
+      "learning_rate": 9.830142920249004e-06,
+      "loss": 0.7315,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9046112111087616,
+      "learning_rate": 9.829874951652807e-06,
+      "loss": 0.9845,
+      "step": 1731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.221014698477186,
+      "learning_rate": 9.829606775506124e-06,
+      "loss": 0.7543,
+      "step": 1732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.040576626895187,
+      "learning_rate": 9.829338391820484e-06,
+      "loss": 0.9537,
+      "step": 1733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1858341175720737,
+      "learning_rate": 9.829069800607418e-06,
+      "loss": 1.0145,
+      "step": 1734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.241588093684202,
+      "learning_rate": 9.828801001878468e-06,
+      "loss": 0.6792,
+      "step": 1735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2089910295369237,
+      "learning_rate": 9.828531995645183e-06,
+      "loss": 0.996,
+      "step": 1736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.56807095532179,
+      "learning_rate": 9.828262781919124e-06,
+      "loss": 1.0335,
+      "step": 1737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8095537441848792,
+      "learning_rate": 9.827993360711863e-06,
+      "loss": 0.8836,
+      "step": 1738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.532490773130744,
+      "learning_rate": 9.827723732034972e-06,
+      "loss": 0.9672,
+      "step": 1739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7516717942233058,
+      "learning_rate": 9.827453895900043e-06,
+      "loss": 1.0275,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7325905138906608,
+      "learning_rate": 9.827183852318666e-06,
+      "loss": 0.8684,
+      "step": 1741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7798210546927482,
+      "learning_rate": 9.82691360130245e-06,
+      "loss": 0.7113,
+      "step": 1742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.935974629406583,
+      "learning_rate": 9.826643142863006e-06,
+      "loss": 1.0284,
+      "step": 1743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5175263089457802,
+      "learning_rate": 9.826372477011956e-06,
+      "loss": 0.9933,
+      "step": 1744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7971285552512435,
+      "learning_rate": 9.826101603760934e-06,
+      "loss": 0.9998,
+      "step": 1745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.340313805328199,
+      "learning_rate": 9.825830523121575e-06,
+      "loss": 0.7686,
+      "step": 1746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1942012385010683,
+      "learning_rate": 9.825559235105532e-06,
+      "loss": 0.6855,
+      "step": 1747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1887377800193284,
+      "learning_rate": 9.82528773972446e-06,
+      "loss": 0.6371,
+      "step": 1748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9509245062507412,
+      "learning_rate": 9.825016036990029e-06,
+      "loss": 0.8293,
+      "step": 1749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.085715466820096,
+      "learning_rate": 9.824744126913914e-06,
+      "loss": 0.9713,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.214490781446619,
+      "learning_rate": 9.824472009507796e-06,
+      "loss": 0.8373,
+      "step": 1751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6336375733033077,
+      "learning_rate": 9.824199684783373e-06,
+      "loss": 0.8531,
+      "step": 1752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2040215781703565,
+      "learning_rate": 9.823927152752344e-06,
+      "loss": 0.917,
+      "step": 1753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7750396849783883,
+      "learning_rate": 9.823654413426424e-06,
+      "loss": 1.0727,
+      "step": 1754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1257502267224835,
+      "learning_rate": 9.82338146681733e-06,
+      "loss": 0.9764,
+      "step": 1755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.326985770654587,
+      "learning_rate": 9.823108312936791e-06,
+      "loss": 0.9609,
+      "step": 1756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.159968704709999,
+      "learning_rate": 9.822834951796547e-06,
+      "loss": 0.7448,
+      "step": 1757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.525897251088983,
+      "learning_rate": 9.822561383408345e-06,
+      "loss": 0.8733,
+      "step": 1758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.991848455610136,
+      "learning_rate": 9.822287607783938e-06,
+      "loss": 0.7347,
+      "step": 1759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0782286729894595,
+      "learning_rate": 9.822013624935093e-06,
+      "loss": 0.9374,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.359833904929645,
+      "learning_rate": 9.821739434873586e-06,
+      "loss": 0.9646,
+      "step": 1761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2079253352350223,
+      "learning_rate": 9.821465037611194e-06,
+      "loss": 0.9786,
+      "step": 1762
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.510673525897429,
+      "learning_rate": 9.821190433159711e-06,
+      "loss": 1.0002,
+      "step": 1763
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0844314592699478,
+      "learning_rate": 9.820915621530939e-06,
+      "loss": 1.0039,
+      "step": 1764
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1310409819099427,
+      "learning_rate": 9.820640602736687e-06,
+      "loss": 1.1301,
+      "step": 1765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4504707227190585,
+      "learning_rate": 9.82036537678877e-06,
+      "loss": 0.9683,
+      "step": 1766
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4793397633809944,
+      "learning_rate": 9.820089943699018e-06,
+      "loss": 0.9602,
+      "step": 1767
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1754934471828875,
+      "learning_rate": 9.819814303479268e-06,
+      "loss": 1.0164,
+      "step": 1768
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1505112639064405,
+      "learning_rate": 9.81953845614136e-06,
+      "loss": 0.8421,
+      "step": 1769
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2214584401539765,
+      "learning_rate": 9.819262401697154e-06,
+      "loss": 0.7262,
+      "step": 1770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.190718238327891,
+      "learning_rate": 9.818986140158507e-06,
+      "loss": 0.8031,
+      "step": 1771
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1914105486899307,
+      "learning_rate": 9.818709671537295e-06,
+      "loss": 0.8742,
+      "step": 1772
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6577970240306015,
+      "learning_rate": 9.818432995845394e-06,
+      "loss": 0.7773,
+      "step": 1773
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2882873269767767,
+      "learning_rate": 9.818156113094699e-06,
+      "loss": 0.9818,
+      "step": 1774
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.908960703977235,
+      "learning_rate": 9.817879023297103e-06,
+      "loss": 0.9602,
+      "step": 1775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.425853938036722,
+      "learning_rate": 9.817601726464516e-06,
+      "loss": 1.0263,
+      "step": 1776
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9186256652672484,
+      "learning_rate": 9.817324222608853e-06,
+      "loss": 0.7734,
+      "step": 1777
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.204107788440804,
+      "learning_rate": 9.817046511742042e-06,
+      "loss": 0.7611,
+      "step": 1778
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.117002474953293,
+      "learning_rate": 9.816768593876012e-06,
+      "loss": 0.901,
+      "step": 1779
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.612842362321342,
+      "learning_rate": 9.816490469022708e-06,
+      "loss": 0.7552,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.525425088989166,
+      "learning_rate": 9.816212137194081e-06,
+      "loss": 0.9293,
+      "step": 1781
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4763495896754875,
+      "learning_rate": 9.815933598402095e-06,
+      "loss": 0.9848,
+      "step": 1782
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9890562086932728,
+      "learning_rate": 9.815654852658714e-06,
+      "loss": 0.7795,
+      "step": 1783
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9086174802592986,
+      "learning_rate": 9.81537589997592e-06,
+      "loss": 0.8494,
+      "step": 1784
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9563942280851778,
+      "learning_rate": 9.815096740365698e-06,
+      "loss": 0.9286,
+      "step": 1785
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9437258719261261,
+      "learning_rate": 9.814817373840046e-06,
+      "loss": 0.8,
+      "step": 1786
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3007451874692815,
+      "learning_rate": 9.814537800410968e-06,
+      "loss": 0.9086,
+      "step": 1787
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2662092881007174,
+      "learning_rate": 9.814258020090478e-06,
+      "loss": 0.7968,
+      "step": 1788
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2217151595885287,
+      "learning_rate": 9.8139780328906e-06,
+      "loss": 0.738,
+      "step": 1789
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.388224085947466,
+      "learning_rate": 9.813697838823364e-06,
+      "loss": 0.9663,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.54476653071069,
+      "learning_rate": 9.813417437900812e-06,
+      "loss": 1.059,
+      "step": 1791
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3030278567571862,
+      "learning_rate": 9.81313683013499e-06,
+      "loss": 1.0391,
+      "step": 1792
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9072766642969066,
+      "learning_rate": 9.81285601553796e-06,
+      "loss": 0.9997,
+      "step": 1793
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2274988783317413,
+      "learning_rate": 9.812574994121791e-06,
+      "loss": 0.6569,
+      "step": 1794
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3267382126190794,
+      "learning_rate": 9.812293765898555e-06,
+      "loss": 1.03,
+      "step": 1795
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.040997143717999,
+      "learning_rate": 9.812012330880337e-06,
+      "loss": 0.7967,
+      "step": 1796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1958630875656513,
+      "learning_rate": 9.811730689079233e-06,
+      "loss": 0.743,
+      "step": 1797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2367167806854833,
+      "learning_rate": 9.811448840507345e-06,
+      "loss": 0.7745,
+      "step": 1798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3865207593642475,
+      "learning_rate": 9.811166785176785e-06,
+      "loss": 0.9131,
+      "step": 1799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.492052339573455,
+      "learning_rate": 9.810884523099673e-06,
+      "loss": 0.9226,
+      "step": 1800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.918231458425151,
+      "learning_rate": 9.810602054288138e-06,
+      "loss": 0.6954,
+      "step": 1801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2516790215937825,
+      "learning_rate": 9.81031937875432e-06,
+      "loss": 0.7683,
+      "step": 1802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3309611699474173,
+      "learning_rate": 9.810036496510364e-06,
+      "loss": 0.8085,
+      "step": 1803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0210078594907295,
+      "learning_rate": 9.809753407568427e-06,
+      "loss": 0.6324,
+      "step": 1804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1651754431830605,
+      "learning_rate": 9.809470111940675e-06,
+      "loss": 0.9589,
+      "step": 1805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6921899140131633,
+      "learning_rate": 9.809186609639281e-06,
+      "loss": 0.9203,
+      "step": 1806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.110223508943187,
+      "learning_rate": 9.808902900676429e-06,
+      "loss": 0.7614,
+      "step": 1807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6570383358060607,
+      "learning_rate": 9.808618985064307e-06,
+      "loss": 0.9451,
+      "step": 1808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1546495058311503,
+      "learning_rate": 9.80833486281512e-06,
+      "loss": 1.0491,
+      "step": 1809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3264716354026453,
+      "learning_rate": 9.808050533941074e-06,
+      "loss": 0.957,
+      "step": 1810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7068159124932576,
+      "learning_rate": 9.807765998454388e-06,
+      "loss": 0.9458,
+      "step": 1811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.743935628231741,
+      "learning_rate": 9.80748125636729e-06,
+      "loss": 0.7721,
+      "step": 1812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0567156207763304,
+      "learning_rate": 9.807196307692015e-06,
+      "loss": 0.7297,
+      "step": 1813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.685036000383235,
+      "learning_rate": 9.80691115244081e-06,
+      "loss": 0.7757,
+      "step": 1814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.125103380667508,
+      "learning_rate": 9.806625790625925e-06,
+      "loss": 0.7362,
+      "step": 1815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.304770174004962,
+      "learning_rate": 9.806340222259627e-06,
+      "loss": 0.8635,
+      "step": 1816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.118610074477973,
+      "learning_rate": 9.806054447354185e-06,
+      "loss": 0.9868,
+      "step": 1817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2935458814127805,
+      "learning_rate": 9.805768465921878e-06,
+      "loss": 1.0847,
+      "step": 1818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.502682846093014,
+      "learning_rate": 9.805482277974999e-06,
+      "loss": 0.9373,
+      "step": 1819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3835797757916073,
+      "learning_rate": 9.805195883525844e-06,
+      "loss": 0.7454,
+      "step": 1820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.118843398467263,
+      "learning_rate": 9.804909282586718e-06,
+      "loss": 0.6825,
+      "step": 1821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.100995348186191,
+      "learning_rate": 9.804622475169943e-06,
+      "loss": 0.9792,
+      "step": 1822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2239182625920506,
+      "learning_rate": 9.804335461287835e-06,
+      "loss": 0.885,
+      "step": 1823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1689883448778822,
+      "learning_rate": 9.804048240952736e-06,
+      "loss": 0.8019,
+      "step": 1824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1601001731902745,
+      "learning_rate": 9.803760814176983e-06,
+      "loss": 0.6352,
+      "step": 1825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5162505715328267,
+      "learning_rate": 9.80347318097293e-06,
+      "loss": 1.0052,
+      "step": 1826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0347653335681097,
+      "learning_rate": 9.803185341352936e-06,
+      "loss": 0.749,
+      "step": 1827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4639404083949272,
+      "learning_rate": 9.802897295329373e-06,
+      "loss": 1.0231,
+      "step": 1828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.789871630519652,
+      "learning_rate": 9.802609042914614e-06,
+      "loss": 0.9093,
+      "step": 1829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0929459852479735,
+      "learning_rate": 9.80232058412105e-06,
+      "loss": 0.647,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3948504177949554,
+      "learning_rate": 9.802031918961073e-06,
+      "loss": 1.0293,
+      "step": 1831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.223290637631493,
+      "learning_rate": 9.801743047447092e-06,
+      "loss": 1.0345,
+      "step": 1832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1071441919360498,
+      "learning_rate": 9.801453969591517e-06,
+      "loss": 0.7392,
+      "step": 1833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2145244391881787,
+      "learning_rate": 9.80116468540677e-06,
+      "loss": 0.8958,
+      "step": 1834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.052238489446444,
+      "learning_rate": 9.800875194905286e-06,
+      "loss": 0.9888,
+      "step": 1835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2286584982687723,
+      "learning_rate": 9.800585498099501e-06,
+      "loss": 0.9281,
+      "step": 1836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2243777146501613,
+      "learning_rate": 9.800295595001868e-06,
+      "loss": 0.8533,
+      "step": 1837
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4189988944428817,
+      "learning_rate": 9.800005485624842e-06,
+      "loss": 1.0372,
+      "step": 1838
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.152406039741701,
+      "learning_rate": 9.79971516998089e-06,
+      "loss": 0.9092,
+      "step": 1839
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3570613424183042,
+      "learning_rate": 9.799424648082486e-06,
+      "loss": 0.9549,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1772857468744558,
+      "learning_rate": 9.799133919942117e-06,
+      "loss": 0.7061,
+      "step": 1841
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.2502923475054555,
+      "learning_rate": 9.798842985572274e-06,
+      "loss": 0.9296,
+      "step": 1842
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.63746941946148,
+      "learning_rate": 9.798551844985463e-06,
+      "loss": 0.9632,
+      "step": 1843
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.918769274554694,
+      "learning_rate": 9.79826049819419e-06,
+      "loss": 0.8117,
+      "step": 1844
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1654727985819466,
+      "learning_rate": 9.797968945210978e-06,
+      "loss": 0.8804,
+      "step": 1845
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6991591571668603,
+      "learning_rate": 9.797677186048356e-06,
+      "loss": 1.0061,
+      "step": 1846
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4496919779178192,
+      "learning_rate": 9.797385220718859e-06,
+      "loss": 0.9665,
+      "step": 1847
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3241854458780793,
+      "learning_rate": 9.797093049235034e-06,
+      "loss": 0.9118,
+      "step": 1848
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0837407618167387,
+      "learning_rate": 9.796800671609436e-06,
+      "loss": 0.6955,
+      "step": 1849
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.862178886451664,
+      "learning_rate": 9.796508087854632e-06,
+      "loss": 0.9351,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.529512562900625,
+      "learning_rate": 9.796215297983193e-06,
+      "loss": 0.9342,
+      "step": 1851
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.511554681106968,
+      "learning_rate": 9.7959223020077e-06,
+      "loss": 0.901,
+      "step": 1852
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2090194572468547,
+      "learning_rate": 9.795629099940746e-06,
+      "loss": 0.7454,
+      "step": 1853
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.371098379557122,
+      "learning_rate": 9.795335691794929e-06,
+      "loss": 0.7463,
+      "step": 1854
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9799243272118003,
+      "learning_rate": 9.795042077582856e-06,
+      "loss": 0.803,
+      "step": 1855
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5358346704736983,
+      "learning_rate": 9.794748257317148e-06,
+      "loss": 0.8998,
+      "step": 1856
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0817711325903208,
+      "learning_rate": 9.794454231010427e-06,
+      "loss": 1.1231,
+      "step": 1857
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4324097760866725,
+      "learning_rate": 9.794159998675331e-06,
+      "loss": 0.8283,
+      "step": 1858
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.557482598384941,
+      "learning_rate": 9.793865560324503e-06,
+      "loss": 0.8862,
+      "step": 1859
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.621789349637084,
+      "learning_rate": 9.793570915970594e-06,
+      "loss": 0.8203,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.997641631323326,
+      "learning_rate": 9.793276065626267e-06,
+      "loss": 0.94,
+      "step": 1861
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6855640303364465,
+      "learning_rate": 9.792981009304192e-06,
+      "loss": 0.8782,
+      "step": 1862
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5395585736196766,
+      "learning_rate": 9.79268574701705e-06,
+      "loss": 0.7519,
+      "step": 1863
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.160131748019118,
+      "learning_rate": 9.792390278777527e-06,
+      "loss": 0.815,
+      "step": 1864
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.578736358243283,
+      "learning_rate": 9.79209460459832e-06,
+      "loss": 0.9239,
+      "step": 1865
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.132321910264576,
+      "learning_rate": 9.791798724492136e-06,
+      "loss": 0.803,
+      "step": 1866
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1728690804080633,
+      "learning_rate": 9.791502638471689e-06,
+      "loss": 0.8814,
+      "step": 1867
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.726499302045631,
+      "learning_rate": 9.7912063465497e-06,
+      "loss": 0.9664,
+      "step": 1868
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.58157605369757,
+      "learning_rate": 9.790909848738907e-06,
+      "loss": 0.7871,
+      "step": 1869
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.113266614566401,
+      "learning_rate": 9.790613145052046e-06,
+      "loss": 0.8374,
+      "step": 1870
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0478414701567136,
+      "learning_rate": 9.790316235501869e-06,
+      "loss": 1.0844,
+      "step": 1871
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.882380789530092,
+      "learning_rate": 9.790019120101135e-06,
+      "loss": 0.6605,
+      "step": 1872
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2764201779462443,
+      "learning_rate": 9.789721798862612e-06,
+      "loss": 0.8679,
+      "step": 1873
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0995177198548287,
+      "learning_rate": 9.789424271799075e-06,
+      "loss": 0.961,
+      "step": 1874
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9791158303651541,
+      "learning_rate": 9.789126538923312e-06,
+      "loss": 0.9836,
+      "step": 1875
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9046140987195064,
+      "learning_rate": 9.788828600248114e-06,
+      "loss": 0.9043,
+      "step": 1876
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9810825975255275,
+      "learning_rate": 9.788530455786288e-06,
+      "loss": 0.6934,
+      "step": 1877
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.416633985976075,
+      "learning_rate": 9.78823210555064e-06,
+      "loss": 0.7758,
+      "step": 1878
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9712047325232425,
+      "learning_rate": 9.787933549553996e-06,
+      "loss": 0.8096,
+      "step": 1879
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3173238708826303,
+      "learning_rate": 9.787634787809185e-06,
+      "loss": 0.7522,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.069613767776303,
+      "learning_rate": 9.787335820329045e-06,
+      "loss": 0.932,
+      "step": 1881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1980327121708307,
+      "learning_rate": 9.787036647126422e-06,
+      "loss": 1.0361,
+      "step": 1882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8448533932795645,
+      "learning_rate": 9.786737268214172e-06,
+      "loss": 0.9742,
+      "step": 1883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.418168361290457,
+      "learning_rate": 9.786437683605161e-06,
+      "loss": 0.8429,
+      "step": 1884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.39698127054315,
+      "learning_rate": 9.786137893312265e-06,
+      "loss": 0.9984,
+      "step": 1885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.985050641694871,
+      "learning_rate": 9.785837897348363e-06,
+      "loss": 0.7817,
+      "step": 1886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9317230881522152,
+      "learning_rate": 9.785537695726348e-06,
+      "loss": 0.8217,
+      "step": 1887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.573913734504405,
+      "learning_rate": 9.78523728845912e-06,
+      "loss": 0.7485,
+      "step": 1888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.200519252759883,
+      "learning_rate": 9.78493667555959e-06,
+      "loss": 0.9526,
+      "step": 1889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.292030515010084,
+      "learning_rate": 9.784635857040672e-06,
+      "loss": 0.937,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.907801799671412,
+      "learning_rate": 9.784334832915297e-06,
+      "loss": 0.9863,
+      "step": 1891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8074253906298363,
+      "learning_rate": 9.784033603196397e-06,
+      "loss": 0.819,
+      "step": 1892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.072507862456159,
+      "learning_rate": 9.78373216789692e-06,
+      "loss": 0.744,
+      "step": 1893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0373196379766183,
+      "learning_rate": 9.783430527029818e-06,
+      "loss": 0.84,
+      "step": 1894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.569029020415512,
+      "learning_rate": 9.783128680608052e-06,
+      "loss": 0.8292,
+      "step": 1895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6069565869761484,
+      "learning_rate": 9.782826628644596e-06,
+      "loss": 0.8992,
+      "step": 1896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5120758914355514,
+      "learning_rate": 9.782524371152425e-06,
+      "loss": 1.1029,
+      "step": 1897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4909821365243903,
+      "learning_rate": 9.782221908144531e-06,
+      "loss": 0.8833,
+      "step": 1898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2833196210377507,
+      "learning_rate": 9.781919239633912e-06,
+      "loss": 0.7975,
+      "step": 1899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.407155002452201,
+      "learning_rate": 9.781616365633574e-06,
+      "loss": 0.8653,
+      "step": 1900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1211480277331907,
+      "learning_rate": 9.78131328615653e-06,
+      "loss": 0.9246,
+      "step": 1901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.763503347791347,
+      "learning_rate": 9.781010001215806e-06,
+      "loss": 0.7718,
+      "step": 1902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.756660352865112,
+      "learning_rate": 9.780706510824436e-06,
+      "loss": 0.9224,
+      "step": 1903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5209103283103613,
+      "learning_rate": 9.780402814995458e-06,
+      "loss": 1.015,
+      "step": 1904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.367810888117082,
+      "learning_rate": 9.780098913741926e-06,
+      "loss": 0.8047,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8576600942863677,
+      "learning_rate": 9.779794807076896e-06,
+      "loss": 0.8476,
+      "step": 1906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.452435778957075,
+      "learning_rate": 9.77949049501344e-06,
+      "loss": 0.9115,
+      "step": 1907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6288222763245486,
+      "learning_rate": 9.779185977564633e-06,
+      "loss": 0.87,
+      "step": 1908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5454184046879473,
+      "learning_rate": 9.77888125474356e-06,
+      "loss": 1.0321,
+      "step": 1909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9998848694299922,
+      "learning_rate": 9.778576326563315e-06,
+      "loss": 0.7242,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6136283521442287,
+      "learning_rate": 9.778271193037003e-06,
+      "loss": 1.0611,
+      "step": 1911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.21589134084003,
+      "learning_rate": 9.777965854177737e-06,
+      "loss": 0.9579,
+      "step": 1912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5911677236752664,
+      "learning_rate": 9.777660309998638e-06,
+      "loss": 0.8463,
+      "step": 1913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.994628662946554,
+      "learning_rate": 9.777354560512835e-06,
+      "loss": 0.9661,
+      "step": 1914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0420206448551497,
+      "learning_rate": 9.777048605733465e-06,
+      "loss": 0.9448,
+      "step": 1915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3202742015078393,
+      "learning_rate": 9.776742445673679e-06,
+      "loss": 0.798,
+      "step": 1916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.494959320212994,
+      "learning_rate": 9.77643608034663e-06,
+      "loss": 0.98,
+      "step": 1917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9019149692832356,
+      "learning_rate": 9.776129509765487e-06,
+      "loss": 1.0196,
+      "step": 1918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2334654164579413,
+      "learning_rate": 9.77582273394342e-06,
+      "loss": 0.7646,
+      "step": 1919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2207531444379836,
+      "learning_rate": 9.775515752893613e-06,
+      "loss": 0.7728,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.017777811733242,
+      "learning_rate": 9.775208566629262e-06,
+      "loss": 0.6519,
+      "step": 1921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1547980819921069,
+      "learning_rate": 9.774901175163562e-06,
+      "loss": 0.6407,
+      "step": 1922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4814175584899605,
+      "learning_rate": 9.774593578509724e-06,
+      "loss": 0.8076,
+      "step": 1923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.074834152093319,
+      "learning_rate": 9.774285776680967e-06,
+      "loss": 0.701,
+      "step": 1924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9068951894837067,
+      "learning_rate": 9.773977769690517e-06,
+      "loss": 1.0017,
+      "step": 1925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9729781371099904,
+      "learning_rate": 9.773669557551609e-06,
+      "loss": 1.035,
+      "step": 1926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9393681262124476,
+      "learning_rate": 9.77336114027749e-06,
+      "loss": 0.8769,
+      "step": 1927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9559211635415392,
+      "learning_rate": 9.773052517881412e-06,
+      "loss": 0.9015,
+      "step": 1928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.562166936546625,
+      "learning_rate": 9.772743690376636e-06,
+      "loss": 0.8545,
+      "step": 1929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.028576845763841,
+      "learning_rate": 9.772434657776435e-06,
+      "loss": 0.9553,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.174179001466808,
+      "learning_rate": 9.772125420094087e-06,
+      "loss": 0.8692,
+      "step": 1931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2915415875431706,
+      "learning_rate": 9.771815977342882e-06,
+      "loss": 0.782,
+      "step": 1932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5926576990174777,
+      "learning_rate": 9.771506329536118e-06,
+      "loss": 0.7276,
+      "step": 1933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7610326247224775,
+      "learning_rate": 9.7711964766871e-06,
+      "loss": 0.8999,
+      "step": 1934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.564812240873764,
+      "learning_rate": 9.770886418809142e-06,
+      "loss": 0.922,
+      "step": 1935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1673800479436456,
+      "learning_rate": 9.77057615591557e-06,
+      "loss": 0.6313,
+      "step": 1936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8952166396906267,
+      "learning_rate": 9.770265688019718e-06,
+      "loss": 0.8507,
+      "step": 1937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.187706688501049,
+      "learning_rate": 9.769955015134923e-06,
+      "loss": 0.8318,
+      "step": 1938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2459397896052034,
+      "learning_rate": 9.76964413727454e-06,
+      "loss": 0.9942,
+      "step": 1939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.341411311517928,
+      "learning_rate": 9.769333054451923e-06,
+      "loss": 0.9728,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.235063899059563,
+      "learning_rate": 9.769021766680444e-06,
+      "loss": 0.8103,
+      "step": 1941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.263558619662416,
+      "learning_rate": 9.768710273973479e-06,
+      "loss": 0.8561,
+      "step": 1942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.061609494840759,
+      "learning_rate": 9.768398576344411e-06,
+      "loss": 0.7592,
+      "step": 1943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.267034273056102,
+      "learning_rate": 9.768086673806638e-06,
+      "loss": 0.7783,
+      "step": 1944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9591101415525476,
+      "learning_rate": 9.767774566373563e-06,
+      "loss": 0.8961,
+      "step": 1945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.987468833697125,
+      "learning_rate": 9.767462254058593e-06,
+      "loss": 0.8404,
+      "step": 1946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.949077735612319,
+      "learning_rate": 9.767149736875156e-06,
+      "loss": 0.8423,
+      "step": 1947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0471384598433002,
+      "learning_rate": 9.766837014836675e-06,
+      "loss": 0.8615,
+      "step": 1948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0306335974948846,
+      "learning_rate": 9.766524087956592e-06,
+      "loss": 0.9288,
+      "step": 1949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.525619373682626,
+      "learning_rate": 9.766210956248354e-06,
+      "loss": 0.9152,
+      "step": 1950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9174242790186438,
+      "learning_rate": 9.765897619725416e-06,
+      "loss": 0.8727,
+      "step": 1951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1747095945500927,
+      "learning_rate": 9.765584078401242e-06,
+      "loss": 0.7608,
+      "step": 1952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2059049807531974,
+      "learning_rate": 9.765270332289307e-06,
+      "loss": 0.6494,
+      "step": 1953
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9226623149480933,
+      "learning_rate": 9.764956381403095e-06,
+      "loss": 0.8358,
+      "step": 1954
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.283112756035866,
+      "learning_rate": 9.764642225756093e-06,
+      "loss": 0.9559,
+      "step": 1955
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6214690376694714,
+      "learning_rate": 9.764327865361805e-06,
+      "loss": 0.8818,
+      "step": 1956
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4695846686795666,
+      "learning_rate": 9.764013300233736e-06,
+      "loss": 0.9409,
+      "step": 1957
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.301266239317653,
+      "learning_rate": 9.763698530385407e-06,
+      "loss": 0.7074,
+      "step": 1958
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2932554730108414,
+      "learning_rate": 9.76338355583034e-06,
+      "loss": 0.8673,
+      "step": 1959
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9216963485575507,
+      "learning_rate": 9.763068376582075e-06,
+      "loss": 0.8117,
+      "step": 1960
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3572042668169164,
+      "learning_rate": 9.762752992654155e-06,
+      "loss": 0.9472,
+      "step": 1961
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6183280285582287,
+      "learning_rate": 9.76243740406013e-06,
+      "loss": 0.9695,
+      "step": 1962
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.617372310141226,
+      "learning_rate": 9.762121610813564e-06,
+      "loss": 0.7573,
+      "step": 1963
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3969907579354,
+      "learning_rate": 9.761805612928025e-06,
+      "loss": 1.0355,
+      "step": 1964
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0809347242645835,
+      "learning_rate": 9.761489410417095e-06,
+      "loss": 0.693,
+      "step": 1965
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.839027153559159,
+      "learning_rate": 9.761173003294362e-06,
+      "loss": 0.9802,
+      "step": 1966
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.11056623470704,
+      "learning_rate": 9.76085639157342e-06,
+      "loss": 0.9785,
+      "step": 1967
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2950166430854213,
+      "learning_rate": 9.760539575267875e-06,
+      "loss": 1.1005,
+      "step": 1968
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9404612055105055,
+      "learning_rate": 9.760222554391343e-06,
+      "loss": 1.0087,
+      "step": 1969
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1252706760723594,
+      "learning_rate": 9.759905328957446e-06,
+      "loss": 0.9109,
+      "step": 1970
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6901087802888988,
+      "learning_rate": 9.759587898979816e-06,
+      "loss": 0.6709,
+      "step": 1971
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7156662651947048,
+      "learning_rate": 9.759270264472097e-06,
+      "loss": 0.9475,
+      "step": 1972
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.718060548740193,
+      "learning_rate": 9.75895242544793e-06,
+      "loss": 0.8456,
+      "step": 1973
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.108634749308191,
+      "learning_rate": 9.758634381920982e-06,
+      "loss": 1.069,
+      "step": 1974
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.972659006541597,
+      "learning_rate": 9.758316133904914e-06,
+      "loss": 0.9299,
+      "step": 1975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.009484539932172,
+      "learning_rate": 9.757997681413407e-06,
+      "loss": 1.0437,
+      "step": 1976
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8638816129660074,
+      "learning_rate": 9.757679024460142e-06,
+      "loss": 0.8134,
+      "step": 1977
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8198673803267313,
+      "learning_rate": 9.757360163058814e-06,
+      "loss": 0.752,
+      "step": 1978
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3345927603020735,
+      "learning_rate": 9.757041097223123e-06,
+      "loss": 0.8559,
+      "step": 1979
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.252858339244004,
+      "learning_rate": 9.756721826966783e-06,
+      "loss": 0.9009,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0742816596989067,
+      "learning_rate": 9.756402352303513e-06,
+      "loss": 0.9464,
+      "step": 1981
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7359790390078844,
+      "learning_rate": 9.756082673247039e-06,
+      "loss": 0.7882,
+      "step": 1982
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3113666864231095,
+      "learning_rate": 9.755762789811101e-06,
+      "loss": 0.9249,
+      "step": 1983
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4572526179799428,
+      "learning_rate": 9.755442702009443e-06,
+      "loss": 0.9228,
+      "step": 1984
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9216418020514576,
+      "learning_rate": 9.755122409855822e-06,
+      "loss": 1.1096,
+      "step": 1985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0154280404406952,
+      "learning_rate": 9.754801913364002e-06,
+      "loss": 0.8496,
+      "step": 1986
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6259209142046962,
+      "learning_rate": 9.754481212547752e-06,
+      "loss": 0.6469,
+      "step": 1987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7821311852058366,
+      "learning_rate": 9.754160307420858e-06,
+      "loss": 0.7615,
+      "step": 1988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.794645668974392,
+      "learning_rate": 9.753839197997105e-06,
+      "loss": 0.9391,
+      "step": 1989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0233260032274694,
+      "learning_rate": 9.753517884290298e-06,
+      "loss": 0.6832,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1561780159503694,
+      "learning_rate": 9.753196366314237e-06,
+      "loss": 0.6311,
+      "step": 1991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.396237548942649,
+      "learning_rate": 9.752874644082743e-06,
+      "loss": 0.9948,
+      "step": 1992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1727630658352073,
+      "learning_rate": 9.752552717609643e-06,
+      "loss": 0.9394,
+      "step": 1993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.492742440559075,
+      "learning_rate": 9.752230586908767e-06,
+      "loss": 0.9475,
+      "step": 1994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0004187596341456,
+      "learning_rate": 9.751908251993956e-06,
+      "loss": 0.8254,
+      "step": 1995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5531712229020442,
+      "learning_rate": 9.751585712879067e-06,
+      "loss": 1.1551,
+      "step": 1996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.820291339547388,
+      "learning_rate": 9.751262969577958e-06,
+      "loss": 0.8851,
+      "step": 1997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.926951316093863,
+      "learning_rate": 9.750940022104497e-06,
+      "loss": 0.9701,
+      "step": 1998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1218364344308847,
+      "learning_rate": 9.75061687047256e-06,
+      "loss": 0.7545,
+      "step": 1999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.231998771730378,
+      "learning_rate": 9.750293514696038e-06,
+      "loss": 0.8748,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.606597232862342,
+      "learning_rate": 9.749969954788825e-06,
+      "loss": 0.892,
+      "step": 2001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.855903091207772,
+      "learning_rate": 9.749646190764823e-06,
+      "loss": 1.0241,
+      "step": 2002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5419465087439654,
+      "learning_rate": 9.749322222637947e-06,
+      "loss": 1.0432,
+      "step": 2003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.29041639902589,
+      "learning_rate": 9.748998050422117e-06,
+      "loss": 0.7332,
+      "step": 2004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7975880693834205,
+      "learning_rate": 9.748673674131264e-06,
+      "loss": 0.9384,
+      "step": 2005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3781926822234003,
+      "learning_rate": 9.748349093779327e-06,
+      "loss": 1.0124,
+      "step": 2006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2936292300873682,
+      "learning_rate": 9.748024309380256e-06,
+      "loss": 0.8157,
+      "step": 2007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3009980026112338,
+      "learning_rate": 9.747699320948006e-06,
+      "loss": 0.8438,
+      "step": 2008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.395580803452439,
+      "learning_rate": 9.747374128496541e-06,
+      "loss": 0.827,
+      "step": 2009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9790406560811322,
+      "learning_rate": 9.747048732039837e-06,
+      "loss": 0.8546,
+      "step": 2010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.085617844771981,
+      "learning_rate": 9.746723131591878e-06,
+      "loss": 0.9127,
+      "step": 2011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5253721204643482,
+      "learning_rate": 9.746397327166653e-06,
+      "loss": 0.903,
+      "step": 2012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7311233266598594,
+      "learning_rate": 9.746071318778165e-06,
+      "loss": 0.7552,
+      "step": 2013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9012478719687693,
+      "learning_rate": 9.745745106440422e-06,
+      "loss": 1.0108,
+      "step": 2014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.60934813146778,
+      "learning_rate": 9.745418690167444e-06,
+      "loss": 0.9249,
+      "step": 2015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.060629527781592,
+      "learning_rate": 9.745092069973254e-06,
+      "loss": 0.9215,
+      "step": 2016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8701304882107583,
+      "learning_rate": 9.74476524587189e-06,
+      "loss": 1.1341,
+      "step": 2017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2091693869965898,
+      "learning_rate": 9.744438217877396e-06,
+      "loss": 0.8407,
+      "step": 2018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.142381372565363,
+      "learning_rate": 9.744110986003826e-06,
+      "loss": 1.1076,
+      "step": 2019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1580001530701398,
+      "learning_rate": 9.743783550265242e-06,
+      "loss": 0.6335,
+      "step": 2020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.419897769359722,
+      "learning_rate": 9.743455910675712e-06,
+      "loss": 0.8552,
+      "step": 2021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.861702874218577,
+      "learning_rate": 9.74312806724932e-06,
+      "loss": 0.8299,
+      "step": 2022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.13473748920817,
+      "learning_rate": 9.74280002000015e-06,
+      "loss": 1.0461,
+      "step": 2023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9375690115074073,
+      "learning_rate": 9.742471768942299e-06,
+      "loss": 0.9587,
+      "step": 2024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.992009547899281,
+      "learning_rate": 9.742143314089877e-06,
+      "loss": 0.9696,
+      "step": 2025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1108493848488266,
+      "learning_rate": 9.741814655456994e-06,
+      "loss": 0.8819,
+      "step": 2026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.711433017682618,
+      "learning_rate": 9.741485793057774e-06,
+      "loss": 1.1696,
+      "step": 2027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.509115217019329,
+      "learning_rate": 9.74115672690635e-06,
+      "loss": 0.8174,
+      "step": 2028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.687046698022144,
+      "learning_rate": 9.740827457016863e-06,
+      "loss": 0.9302,
+      "step": 2029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1710012246438655,
+      "learning_rate": 9.74049798340346e-06,
+      "loss": 0.6366,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.658778621894954,
+      "learning_rate": 9.740168306080303e-06,
+      "loss": 0.7274,
+      "step": 2031
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.071147434467376,
+      "learning_rate": 9.739838425061557e-06,
+      "loss": 0.9266,
+      "step": 2032
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1739489649645891,
+      "learning_rate": 9.739508340361397e-06,
+      "loss": 0.674,
+      "step": 2033
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.436149194748074,
+      "learning_rate": 9.739178051994008e-06,
+      "loss": 0.9224,
+      "step": 2034
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.553771627593203,
+      "learning_rate": 9.738847559973584e-06,
+      "loss": 0.6411,
+      "step": 2035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.632523432136379,
+      "learning_rate": 9.738516864314324e-06,
+      "loss": 0.8335,
+      "step": 2036
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.172440035986641,
+      "learning_rate": 9.738185965030444e-06,
+      "loss": 0.9378,
+      "step": 2037
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1472638647511686,
+      "learning_rate": 9.73785486213616e-06,
+      "loss": 0.8213,
+      "step": 2038
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.548049910113574,
+      "learning_rate": 9.7375235556457e-06,
+      "loss": 0.9688,
+      "step": 2039
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.473111098468926,
+      "learning_rate": 9.7371920455733e-06,
+      "loss": 1.0286,
+      "step": 2040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.293757046662831,
+      "learning_rate": 9.736860331933212e-06,
+      "loss": 0.7976,
+      "step": 2041
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7502800455860146,
+      "learning_rate": 9.736528414739681e-06,
+      "loss": 0.7994,
+      "step": 2042
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3064539851346377,
+      "learning_rate": 9.736196294006978e-06,
+      "loss": 0.8609,
+      "step": 2043
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.198666079609974,
+      "learning_rate": 9.735863969749373e-06,
+      "loss": 0.899,
+      "step": 2044
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2689011088826272,
+      "learning_rate": 9.735531441981145e-06,
+      "loss": 0.8131,
+      "step": 2045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.856307080331479,
+      "learning_rate": 9.735198710716585e-06,
+      "loss": 1.102,
+      "step": 2046
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0675026508028784,
+      "learning_rate": 9.734865775969988e-06,
+      "loss": 0.6409,
+      "step": 2047
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1274964266717182,
+      "learning_rate": 9.734532637755665e-06,
+      "loss": 0.7718,
+      "step": 2048
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.800018463562249,
+      "learning_rate": 9.734199296087932e-06,
+      "loss": 0.9635,
+      "step": 2049
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2815204435325533,
+      "learning_rate": 9.73386575098111e-06,
+      "loss": 0.8634,
+      "step": 2050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9939656422728125,
+      "learning_rate": 9.733532002449533e-06,
+      "loss": 1.0393,
+      "step": 2051
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4091078461550244,
+      "learning_rate": 9.733198050507544e-06,
+      "loss": 1.0859,
+      "step": 2052
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.048240630213825,
+      "learning_rate": 9.732863895169496e-06,
+      "loss": 0.9128,
+      "step": 2053
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7854228150236002,
+      "learning_rate": 9.732529536449741e-06,
+      "loss": 0.9586,
+      "step": 2054
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.187984038475194,
+      "learning_rate": 9.732194974362657e-06,
+      "loss": 0.8242,
+      "step": 2055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.595720040257059,
+      "learning_rate": 9.731860208922612e-06,
+      "loss": 0.9346,
+      "step": 2056
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2960687774003476,
+      "learning_rate": 9.731525240143998e-06,
+      "loss": 1.0882,
+      "step": 2057
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.85885152516942,
+      "learning_rate": 9.731190068041205e-06,
+      "loss": 0.8271,
+      "step": 2058
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1093582681577967,
+      "learning_rate": 9.730854692628637e-06,
+      "loss": 0.7602,
+      "step": 2059
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0334963188506032,
+      "learning_rate": 9.730519113920708e-06,
+      "loss": 0.9636,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.396978045727194,
+      "learning_rate": 9.730183331931838e-06,
+      "loss": 1.0062,
+      "step": 2061
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.478545458014925,
+      "learning_rate": 9.729847346676455e-06,
+      "loss": 0.8625,
+      "step": 2062
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8123221143153128,
+      "learning_rate": 9.729511158168999e-06,
+      "loss": 0.801,
+      "step": 2063
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.380052684302814,
+      "learning_rate": 9.729174766423912e-06,
+      "loss": 0.7883,
+      "step": 2064
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.506753855785254,
+      "learning_rate": 9.728838171455655e-06,
+      "loss": 0.8041,
+      "step": 2065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9752412193236424,
+      "learning_rate": 9.72850137327869e-06,
+      "loss": 0.782,
+      "step": 2066
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.413913719778843,
+      "learning_rate": 9.728164371907489e-06,
+      "loss": 1.0191,
+      "step": 2067
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3172184354847567,
+      "learning_rate": 9.727827167356537e-06,
+      "loss": 1.0095,
+      "step": 2068
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1367583916794843,
+      "learning_rate": 9.72748975964032e-06,
+      "loss": 0.9754,
+      "step": 2069
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.85114125957553,
+      "learning_rate": 9.72715214877334e-06,
+      "loss": 0.732,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0625480876311157,
+      "learning_rate": 9.726814334770106e-06,
+      "loss": 0.9157,
+      "step": 2071
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0965706107819457,
+      "learning_rate": 9.72647631764513e-06,
+      "loss": 0.7638,
+      "step": 2072
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.500092452383224,
+      "learning_rate": 9.72613809741294e-06,
+      "loss": 0.9881,
+      "step": 2073
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1072126083563587,
+      "learning_rate": 9.725799674088072e-06,
+      "loss": 0.6978,
+      "step": 2074
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9274188831953616,
+      "learning_rate": 9.725461047685067e-06,
+      "loss": 0.8336,
+      "step": 2075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6373417398119425,
+      "learning_rate": 9.725122218218477e-06,
+      "loss": 0.8785,
+      "step": 2076
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5024875988795316,
+      "learning_rate": 9.72478318570286e-06,
+      "loss": 0.9652,
+      "step": 2077
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4025185595631546,
+      "learning_rate": 9.724443950152789e-06,
+      "loss": 0.934,
+      "step": 2078
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.280608906688087,
+      "learning_rate": 9.724104511582838e-06,
+      "loss": 0.8567,
+      "step": 2079
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2353547392096766,
+      "learning_rate": 9.723764870007596e-06,
+      "loss": 0.8736,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6688720220841065,
+      "learning_rate": 9.723425025441657e-06,
+      "loss": 0.9629,
+      "step": 2081
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.636859819248619,
+      "learning_rate": 9.723084977899624e-06,
+      "loss": 0.7932,
+      "step": 2082
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.449079861868277,
+      "learning_rate": 9.722744727396111e-06,
+      "loss": 0.7643,
+      "step": 2083
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6902658704356277,
+      "learning_rate": 9.72240427394574e-06,
+      "loss": 0.9322,
+      "step": 2084
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3198196806672824,
+      "learning_rate": 9.72206361756314e-06,
+      "loss": 0.948,
+      "step": 2085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.823116783220429,
+      "learning_rate": 9.721722758262948e-06,
+      "loss": 0.8371,
+      "step": 2086
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.458326581023423,
+      "learning_rate": 9.721381696059815e-06,
+      "loss": 0.5993,
+      "step": 2087
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.00253183747065,
+      "learning_rate": 9.721040430968393e-06,
+      "loss": 0.9952,
+      "step": 2088
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.141540856292212,
+      "learning_rate": 9.720698963003351e-06,
+      "loss": 0.8234,
+      "step": 2089
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4183175065894202,
+      "learning_rate": 9.720357292179363e-06,
+      "loss": 0.8886,
+      "step": 2090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2977021514519125,
+      "learning_rate": 9.720015418511108e-06,
+      "loss": 0.9566,
+      "step": 2091
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.333361248218109,
+      "learning_rate": 9.719673342013277e-06,
+      "loss": 0.9207,
+      "step": 2092
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.050658589854055,
+      "learning_rate": 9.719331062700572e-06,
+      "loss": 0.9764,
+      "step": 2093
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.19806617255031,
+      "learning_rate": 9.7189885805877e-06,
+      "loss": 0.7595,
+      "step": 2094
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0514413265370044,
+      "learning_rate": 9.71864589568938e-06,
+      "loss": 0.6699,
+      "step": 2095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5186195032866237,
+      "learning_rate": 9.718303008020338e-06,
+      "loss": 0.8765,
+      "step": 2096
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8983716082248756,
+      "learning_rate": 9.717959917595305e-06,
+      "loss": 0.8727,
+      "step": 2097
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0947667946492352,
+      "learning_rate": 9.71761662442903e-06,
+      "loss": 0.6736,
+      "step": 2098
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2173599177306396,
+      "learning_rate": 9.717273128536259e-06,
+      "loss": 0.6468,
+      "step": 2099
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0650716546998504,
+      "learning_rate": 9.716929429931757e-06,
+      "loss": 0.8551,
+      "step": 2100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.024666441798668,
+      "learning_rate": 9.716585528630291e-06,
+      "loss": 1.0521,
+      "step": 2101
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.179633629166904,
+      "learning_rate": 9.716241424646642e-06,
+      "loss": 0.9521,
+      "step": 2102
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.74027778059854,
+      "learning_rate": 9.715897117995594e-06,
+      "loss": 0.876,
+      "step": 2103
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8025939069726795,
+      "learning_rate": 9.715552608691944e-06,
+      "loss": 0.9435,
+      "step": 2104
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0689035267535982,
+      "learning_rate": 9.715207896750499e-06,
+      "loss": 0.8832,
+      "step": 2105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1701533936493784,
+      "learning_rate": 9.714862982186067e-06,
+      "loss": 0.6968,
+      "step": 2106
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9756876125938423,
+      "learning_rate": 9.714517865013473e-06,
+      "loss": 0.7876,
+      "step": 2107
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.746117630955115,
+      "learning_rate": 9.714172545247544e-06,
+      "loss": 0.9115,
+      "step": 2108
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9757891802186425,
+      "learning_rate": 9.713827022903124e-06,
+      "loss": 0.8026,
+      "step": 2109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0935297662205983,
+      "learning_rate": 9.71348129799506e-06,
+      "loss": 0.8937,
+      "step": 2110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1952503897268327,
+      "learning_rate": 9.713135370538204e-06,
+      "loss": 0.7752,
+      "step": 2111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5453552002469153,
+      "learning_rate": 9.712789240547426e-06,
+      "loss": 0.8886,
+      "step": 2112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2010757612858043,
+      "learning_rate": 9.712442908037599e-06,
+      "loss": 0.7043,
+      "step": 2113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2851912920231285,
+      "learning_rate": 9.712096373023603e-06,
+      "loss": 0.9003,
+      "step": 2114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.180894106053688,
+      "learning_rate": 9.711749635520333e-06,
+      "loss": 0.8875,
+      "step": 2115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.042902958270119,
+      "learning_rate": 9.711402695542689e-06,
+      "loss": 0.648,
+      "step": 2116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0425301234317774,
+      "learning_rate": 9.711055553105575e-06,
+      "loss": 0.6698,
+      "step": 2117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0966188830667063,
+      "learning_rate": 9.710708208223914e-06,
+      "loss": 0.8613,
+      "step": 2118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0309479288028958,
+      "learning_rate": 9.710360660912629e-06,
+      "loss": 0.8801,
+      "step": 2119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6592048627169733,
+      "learning_rate": 9.710012911186657e-06,
+      "loss": 0.8836,
+      "step": 2120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.45595412955799,
+      "learning_rate": 9.70966495906094e-06,
+      "loss": 1.09,
+      "step": 2121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.25011710007709,
+      "learning_rate": 9.70931680455043e-06,
+      "loss": 0.8379,
+      "step": 2122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4940703478151502,
+      "learning_rate": 9.708968447670088e-06,
+      "loss": 0.9377,
+      "step": 2123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0609765396890836,
+      "learning_rate": 9.708619888434887e-06,
+      "loss": 0.6967,
+      "step": 2124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9621176909809637,
+      "learning_rate": 9.708271126859801e-06,
+      "loss": 1.0554,
+      "step": 2125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4286238578067234,
+      "learning_rate": 9.70792216295982e-06,
+      "loss": 0.9149,
+      "step": 2126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.435667608629415,
+      "learning_rate": 9.707572996749936e-06,
+      "loss": 0.7932,
+      "step": 2127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6392217565196927,
+      "learning_rate": 9.707223628245157e-06,
+      "loss": 1.0142,
+      "step": 2128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2237289377770395,
+      "learning_rate": 9.706874057460497e-06,
+      "loss": 0.8434,
+      "step": 2129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0205018575922944,
+      "learning_rate": 9.706524284410973e-06,
+      "loss": 0.9583,
+      "step": 2130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8990796599121806,
+      "learning_rate": 9.70617430911162e-06,
+      "loss": 0.7242,
+      "step": 2131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3224148843263843,
+      "learning_rate": 9.705824131577476e-06,
+      "loss": 0.7709,
+      "step": 2132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4462856424217176,
+      "learning_rate": 9.70547375182359e-06,
+      "loss": 0.8134,
+      "step": 2133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1413265187755086,
+      "learning_rate": 9.705123169865016e-06,
+      "loss": 0.7301,
+      "step": 2134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.134262231843198,
+      "learning_rate": 9.70477238571682e-06,
+      "loss": 0.9479,
+      "step": 2135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7455452970803185,
+      "learning_rate": 9.704421399394077e-06,
+      "loss": 0.8586,
+      "step": 2136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7375753121292066,
+      "learning_rate": 9.704070210911868e-06,
+      "loss": 0.988,
+      "step": 2137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6764374134350084,
+      "learning_rate": 9.703718820285288e-06,
+      "loss": 0.9056,
+      "step": 2138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.253819351780303,
+      "learning_rate": 9.703367227529432e-06,
+      "loss": 0.697,
+      "step": 2139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.126778030216531,
+      "learning_rate": 9.703015432659413e-06,
+      "loss": 0.8304,
+      "step": 2140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8892331347656437,
+      "learning_rate": 9.702663435690344e-06,
+      "loss": 1.0464,
+      "step": 2141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3743433099987907,
+      "learning_rate": 9.702311236637357e-06,
+      "loss": 0.9293,
+      "step": 2142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8248710747690016,
+      "learning_rate": 9.701958835515582e-06,
+      "loss": 0.7832,
+      "step": 2143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9978465724102867,
+      "learning_rate": 9.701606232340165e-06,
+      "loss": 0.8896,
+      "step": 2144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4925703685219367,
+      "learning_rate": 9.701253427126257e-06,
+      "loss": 1.1093,
+      "step": 2145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.467856606327108,
+      "learning_rate": 9.700900419889018e-06,
+      "loss": 0.9016,
+      "step": 2146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4054372387861434,
+      "learning_rate": 9.700547210643618e-06,
+      "loss": 1.0142,
+      "step": 2147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9354803447779223,
+      "learning_rate": 9.700193799405238e-06,
+      "loss": 0.9076,
+      "step": 2148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4589989503117993,
+      "learning_rate": 9.699840186189061e-06,
+      "loss": 1.0246,
+      "step": 2149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1746271101532533,
+      "learning_rate": 9.699486371010285e-06,
+      "loss": 0.9631,
+      "step": 2150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.963130782777976,
+      "learning_rate": 9.69913235388411e-06,
+      "loss": 0.8674,
+      "step": 2151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2676799871059004,
+      "learning_rate": 9.698778134825755e-06,
+      "loss": 0.7911,
+      "step": 2152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5137281090889627,
+      "learning_rate": 9.69842371385044e-06,
+      "loss": 0.898,
+      "step": 2153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2449077943043276,
+      "learning_rate": 9.698069090973391e-06,
+      "loss": 0.8148,
+      "step": 2154
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.208554533320447,
+      "learning_rate": 9.697714266209853e-06,
+      "loss": 0.7651,
+      "step": 2155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.989357972426514,
+      "learning_rate": 9.697359239575069e-06,
+      "loss": 0.846,
+      "step": 2156
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.066426554434464,
+      "learning_rate": 9.697004011084298e-06,
+      "loss": 1.025,
+      "step": 2157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.071104587087026,
+      "learning_rate": 9.696648580752803e-06,
+      "loss": 0.8485,
+      "step": 2158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0639832658530066,
+      "learning_rate": 9.696292948595857e-06,
+      "loss": 0.9599,
+      "step": 2159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5599432340346846,
+      "learning_rate": 9.695937114628746e-06,
+      "loss": 0.7676,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2554229854487473,
+      "learning_rate": 9.695581078866759e-06,
+      "loss": 0.9707,
+      "step": 2161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8163896047430117,
+      "learning_rate": 9.695224841325194e-06,
+      "loss": 1.0519,
+      "step": 2162
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.926231740446241,
+      "learning_rate": 9.694868402019362e-06,
+      "loss": 0.8515,
+      "step": 2163
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.206959381586948,
+      "learning_rate": 9.694511760964578e-06,
+      "loss": 0.6113,
+      "step": 2164
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2144320921163643,
+      "learning_rate": 9.694154918176168e-06,
+      "loss": 0.9093,
+      "step": 2165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.310686780909907,
+      "learning_rate": 9.693797873669466e-06,
+      "loss": 0.8595,
+      "step": 2166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.226460722927712,
+      "learning_rate": 9.693440627459817e-06,
+      "loss": 0.6894,
+      "step": 2167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6258020232646877,
+      "learning_rate": 9.693083179562573e-06,
+      "loss": 0.994,
+      "step": 2168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.276796516014372,
+      "learning_rate": 9.69272552999309e-06,
+      "loss": 0.9464,
+      "step": 2169
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1370941678754711,
+      "learning_rate": 9.69236767876674e-06,
+      "loss": 0.739,
+      "step": 2170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.209203929780559,
+      "learning_rate": 9.692009625898902e-06,
+      "loss": 0.9153,
+      "step": 2171
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.553981411688191,
+      "learning_rate": 9.69165137140496e-06,
+      "loss": 0.8281,
+      "step": 2172
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.099995398602221,
+      "learning_rate": 9.69129291530031e-06,
+      "loss": 0.9246,
+      "step": 2173
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.277279687923281,
+      "learning_rate": 9.690934257600353e-06,
+      "loss": 0.8014,
+      "step": 2174
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8982359755274685,
+      "learning_rate": 9.690575398320504e-06,
+      "loss": 0.9989,
+      "step": 2175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.051822269563464,
+      "learning_rate": 9.690216337476188e-06,
+      "loss": 0.8475,
+      "step": 2176
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.215161965166036,
+      "learning_rate": 9.689857075082828e-06,
+      "loss": 0.6881,
+      "step": 2177
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.168874515915189,
+      "learning_rate": 9.689497611155862e-06,
+      "loss": 0.7689,
+      "step": 2178
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4782050034252627,
+      "learning_rate": 9.689137945710742e-06,
+      "loss": 0.8166,
+      "step": 2179
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0613280078588432,
+      "learning_rate": 9.688778078762922e-06,
+      "loss": 0.5873,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.598335120196963,
+      "learning_rate": 9.688418010327862e-06,
+      "loss": 0.8686,
+      "step": 2181
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.637039686646624,
+      "learning_rate": 9.688057740421042e-06,
+      "loss": 0.8042,
+      "step": 2182
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1513483488702936,
+      "learning_rate": 9.68769726905794e-06,
+      "loss": 0.8443,
+      "step": 2183
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9124912353293766,
+      "learning_rate": 9.687336596254045e-06,
+      "loss": 0.9386,
+      "step": 2184
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4041155737602167,
+      "learning_rate": 9.686975722024856e-06,
+      "loss": 0.7203,
+      "step": 2185
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9202530608283392,
+      "learning_rate": 9.686614646385883e-06,
+      "loss": 0.9483,
+      "step": 2186
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2029270601644546,
+      "learning_rate": 9.68625336935264e-06,
+      "loss": 0.9035,
+      "step": 2187
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0444176329847101,
+      "learning_rate": 9.685891890940655e-06,
+      "loss": 0.6098,
+      "step": 2188
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0497392716925065,
+      "learning_rate": 9.685530211165459e-06,
+      "loss": 0.7463,
+      "step": 2189
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.537675620458583,
+      "learning_rate": 9.685168330042594e-06,
+      "loss": 0.8238,
+      "step": 2190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.182486817977268,
+      "learning_rate": 9.68480624758761e-06,
+      "loss": 0.7994,
+      "step": 2191
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.527681290224187,
+      "learning_rate": 9.68444396381607e-06,
+      "loss": 0.9332,
+      "step": 2192
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3028089314653934,
+      "learning_rate": 9.68408147874354e-06,
+      "loss": 0.8731,
+      "step": 2193
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.167303304295371,
+      "learning_rate": 9.683718792385595e-06,
+      "loss": 0.7097,
+      "step": 2194
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7932231220741293,
+      "learning_rate": 9.683355904757822e-06,
+      "loss": 0.8973,
+      "step": 2195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.174445973723725,
+      "learning_rate": 9.682992815875818e-06,
+      "loss": 0.8552,
+      "step": 2196
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2960817213148235,
+      "learning_rate": 9.682629525755182e-06,
+      "loss": 0.9164,
+      "step": 2197
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1254220828440684,
+      "learning_rate": 9.682266034411527e-06,
+      "loss": 0.8487,
+      "step": 2198
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.202119962193055,
+      "learning_rate": 9.681902341860471e-06,
+      "loss": 0.7853,
+      "step": 2199
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.537932737657711,
+      "learning_rate": 9.681538448117647e-06,
+      "loss": 1.0256,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1256826540872473,
+      "learning_rate": 9.681174353198687e-06,
+      "loss": 1.014,
+      "step": 2201
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.741146782368245,
+      "learning_rate": 9.68081005711924e-06,
+      "loss": 0.6442,
+      "step": 2202
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3191593519969556,
+      "learning_rate": 9.68044555989496e-06,
+      "loss": 1.0181,
+      "step": 2203
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.235726033945763,
+      "learning_rate": 9.680080861541511e-06,
+      "loss": 0.8658,
+      "step": 2204
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.616926292674678,
+      "learning_rate": 9.679715962074566e-06,
+      "loss": 0.9209,
+      "step": 2205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.076969962108019,
+      "learning_rate": 9.6793508615098e-06,
+      "loss": 0.7401,
+      "step": 2206
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.771910009422206,
+      "learning_rate": 9.678985559862909e-06,
+      "loss": 0.8572,
+      "step": 2207
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8383448634122352,
+      "learning_rate": 9.678620057149586e-06,
+      "loss": 0.8704,
+      "step": 2208
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0051363407912537,
+      "learning_rate": 9.678254353385538e-06,
+      "loss": 1.0241,
+      "step": 2209
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3543291264213855,
+      "learning_rate": 9.677888448586485e-06,
+      "loss": 0.7647,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9917316737955009,
+      "learning_rate": 9.677522342768145e-06,
+      "loss": 0.7366,
+      "step": 2211
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1450614753621475,
+      "learning_rate": 9.677156035946253e-06,
+      "loss": 1.1081,
+      "step": 2212
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8818715797168053,
+      "learning_rate": 9.67678952813655e-06,
+      "loss": 0.9866,
+      "step": 2213
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3026390732979936,
+      "learning_rate": 9.676422819354785e-06,
+      "loss": 1.0702,
+      "step": 2214
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.150743256893927,
+      "learning_rate": 9.676055909616716e-06,
+      "loss": 1.0153,
+      "step": 2215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2281711907012023,
+      "learning_rate": 9.675688798938112e-06,
+      "loss": 0.5779,
+      "step": 2216
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1130440344888193,
+      "learning_rate": 9.675321487334744e-06,
+      "loss": 0.8472,
+      "step": 2217
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.077143476719788,
+      "learning_rate": 9.674953974822402e-06,
+      "loss": 0.9115,
+      "step": 2218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9760178238469996,
+      "learning_rate": 9.674586261416874e-06,
+      "loss": 0.9598,
+      "step": 2219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.744700411276636,
+      "learning_rate": 9.674218347133965e-06,
+      "loss": 0.7412,
+      "step": 2220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.751426788922555,
+      "learning_rate": 9.673850231989484e-06,
+      "loss": 0.8189,
+      "step": 2221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3007769191076817,
+      "learning_rate": 9.673481915999249e-06,
+      "loss": 0.663,
+      "step": 2222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7172531819141597,
+      "learning_rate": 9.673113399179085e-06,
+      "loss": 0.818,
+      "step": 2223
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1646779427828724,
+      "learning_rate": 9.672744681544834e-06,
+      "loss": 0.6706,
+      "step": 2224
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1608445332247896,
+      "learning_rate": 9.672375763112335e-06,
+      "loss": 0.8685,
+      "step": 2225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0485111173782684,
+      "learning_rate": 9.672006643897444e-06,
+      "loss": 1.0189,
+      "step": 2226
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3306431647883918,
+      "learning_rate": 9.671637323916023e-06,
+      "loss": 0.6385,
+      "step": 2227
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0558610384244767,
+      "learning_rate": 9.671267803183942e-06,
+      "loss": 0.9761,
+      "step": 2228
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3235406620830807,
+      "learning_rate": 9.670898081717079e-06,
+      "loss": 0.848,
+      "step": 2229
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.254706182084915,
+      "learning_rate": 9.670528159531325e-06,
+      "loss": 0.881,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9713053437975494,
+      "learning_rate": 9.670158036642572e-06,
+      "loss": 0.8675,
+      "step": 2231
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1352445050327329,
+      "learning_rate": 9.669787713066728e-06,
+      "loss": 0.7619,
+      "step": 2232
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0956441944715176,
+      "learning_rate": 9.669417188819704e-06,
+      "loss": 0.948,
+      "step": 2233
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.221138703567473,
+      "learning_rate": 9.669046463917427e-06,
+      "loss": 0.7785,
+      "step": 2234
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1687906947913866,
+      "learning_rate": 9.668675538375823e-06,
+      "loss": 1.1289,
+      "step": 2235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8866731027238526,
+      "learning_rate": 9.668304412210835e-06,
+      "loss": 0.9242,
+      "step": 2236
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.63211005855444,
+      "learning_rate": 9.667933085438407e-06,
+      "loss": 1.0141,
+      "step": 2237
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5164570792138843,
+      "learning_rate": 9.6675615580745e-06,
+      "loss": 0.8859,
+      "step": 2238
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1866187377588917,
+      "learning_rate": 9.667189830135078e-06,
+      "loss": 0.638,
+      "step": 2239
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.902248355397232,
+      "learning_rate": 9.666817901636115e-06,
+      "loss": 0.8145,
+      "step": 2240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.300469022607953,
+      "learning_rate": 9.666445772593592e-06,
+      "loss": 0.8121,
+      "step": 2241
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.401179143133987,
+      "learning_rate": 9.666073443023502e-06,
+      "loss": 0.8687,
+      "step": 2242
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.467385565716672,
+      "learning_rate": 9.665700912941844e-06,
+      "loss": 0.8018,
+      "step": 2243
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7688153715653776,
+      "learning_rate": 9.665328182364627e-06,
+      "loss": 0.8067,
+      "step": 2244
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6637999446917426,
+      "learning_rate": 9.664955251307867e-06,
+      "loss": 1.0694,
+      "step": 2245
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.692420820899958,
+      "learning_rate": 9.66458211978759e-06,
+      "loss": 0.8708,
+      "step": 2246
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2360337588203087,
+      "learning_rate": 9.664208787819833e-06,
+      "loss": 0.9001,
+      "step": 2247
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4285847578758344,
+      "learning_rate": 9.663835255420635e-06,
+      "loss": 0.9282,
+      "step": 2248
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3296234811890617,
+      "learning_rate": 9.663461522606049e-06,
+      "loss": 0.7751,
+      "step": 2249
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4595428778405566,
+      "learning_rate": 9.663087589392135e-06,
+      "loss": 0.8729,
+      "step": 2250
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.882118630216425,
+      "learning_rate": 9.662713455794963e-06,
+      "loss": 0.946,
+      "step": 2251
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.316478253999047,
+      "learning_rate": 9.66233912183061e-06,
+      "loss": 0.8583,
+      "step": 2252
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5182253784018944,
+      "learning_rate": 9.66196458751516e-06,
+      "loss": 1.1074,
+      "step": 2253
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3132070985811595,
+      "learning_rate": 9.66158985286471e-06,
+      "loss": 0.8997,
+      "step": 2254
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.978046654816286,
+      "learning_rate": 9.661214917895361e-06,
+      "loss": 0.7838,
+      "step": 2255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.314902658094912,
+      "learning_rate": 9.660839782623226e-06,
+      "loss": 0.9004,
+      "step": 2256
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6005456233514437,
+      "learning_rate": 9.660464447064427e-06,
+      "loss": 0.8318,
+      "step": 2257
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8646631862884524,
+      "learning_rate": 9.660088911235088e-06,
+      "loss": 1.054,
+      "step": 2258
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3170915867755895,
+      "learning_rate": 9.659713175151352e-06,
+      "loss": 0.9453,
+      "step": 2259
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2374092320753793,
+      "learning_rate": 9.659337238829364e-06,
+      "loss": 0.8956,
+      "step": 2260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.414704297582489,
+      "learning_rate": 9.658961102285276e-06,
+      "loss": 0.8416,
+      "step": 2261
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.332590167368533,
+      "learning_rate": 9.658584765535254e-06,
+      "loss": 0.8914,
+      "step": 2262
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.508766373786291,
+      "learning_rate": 9.65820822859547e-06,
+      "loss": 0.9434,
+      "step": 2263
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2617705108776325,
+      "learning_rate": 9.657831491482103e-06,
+      "loss": 0.9381,
+      "step": 2264
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.00458549917529,
+      "learning_rate": 9.657454554211344e-06,
+      "loss": 0.9244,
+      "step": 2265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.085641244962591,
+      "learning_rate": 9.657077416799392e-06,
+      "loss": 0.6354,
+      "step": 2266
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2292674455902852,
+      "learning_rate": 9.656700079262448e-06,
+      "loss": 0.7296,
+      "step": 2267
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2916458701510707,
+      "learning_rate": 9.656322541616734e-06,
+      "loss": 0.6482,
+      "step": 2268
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9249975584434997,
+      "learning_rate": 9.655944803878467e-06,
+      "loss": 0.8903,
+      "step": 2269
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6847830488712967,
+      "learning_rate": 9.655566866063885e-06,
+      "loss": 0.7956,
+      "step": 2270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.857572811840922,
+      "learning_rate": 9.655188728189225e-06,
+      "loss": 0.887,
+      "step": 2271
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9982578626932197,
+      "learning_rate": 9.654810390270737e-06,
+      "loss": 0.7686,
+      "step": 2272
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.252593143115697,
+      "learning_rate": 9.654431852324682e-06,
+      "loss": 0.825,
+      "step": 2273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2877569525233694,
+      "learning_rate": 9.654053114367321e-06,
+      "loss": 0.8033,
+      "step": 2274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2537569759876126,
+      "learning_rate": 9.653674176414936e-06,
+      "loss": 0.8981,
+      "step": 2275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3239411909866,
+      "learning_rate": 9.653295038483806e-06,
+      "loss": 0.9123,
+      "step": 2276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.412676613331747,
+      "learning_rate": 9.652915700590224e-06,
+      "loss": 0.8721,
+      "step": 2277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3121561037460048,
+      "learning_rate": 9.652536162750492e-06,
+      "loss": 0.6996,
+      "step": 2278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.928036862246456,
+      "learning_rate": 9.65215642498092e-06,
+      "loss": 0.8603,
+      "step": 2279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5397410763094648,
+      "learning_rate": 9.651776487297823e-06,
+      "loss": 0.9208,
+      "step": 2280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8220968196082055,
+      "learning_rate": 9.651396349717533e-06,
+      "loss": 0.9231,
+      "step": 2281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.379737172034971,
+      "learning_rate": 9.651016012256382e-06,
+      "loss": 0.7098,
+      "step": 2282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.141425535226531,
+      "learning_rate": 9.650635474930716e-06,
+      "loss": 0.6505,
+      "step": 2283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2319959744400264,
+      "learning_rate": 9.650254737756883e-06,
+      "loss": 0.6865,
+      "step": 2284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.506072150851678,
+      "learning_rate": 9.64987380075125e-06,
+      "loss": 1.0254,
+      "step": 2285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.43682919314133,
+      "learning_rate": 9.649492663930182e-06,
+      "loss": 0.8505,
+      "step": 2286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4116408164600727,
+      "learning_rate": 9.64911132731006e-06,
+      "loss": 1.0052,
+      "step": 2287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.701376272912519,
+      "learning_rate": 9.648729790907271e-06,
+      "loss": 0.9814,
+      "step": 2288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.159937359541095,
+      "learning_rate": 9.648348054738208e-06,
+      "loss": 0.704,
+      "step": 2289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.029865454446369,
+      "learning_rate": 9.647966118819278e-06,
+      "loss": 0.922,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.46879028818401,
+      "learning_rate": 9.647583983166892e-06,
+      "loss": 0.6443,
+      "step": 2291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0226153499730564,
+      "learning_rate": 9.647201647797472e-06,
+      "loss": 0.7208,
+      "step": 2292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3573262855945036,
+      "learning_rate": 9.646819112727445e-06,
+      "loss": 0.9447,
+      "step": 2293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.210649414907495,
+      "learning_rate": 9.646436377973253e-06,
+      "loss": 0.9315,
+      "step": 2294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.421029042402498,
+      "learning_rate": 9.646053443551344e-06,
+      "loss": 0.995,
+      "step": 2295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3108982159062927,
+      "learning_rate": 9.64567030947817e-06,
+      "loss": 0.9328,
+      "step": 2296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.017615293384587,
+      "learning_rate": 9.645286975770194e-06,
+      "loss": 0.894,
+      "step": 2297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0731700473758976,
+      "learning_rate": 9.644903442443894e-06,
+      "loss": 0.9093,
+      "step": 2298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.304572729439097,
+      "learning_rate": 9.644519709515746e-06,
+      "loss": 0.7368,
+      "step": 2299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.858969636423153,
+      "learning_rate": 9.644135777002244e-06,
+      "loss": 0.7668,
+      "step": 2300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.059037200743867,
+      "learning_rate": 9.643751644919883e-06,
+      "loss": 0.8368,
+      "step": 2301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.353655697434197,
+      "learning_rate": 9.643367313285172e-06,
+      "loss": 0.928,
+      "step": 2302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.902244779575429,
+      "learning_rate": 9.642982782114628e-06,
+      "loss": 0.7831,
+      "step": 2303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9926991870923365,
+      "learning_rate": 9.642598051424772e-06,
+      "loss": 0.8521,
+      "step": 2304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3106901393579582,
+      "learning_rate": 9.642213121232137e-06,
+      "loss": 0.767,
+      "step": 2305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2674895884734465,
+      "learning_rate": 9.641827991553267e-06,
+      "loss": 0.6624,
+      "step": 2306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2457630302983618,
+      "learning_rate": 9.64144266240471e-06,
+      "loss": 1.0487,
+      "step": 2307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5856053750334778,
+      "learning_rate": 9.641057133803024e-06,
+      "loss": 1.0776,
+      "step": 2308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1165431240498482,
+      "learning_rate": 9.640671405764777e-06,
+      "loss": 0.6348,
+      "step": 2309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9440599809648775,
+      "learning_rate": 9.640285478306546e-06,
+      "loss": 0.8047,
+      "step": 2310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.386106390302779,
+      "learning_rate": 9.639899351444911e-06,
+      "loss": 1.073,
+      "step": 2311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3921496951609393,
+      "learning_rate": 9.63951302519647e-06,
+      "loss": 1.0022,
+      "step": 2312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.098985855733678,
+      "learning_rate": 9.639126499577818e-06,
+      "loss": 0.6975,
+      "step": 2313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9452276473320589,
+      "learning_rate": 9.638739774605572e-06,
+      "loss": 0.7523,
+      "step": 2314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.073292790639716,
+      "learning_rate": 9.638352850296344e-06,
+      "loss": 0.6236,
+      "step": 2315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.265412687729197,
+      "learning_rate": 9.637965726666766e-06,
+      "loss": 0.9956,
+      "step": 2316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6309723065935717,
+      "learning_rate": 9.63757840373347e-06,
+      "loss": 0.8039,
+      "step": 2317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.088214967603786,
+      "learning_rate": 9.637190881513103e-06,
+      "loss": 0.7995,
+      "step": 2318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.984165167522704,
+      "learning_rate": 9.636803160022314e-06,
+      "loss": 0.9321,
+      "step": 2319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.337616792045294,
+      "learning_rate": 9.63641523927777e-06,
+      "loss": 0.7607,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.645905132012104,
+      "learning_rate": 9.636027119296133e-06,
+      "loss": 0.9606,
+      "step": 2321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.0593531949318455,
+      "learning_rate": 9.63563880009409e-06,
+      "loss": 0.7291,
+      "step": 2322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.479059411601704,
+      "learning_rate": 9.63525028168832e-06,
+      "loss": 0.8724,
+      "step": 2323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.489259017092427,
+      "learning_rate": 9.634861564095525e-06,
+      "loss": 0.6224,
+      "step": 2324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0879407954344127,
+      "learning_rate": 9.634472647332405e-06,
+      "loss": 0.752,
+      "step": 2325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.604358164874514,
+      "learning_rate": 9.634083531415672e-06,
+      "loss": 0.8628,
+      "step": 2326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1383889638941962,
+      "learning_rate": 9.633694216362051e-06,
+      "loss": 0.6635,
+      "step": 2327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0470770375910012,
+      "learning_rate": 9.633304702188271e-06,
+      "loss": 0.906,
+      "step": 2328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.752213973248303,
+      "learning_rate": 9.632914988911066e-06,
+      "loss": 0.7905,
+      "step": 2329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.703377329757559,
+      "learning_rate": 9.632525076547187e-06,
+      "loss": 0.9114,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.168115612900132,
+      "learning_rate": 9.632134965113389e-06,
+      "loss": 0.6387,
+      "step": 2331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.164314734562098,
+      "learning_rate": 9.631744654626435e-06,
+      "loss": 0.8511,
+      "step": 2332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.944840393485437,
+      "learning_rate": 9.631354145103098e-06,
+      "loss": 0.7604,
+      "step": 2333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.283902205221471,
+      "learning_rate": 9.63096343656016e-06,
+      "loss": 0.9327,
+      "step": 2334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.91044077012312,
+      "learning_rate": 9.630572529014406e-06,
+      "loss": 0.7797,
+      "step": 2335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6237953530681355,
+      "learning_rate": 9.630181422482639e-06,
+      "loss": 1.0155,
+      "step": 2336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.248845474783834,
+      "learning_rate": 9.629790116981665e-06,
+      "loss": 0.966,
+      "step": 2337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.103199128158375,
+      "learning_rate": 9.629398612528299e-06,
+      "loss": 0.7927,
+      "step": 2338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.707989783313987,
+      "learning_rate": 9.629006909139363e-06,
+      "loss": 0.8382,
+      "step": 2339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.793883032124016,
+      "learning_rate": 9.628615006831692e-06,
+      "loss": 0.9738,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8250461026057847,
+      "learning_rate": 9.628222905622124e-06,
+      "loss": 0.7426,
+      "step": 2341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.233481150039175,
+      "learning_rate": 9.627830605527513e-06,
+      "loss": 0.8755,
+      "step": 2342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.049087522407139,
+      "learning_rate": 9.627438106564711e-06,
+      "loss": 0.9916,
+      "step": 2343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4272500878117413,
+      "learning_rate": 9.62704540875059e-06,
+      "loss": 0.7956,
+      "step": 2344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.544762168126381,
+      "learning_rate": 9.626652512102021e-06,
+      "loss": 0.9042,
+      "step": 2345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2407967164624594,
+      "learning_rate": 9.62625941663589e-06,
+      "loss": 0.7572,
+      "step": 2346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.132746792715582,
+      "learning_rate": 9.625866122369089e-06,
+      "loss": 0.9318,
+      "step": 2347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1508199550615723,
+      "learning_rate": 9.62547262931852e-06,
+      "loss": 1.0321,
+      "step": 2348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0913491088893283,
+      "learning_rate": 9.625078937501089e-06,
+      "loss": 0.8269,
+      "step": 2349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.725605979438532,
+      "learning_rate": 9.624685046933715e-06,
+      "loss": 0.99,
+      "step": 2350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0067628406802043,
+      "learning_rate": 9.624290957633325e-06,
+      "loss": 0.9439,
+      "step": 2351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.947969496995403,
+      "learning_rate": 9.623896669616855e-06,
+      "loss": 0.8787,
+      "step": 2352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0843350619209184,
+      "learning_rate": 9.623502182901245e-06,
+      "loss": 0.8124,
+      "step": 2353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3466177700286064,
+      "learning_rate": 9.62310749750345e-06,
+      "loss": 0.8421,
+      "step": 2354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.310369913730796,
+      "learning_rate": 9.62271261344043e-06,
+      "loss": 0.8541,
+      "step": 2355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1684998886688884,
+      "learning_rate": 9.622317530729153e-06,
+      "loss": 0.7265,
+      "step": 2356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2782679496949685,
+      "learning_rate": 9.621922249386598e-06,
+      "loss": 0.7776,
+      "step": 2357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4225868284851853,
+      "learning_rate": 9.621526769429749e-06,
+      "loss": 1.0136,
+      "step": 2358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.767631360955926,
+      "learning_rate": 9.621131090875603e-06,
+      "loss": 0.7916,
+      "step": 2359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0959770394336332,
+      "learning_rate": 9.620735213741163e-06,
+      "loss": 0.932,
+      "step": 2360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1228374543027475,
+      "learning_rate": 9.62033913804344e-06,
+      "loss": 1.003,
+      "step": 2361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5457583895273337,
+      "learning_rate": 9.619942863799454e-06,
+      "loss": 0.9536,
+      "step": 2362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.052314910348969,
+      "learning_rate": 9.619546391026234e-06,
+      "loss": 0.9353,
+      "step": 2363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1066705572467708,
+      "learning_rate": 9.619149719740817e-06,
+      "loss": 0.828,
+      "step": 2364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4807950209563687,
+      "learning_rate": 9.618752849960248e-06,
+      "loss": 0.933,
+      "step": 2365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8384725393125876,
+      "learning_rate": 9.618355781701584e-06,
+      "loss": 0.9949,
+      "step": 2366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.672891509618916,
+      "learning_rate": 9.617958514981885e-06,
+      "loss": 0.9594,
+      "step": 2367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8018572128998733,
+      "learning_rate": 9.617561049818225e-06,
+      "loss": 0.7784,
+      "step": 2368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1990446918409665,
+      "learning_rate": 9.617163386227683e-06,
+      "loss": 0.9834,
+      "step": 2369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.559305445545423,
+      "learning_rate": 9.616765524227348e-06,
+      "loss": 0.7603,
+      "step": 2370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.051132102175693,
+      "learning_rate": 9.616367463834313e-06,
+      "loss": 0.9484,
+      "step": 2371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8249044086740909,
+      "learning_rate": 9.61596920506569e-06,
+      "loss": 1.0195,
+      "step": 2372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3476539028014596,
+      "learning_rate": 9.61557074793859e-06,
+      "loss": 0.9766,
+      "step": 2373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.124470740437923,
+      "learning_rate": 9.615172092470134e-06,
+      "loss": 0.7513,
+      "step": 2374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.931210726947014,
+      "learning_rate": 9.614773238677455e-06,
+      "loss": 0.9311,
+      "step": 2375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5456870643537535,
+      "learning_rate": 9.614374186577693e-06,
+      "loss": 0.9365,
+      "step": 2376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.872866093138009,
+      "learning_rate": 9.613974936187993e-06,
+      "loss": 0.8672,
+      "step": 2377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0622196349466146,
+      "learning_rate": 9.613575487525517e-06,
+      "loss": 0.9645,
+      "step": 2378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.138215856358523,
+      "learning_rate": 9.613175840607428e-06,
+      "loss": 0.7323,
+      "step": 2379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3725833586381142,
+      "learning_rate": 9.612775995450896e-06,
+      "loss": 0.6545,
+      "step": 2380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4562719724827486,
+      "learning_rate": 9.612375952073108e-06,
+      "loss": 0.8189,
+      "step": 2381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0633507715517787,
+      "learning_rate": 9.611975710491252e-06,
+      "loss": 0.7623,
+      "step": 2382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.98310090761249,
+      "learning_rate": 9.611575270722529e-06,
+      "loss": 0.6475,
+      "step": 2383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3259431487758953,
+      "learning_rate": 9.611174632784147e-06,
+      "loss": 0.9416,
+      "step": 2384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.098400540215401,
+      "learning_rate": 9.61077379669332e-06,
+      "loss": 0.8579,
+      "step": 2385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.122965562063497,
+      "learning_rate": 9.610372762467275e-06,
+      "loss": 0.7102,
+      "step": 2386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3578067373496263,
+      "learning_rate": 9.609971530123243e-06,
+      "loss": 1.0233,
+      "step": 2387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6735129695800337,
+      "learning_rate": 9.60957009967847e-06,
+      "loss": 0.9864,
+      "step": 2388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0384373146235477,
+      "learning_rate": 9.609168471150202e-06,
+      "loss": 0.9716,
+      "step": 2389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.534728889990695,
+      "learning_rate": 9.6087666445557e-06,
+      "loss": 0.8018,
+      "step": 2390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9182343226921568,
+      "learning_rate": 9.60836461991223e-06,
+      "loss": 0.8956,
+      "step": 2391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.312365448049255,
+      "learning_rate": 9.607962397237071e-06,
+      "loss": 1.0355,
+      "step": 2392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5281304833343625,
+      "learning_rate": 9.607559976547504e-06,
+      "loss": 0.9184,
+      "step": 2393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9358864060069856,
+      "learning_rate": 9.607157357860823e-06,
+      "loss": 0.8624,
+      "step": 2394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3769543670021713,
+      "learning_rate": 9.606754541194329e-06,
+      "loss": 0.8253,
+      "step": 2395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4186184948325375,
+      "learning_rate": 9.606351526565336e-06,
+      "loss": 0.8127,
+      "step": 2396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.991544508512453,
+      "learning_rate": 9.605948313991156e-06,
+      "loss": 0.9136,
+      "step": 2397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.250154324012038,
+      "learning_rate": 9.60554490348912e-06,
+      "loss": 0.7718,
+      "step": 2398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1240911074701956,
+      "learning_rate": 9.605141295076561e-06,
+      "loss": 0.6588,
+      "step": 2399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.095660862296258,
+      "learning_rate": 9.604737488770826e-06,
+      "loss": 0.9902,
+      "step": 2400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6105794953209607,
+      "learning_rate": 9.604333484589266e-06,
+      "loss": 0.8684,
+      "step": 2401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.222784024116072,
+      "learning_rate": 9.60392928254924e-06,
+      "loss": 0.6938,
+      "step": 2402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.453174698463195,
+      "learning_rate": 9.60352488266812e-06,
+      "loss": 0.9155,
+      "step": 2403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7575994069373353,
+      "learning_rate": 9.603120284963284e-06,
+      "loss": 0.8831,
+      "step": 2404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1771650955305428,
+      "learning_rate": 9.602715489452117e-06,
+      "loss": 0.6857,
+      "step": 2405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.283409934760845,
+      "learning_rate": 9.602310496152014e-06,
+      "loss": 0.6984,
+      "step": 2406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.580185302259695,
+      "learning_rate": 9.60190530508038e-06,
+      "loss": 0.8063,
+      "step": 2407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0726064910377326,
+      "learning_rate": 9.601499916254626e-06,
+      "loss": 0.8382,
+      "step": 2408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.253698212639115,
+      "learning_rate": 9.601094329692173e-06,
+      "loss": 0.9834,
+      "step": 2409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.379259099895284,
+      "learning_rate": 9.600688545410449e-06,
+      "loss": 1.1021,
+      "step": 2410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4738254653775447,
+      "learning_rate": 9.60028256342689e-06,
+      "loss": 0.9291,
+      "step": 2411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5130218531555,
+      "learning_rate": 9.599876383758946e-06,
+      "loss": 0.7299,
+      "step": 2412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.995385479896558,
+      "learning_rate": 9.599470006424069e-06,
+      "loss": 0.8876,
+      "step": 2413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1563094790954964,
+      "learning_rate": 9.599063431439721e-06,
+      "loss": 0.9734,
+      "step": 2414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6926237597438005,
+      "learning_rate": 9.598656658823378e-06,
+      "loss": 0.8676,
+      "step": 2415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.984978013597154,
+      "learning_rate": 9.598249688592514e-06,
+      "loss": 0.9329,
+      "step": 2416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1253076728944693,
+      "learning_rate": 9.597842520764619e-06,
+      "loss": 0.8873,
+      "step": 2417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.137589131212165,
+      "learning_rate": 9.597435155357192e-06,
+      "loss": 0.9527,
+      "step": 2418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.610095151484005,
+      "learning_rate": 9.597027592387739e-06,
+      "loss": 0.9159,
+      "step": 2419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.239792175499802,
+      "learning_rate": 9.59661983187377e-06,
+      "loss": 0.9418,
+      "step": 2420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.873718290031374,
+      "learning_rate": 9.59621187383281e-06,
+      "loss": 0.8288,
+      "step": 2421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.162309590375917,
+      "learning_rate": 9.595803718282391e-06,
+      "loss": 0.7457,
+      "step": 2422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.358448296890817,
+      "learning_rate": 9.59539536524005e-06,
+      "loss": 0.6975,
+      "step": 2423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.113603579057796,
+      "learning_rate": 9.594986814723335e-06,
+      "loss": 1.0113,
+      "step": 2424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7204717314551792,
+      "learning_rate": 9.594578066749804e-06,
+      "loss": 0.7724,
+      "step": 2425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8659925905530597,
+      "learning_rate": 9.59416912133702e-06,
+      "loss": 0.8097,
+      "step": 2426
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3942250816123796,
+      "learning_rate": 9.593759978502556e-06,
+      "loss": 0.9817,
+      "step": 2427
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9229086137604092,
+      "learning_rate": 9.593350638263998e-06,
+      "loss": 0.737,
+      "step": 2428
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6058522704153444,
+      "learning_rate": 9.59294110063893e-06,
+      "loss": 0.895,
+      "step": 2429
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.007956011504491,
+      "learning_rate": 9.592531365644957e-06,
+      "loss": 0.8278,
+      "step": 2430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5303383734083322,
+      "learning_rate": 9.592121433299682e-06,
+      "loss": 0.7941,
+      "step": 2431
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.537828964513514,
+      "learning_rate": 9.59171130362072e-06,
+      "loss": 0.9291,
+      "step": 2432
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.632297383828382,
+      "learning_rate": 9.591300976625697e-06,
+      "loss": 0.8507,
+      "step": 2433
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1295505122323632,
+      "learning_rate": 9.590890452332249e-06,
+      "loss": 0.899,
+      "step": 2434
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.526893934832527,
+      "learning_rate": 9.59047973075801e-06,
+      "loss": 0.8902,
+      "step": 2435
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 27.902192329028008,
+      "learning_rate": 9.590068811920637e-06,
+      "loss": 0.7926,
+      "step": 2436
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0092741986290568,
+      "learning_rate": 9.589657695837781e-06,
+      "loss": 0.943,
+      "step": 2437
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.697217577339657,
+      "learning_rate": 9.589246382527113e-06,
+      "loss": 0.954,
+      "step": 2438
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.198046662960297,
+      "learning_rate": 9.588834872006308e-06,
+      "loss": 0.7935,
+      "step": 2439
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.273890134289108,
+      "learning_rate": 9.588423164293049e-06,
+      "loss": 0.8606,
+      "step": 2440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.098305652730905,
+      "learning_rate": 9.588011259405028e-06,
+      "loss": 0.7714,
+      "step": 2441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5759729326208083,
+      "learning_rate": 9.587599157359943e-06,
+      "loss": 0.8286,
+      "step": 2442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.971200904852272,
+      "learning_rate": 9.587186858175507e-06,
+      "loss": 0.6669,
+      "step": 2443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9791754412977707,
+      "learning_rate": 9.586774361869436e-06,
+      "loss": 0.8529,
+      "step": 2444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9358646902072136,
+      "learning_rate": 9.586361668459452e-06,
+      "loss": 0.9134,
+      "step": 2445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9474118820611859,
+      "learning_rate": 9.585948777963296e-06,
+      "loss": 0.9185,
+      "step": 2446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.489395048015992,
+      "learning_rate": 9.585535690398706e-06,
+      "loss": 0.8722,
+      "step": 2447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.470988385876614,
+      "learning_rate": 9.585122405783435e-06,
+      "loss": 0.9359,
+      "step": 2448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2205670368536166,
+      "learning_rate": 9.584708924135245e-06,
+      "loss": 0.853,
+      "step": 2449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.061129392825398,
+      "learning_rate": 9.584295245471898e-06,
+      "loss": 0.8968,
+      "step": 2450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1392009521515005,
+      "learning_rate": 9.583881369811176e-06,
+      "loss": 0.8956,
+      "step": 2451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3778949702561603,
+      "learning_rate": 9.583467297170863e-06,
+      "loss": 1.0325,
+      "step": 2452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0592947646026625,
+      "learning_rate": 9.583053027568754e-06,
+      "loss": 0.9152,
+      "step": 2453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1059754684833867,
+      "learning_rate": 9.582638561022646e-06,
+      "loss": 0.7087,
+      "step": 2454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.831236483851464,
+      "learning_rate": 9.582223897550356e-06,
+      "loss": 0.6648,
+      "step": 2455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4363096645439173,
+      "learning_rate": 9.581809037169698e-06,
+      "loss": 0.9434,
+      "step": 2456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4764042687510415,
+      "learning_rate": 9.581393979898502e-06,
+      "loss": 0.9376,
+      "step": 2457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1359497962354066,
+      "learning_rate": 9.580978725754604e-06,
+      "loss": 0.7997,
+      "step": 2458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.630016165535835,
+      "learning_rate": 9.580563274755848e-06,
+      "loss": 0.9363,
+      "step": 2459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6810784746441425,
+      "learning_rate": 9.580147626920084e-06,
+      "loss": 0.9086,
+      "step": 2460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.256316544965505,
+      "learning_rate": 9.579731782265178e-06,
+      "loss": 0.9411,
+      "step": 2461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5040800909603664,
+      "learning_rate": 9.579315740808998e-06,
+      "loss": 0.9322,
+      "step": 2462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4840848425700486,
+      "learning_rate": 9.578899502569421e-06,
+      "loss": 0.8736,
+      "step": 2463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0634862763429846,
+      "learning_rate": 9.578483067564335e-06,
+      "loss": 0.7472,
+      "step": 2464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0639243029587733,
+      "learning_rate": 9.578066435811636e-06,
+      "loss": 0.9228,
+      "step": 2465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7947937466179047,
+      "learning_rate": 9.577649607329225e-06,
+      "loss": 1.1372,
+      "step": 2466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0815384499389815,
+      "learning_rate": 9.577232582135014e-06,
+      "loss": 0.7083,
+      "step": 2467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7928064901559857,
+      "learning_rate": 9.576815360246926e-06,
+      "loss": 0.9917,
+      "step": 2468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.333642852924852,
+      "learning_rate": 9.576397941682891e-06,
+      "loss": 0.6121,
+      "step": 2469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7315014632906873,
+      "learning_rate": 9.575980326460841e-06,
+      "loss": 1.0373,
+      "step": 2470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3426971782637127,
+      "learning_rate": 9.575562514598727e-06,
+      "loss": 0.7427,
+      "step": 2471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.135360495363509,
+      "learning_rate": 9.575144506114501e-06,
+      "loss": 0.6606,
+      "step": 2472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.142203517394497,
+      "learning_rate": 9.574726301026127e-06,
+      "loss": 0.6641,
+      "step": 2473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1607832999273517,
+      "learning_rate": 9.574307899351574e-06,
+      "loss": 0.6933,
+      "step": 2474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5241743595511337,
+      "learning_rate": 9.573889301108823e-06,
+      "loss": 0.7574,
+      "step": 2475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3252152518852003,
+      "learning_rate": 9.573470506315864e-06,
+      "loss": 0.8851,
+      "step": 2476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.008040805399465,
+      "learning_rate": 9.57305151499069e-06,
+      "loss": 0.9389,
+      "step": 2477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.861578387650215,
+      "learning_rate": 9.572632327151309e-06,
+      "loss": 0.9002,
+      "step": 2478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3094691822057056,
+      "learning_rate": 9.572212942815734e-06,
+      "loss": 0.8823,
+      "step": 2479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.459590782371774,
+      "learning_rate": 9.571793362001983e-06,
+      "loss": 0.6889,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.822226910760496,
+      "learning_rate": 9.57137358472809e-06,
+      "loss": 1.0306,
+      "step": 2481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6006345923507954,
+      "learning_rate": 9.570953611012092e-06,
+      "loss": 0.9784,
+      "step": 2482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2838451481195077,
+      "learning_rate": 9.570533440872041e-06,
+      "loss": 0.9012,
+      "step": 2483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.775434865820244,
+      "learning_rate": 9.570113074325986e-06,
+      "loss": 0.8436,
+      "step": 2484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4111544736884727,
+      "learning_rate": 9.569692511391995e-06,
+      "loss": 1.0728,
+      "step": 2485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.490828824934825,
+      "learning_rate": 9.569271752088139e-06,
+      "loss": 0.7901,
+      "step": 2486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9037223820395672,
+      "learning_rate": 9.5688507964325e-06,
+      "loss": 0.8157,
+      "step": 2487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.130941479848446,
+      "learning_rate": 9.568429644443167e-06,
+      "loss": 0.6951,
+      "step": 2488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8254830069073837,
+      "learning_rate": 9.568008296138238e-06,
+      "loss": 0.8193,
+      "step": 2489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.535009965332523,
+      "learning_rate": 9.567586751535818e-06,
+      "loss": 0.8805,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.144129818423914,
+      "learning_rate": 9.567165010654022e-06,
+      "loss": 0.9422,
+      "step": 2491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8704892784021787,
+      "learning_rate": 9.566743073510976e-06,
+      "loss": 1.0822,
+      "step": 2492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1355182936346895,
+      "learning_rate": 9.566320940124809e-06,
+      "loss": 0.6473,
+      "step": 2493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.383056220681265,
+      "learning_rate": 9.565898610513661e-06,
+      "loss": 0.9758,
+      "step": 2494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3994554054952575,
+      "learning_rate": 9.565476084695681e-06,
+      "loss": 0.7568,
+      "step": 2495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.030990087037302,
+      "learning_rate": 9.565053362689026e-06,
+      "loss": 0.7477,
+      "step": 2496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.777788470202213,
+      "learning_rate": 9.56463044451186e-06,
+      "loss": 0.8558,
+      "step": 2497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7066538364225794,
+      "learning_rate": 9.564207330182359e-06,
+      "loss": 0.8278,
+      "step": 2498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.892682066037694,
+      "learning_rate": 9.563784019718704e-06,
+      "loss": 0.9005,
+      "step": 2499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.371342373986542,
+      "learning_rate": 9.563360513139085e-06,
+      "loss": 0.7954,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.027839513597188,
+      "learning_rate": 9.562936810461701e-06,
+      "loss": 1.0431,
+      "step": 2501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3591812821173554,
+      "learning_rate": 9.562512911704761e-06,
+      "loss": 0.77,
+      "step": 2502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3020992002553389,
+      "learning_rate": 9.56208881688648e-06,
+      "loss": 0.6923,
+      "step": 2503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1652306163588024,
+      "learning_rate": 9.561664526025082e-06,
+      "loss": 1.1216,
+      "step": 2504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.051510098799362,
+      "learning_rate": 9.5612400391388e-06,
+      "loss": 0.8724,
+      "step": 2505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.439897161315502,
+      "learning_rate": 9.560815356245875e-06,
+      "loss": 0.83,
+      "step": 2506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2536257203345327,
+      "learning_rate": 9.560390477364555e-06,
+      "loss": 0.677,
+      "step": 2507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0448434370287334,
+      "learning_rate": 9.559965402513103e-06,
+      "loss": 1.0416,
+      "step": 2508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.931431324026675,
+      "learning_rate": 9.55954013170978e-06,
+      "loss": 0.8805,
+      "step": 2509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.41746936701894,
+      "learning_rate": 9.559114664972864e-06,
+      "loss": 0.9528,
+      "step": 2510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.040504229644214,
+      "learning_rate": 9.558689002320635e-06,
+      "loss": 1.0726,
+      "step": 2511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.301972722635144,
+      "learning_rate": 9.55826314377139e-06,
+      "loss": 0.9293,
+      "step": 2512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.3280384515225325,
+      "learning_rate": 9.557837089343424e-06,
+      "loss": 0.8094,
+      "step": 2513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.883588838320448,
+      "learning_rate": 9.557410839055047e-06,
+      "loss": 0.8696,
+      "step": 2514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3264052080685755,
+      "learning_rate": 9.556984392924577e-06,
+      "loss": 0.8003,
+      "step": 2515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9573300587956468,
+      "learning_rate": 9.55655775097034e-06,
+      "loss": 0.782,
+      "step": 2516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3298017152770245,
+      "learning_rate": 9.556130913210666e-06,
+      "loss": 0.8005,
+      "step": 2517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.254095112673221,
+      "learning_rate": 9.5557038796639e-06,
+      "loss": 0.9361,
+      "step": 2518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7196421383791507,
+      "learning_rate": 9.555276650348393e-06,
+      "loss": 0.8951,
+      "step": 2519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6232402164310888,
+      "learning_rate": 9.554849225282503e-06,
+      "loss": 0.6103,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0438855475102278,
+      "learning_rate": 9.554421604484599e-06,
+      "loss": 0.8181,
+      "step": 2521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1986967175707433,
+      "learning_rate": 9.553993787973053e-06,
+      "loss": 0.8133,
+      "step": 2522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.114090624013568,
+      "learning_rate": 9.553565775766253e-06,
+      "loss": 0.8573,
+      "step": 2523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1658541380179095,
+      "learning_rate": 9.55313756788259e-06,
+      "loss": 0.7971,
+      "step": 2524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.945263936892752,
+      "learning_rate": 9.552709164340466e-06,
+      "loss": 0.9271,
+      "step": 2525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.616179068368502,
+      "learning_rate": 9.552280565158289e-06,
+      "loss": 0.9066,
+      "step": 2526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9744409565807053,
+      "learning_rate": 9.551851770354477e-06,
+      "loss": 0.9609,
+      "step": 2527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.872823693590382,
+      "learning_rate": 9.551422779947458e-06,
+      "loss": 0.9832,
+      "step": 2528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9470530365596115,
+      "learning_rate": 9.550993593955665e-06,
+      "loss": 0.8561,
+      "step": 2529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.004753074953651,
+      "learning_rate": 9.550564212397541e-06,
+      "loss": 1.0003,
+      "step": 2530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4148541515925097,
+      "learning_rate": 9.55013463529154e-06,
+      "loss": 0.8293,
+      "step": 2531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4481851150410794,
+      "learning_rate": 9.54970486265612e-06,
+      "loss": 0.8605,
+      "step": 2532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6744448105989105,
+      "learning_rate": 9.549274894509748e-06,
+      "loss": 0.9022,
+      "step": 2533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8420753050554675,
+      "learning_rate": 9.548844730870903e-06,
+      "loss": 0.8202,
+      "step": 2534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8476463743973435,
+      "learning_rate": 9.548414371758068e-06,
+      "loss": 0.8461,
+      "step": 2535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.153837270658438,
+      "learning_rate": 9.547983817189738e-06,
+      "loss": 0.8789,
+      "step": 2536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1210795687660713,
+      "learning_rate": 9.547553067184413e-06,
+      "loss": 0.6397,
+      "step": 2537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7120516207372853,
+      "learning_rate": 9.547122121760607e-06,
+      "loss": 0.9062,
+      "step": 2538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.015920964360898,
+      "learning_rate": 9.546690980936836e-06,
+      "loss": 0.8298,
+      "step": 2539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.149807499535779,
+      "learning_rate": 9.546259644731628e-06,
+      "loss": 1.0341,
+      "step": 2540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.714007349312216,
+      "learning_rate": 9.545828113163516e-06,
+      "loss": 0.7875,
+      "step": 2541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.360317943789131,
+      "learning_rate": 9.545396386251048e-06,
+      "loss": 0.9012,
+      "step": 2542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.570666514764328,
+      "learning_rate": 9.544964464012774e-06,
+      "loss": 1.0881,
+      "step": 2543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.027861911825199,
+      "learning_rate": 9.544532346467254e-06,
+      "loss": 0.9129,
+      "step": 2544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7558687703195766,
+      "learning_rate": 9.544100033633057e-06,
+      "loss": 0.7975,
+      "step": 2545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.157174509255589,
+      "learning_rate": 9.543667525528762e-06,
+      "loss": 0.8135,
+      "step": 2546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0710359148878648,
+      "learning_rate": 9.543234822172956e-06,
+      "loss": 0.9345,
+      "step": 2547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1136479302981352,
+      "learning_rate": 9.542801923584228e-06,
+      "loss": 0.739,
+      "step": 2548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.377045126469465,
+      "learning_rate": 9.542368829781186e-06,
+      "loss": 0.9237,
+      "step": 2549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2565896557495133,
+      "learning_rate": 9.541935540782439e-06,
+      "loss": 0.9704,
+      "step": 2550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3567054160225633,
+      "learning_rate": 9.541502056606606e-06,
+      "loss": 0.9991,
+      "step": 2551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0336593842512853,
+      "learning_rate": 9.541068377272315e-06,
+      "loss": 0.6375,
+      "step": 2552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.156446247185135,
+      "learning_rate": 9.540634502798203e-06,
+      "loss": 0.8971,
+      "step": 2553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.534626084246941,
+      "learning_rate": 9.540200433202913e-06,
+      "loss": 0.924,
+      "step": 2554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.250223183049308,
+      "learning_rate": 9.5397661685051e-06,
+      "loss": 1.0035,
+      "step": 2555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0583980176595187,
+      "learning_rate": 9.539331708723423e-06,
+      "loss": 0.7266,
+      "step": 2556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5761268821652226,
+      "learning_rate": 9.538897053876553e-06,
+      "loss": 0.8959,
+      "step": 2557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2178031261659785,
+      "learning_rate": 9.538462203983169e-06,
+      "loss": 0.8363,
+      "step": 2558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2589438249621607,
+      "learning_rate": 9.538027159061955e-06,
+      "loss": 0.9819,
+      "step": 2559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.332354500252085,
+      "learning_rate": 9.537591919131608e-06,
+      "loss": 0.8218,
+      "step": 2560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3287353499258945,
+      "learning_rate": 9.53715648421083e-06,
+      "loss": 0.6414,
+      "step": 2561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2062558007333997,
+      "learning_rate": 9.536720854318333e-06,
+      "loss": 0.9085,
+      "step": 2562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0555948851838326,
+      "learning_rate": 9.536285029472838e-06,
+      "loss": 0.6304,
+      "step": 2563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9693618386097294,
+      "learning_rate": 9.535849009693072e-06,
+      "loss": 0.7217,
+      "step": 2564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.860070959346095,
+      "learning_rate": 9.535412794997772e-06,
+      "loss": 0.8171,
+      "step": 2565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2126375998497654,
+      "learning_rate": 9.534976385405683e-06,
+      "loss": 0.787,
+      "step": 2566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0991965096185323,
+      "learning_rate": 9.53453978093556e-06,
+      "loss": 0.975,
+      "step": 2567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5347110552178624,
+      "learning_rate": 9.534102981606162e-06,
+      "loss": 1.004,
+      "step": 2568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.08697817905623,
+      "learning_rate": 9.533665987436262e-06,
+      "loss": 0.8592,
+      "step": 2569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.113041334927185,
+      "learning_rate": 9.533228798444637e-06,
+      "loss": 0.7933,
+      "step": 2570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5395738703622253,
+      "learning_rate": 9.532791414650075e-06,
+      "loss": 0.9141,
+      "step": 2571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5056089913517579,
+      "learning_rate": 9.53235383607137e-06,
+      "loss": 0.751,
+      "step": 2572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.33677528777256,
+      "learning_rate": 9.531916062727328e-06,
+      "loss": 0.7821,
+      "step": 2573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9992786358553234,
+      "learning_rate": 9.531478094636758e-06,
+      "loss": 0.8641,
+      "step": 2574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.621023297034126,
+      "learning_rate": 9.531039931818485e-06,
+      "loss": 0.9664,
+      "step": 2575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.362972006705663,
+      "learning_rate": 9.530601574291331e-06,
+      "loss": 0.9754,
+      "step": 2576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.490100731240754,
+      "learning_rate": 9.530163022074139e-06,
+      "loss": 1.0584,
+      "step": 2577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0046887393317743,
+      "learning_rate": 9.529724275185753e-06,
+      "loss": 0.6823,
+      "step": 2578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.488404637539367,
+      "learning_rate": 9.529285333645027e-06,
+      "loss": 0.7913,
+      "step": 2579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.180667146173916,
+      "learning_rate": 9.528846197470822e-06,
+      "loss": 0.8361,
+      "step": 2580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.746298021912139,
+      "learning_rate": 9.528406866682009e-06,
+      "loss": 0.8268,
+      "step": 2581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8580396257873546,
+      "learning_rate": 9.527967341297471e-06,
+      "loss": 0.7595,
+      "step": 2582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.752694550534503,
+      "learning_rate": 9.527527621336087e-06,
+      "loss": 1.0991,
+      "step": 2583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.270170448772418,
+      "learning_rate": 9.527087706816762e-06,
+      "loss": 0.8801,
+      "step": 2584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.310355062562527,
+      "learning_rate": 9.526647597758394e-06,
+      "loss": 0.8609,
+      "step": 2585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.773511979239335,
+      "learning_rate": 9.526207294179897e-06,
+      "loss": 1.0104,
+      "step": 2586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6705263930975232,
+      "learning_rate": 9.525766796100194e-06,
+      "loss": 0.7378,
+      "step": 2587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2427324246573455,
+      "learning_rate": 9.52532610353821e-06,
+      "loss": 0.6733,
+      "step": 2588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.171892855305902,
+      "learning_rate": 9.524885216512887e-06,
+      "loss": 0.7766,
+      "step": 2589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.332248712697839,
+      "learning_rate": 9.524444135043168e-06,
+      "loss": 0.8071,
+      "step": 2590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2645856812982035,
+      "learning_rate": 9.524002859148006e-06,
+      "loss": 0.8042,
+      "step": 2591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3001905103832105,
+      "learning_rate": 9.523561388846368e-06,
+      "loss": 0.9248,
+      "step": 2592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0645906169334145,
+      "learning_rate": 9.523119724157222e-06,
+      "loss": 0.7868,
+      "step": 2593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3847963702584143,
+      "learning_rate": 9.522677865099548e-06,
+      "loss": 0.9917,
+      "step": 2594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1059606397895454,
+      "learning_rate": 9.522235811692335e-06,
+      "loss": 0.8533,
+      "step": 2595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.898065038149568,
+      "learning_rate": 9.521793563954575e-06,
+      "loss": 0.7202,
+      "step": 2596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.296434686366211,
+      "learning_rate": 9.521351121905278e-06,
+      "loss": 0.8528,
+      "step": 2597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7664895570443928,
+      "learning_rate": 9.520908485563452e-06,
+      "loss": 0.8342,
+      "step": 2598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2515225217589925,
+      "learning_rate": 9.520465654948119e-06,
+      "loss": 0.9076,
+      "step": 2599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.853831745245739,
+      "learning_rate": 9.52002263007831e-06,
+      "loss": 0.7395,
+      "step": 2600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.171791273565816,
+      "learning_rate": 9.519579410973061e-06,
+      "loss": 1.0701,
+      "step": 2601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2471532929777718,
+      "learning_rate": 9.51913599765142e-06,
+      "loss": 0.9446,
+      "step": 2602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3135915696803937,
+      "learning_rate": 9.518692390132441e-06,
+      "loss": 0.9578,
+      "step": 2603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2964675953944136,
+      "learning_rate": 9.518248588435185e-06,
+      "loss": 0.8873,
+      "step": 2604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3204520480433035,
+      "learning_rate": 9.517804592578725e-06,
+      "loss": 0.9002,
+      "step": 2605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1962424206367976,
+      "learning_rate": 9.517360402582141e-06,
+      "loss": 0.9601,
+      "step": 2606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.167532444913423,
+      "learning_rate": 9.51691601846452e-06,
+      "loss": 0.8725,
+      "step": 2607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4284202119829863,
+      "learning_rate": 9.516471440244956e-06,
+      "loss": 0.7203,
+      "step": 2608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3552036796248252,
+      "learning_rate": 9.516026667942557e-06,
+      "loss": 0.987,
+      "step": 2609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.965344572757636,
+      "learning_rate": 9.515581701576433e-06,
+      "loss": 0.8817,
+      "step": 2610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0398273805566576,
+      "learning_rate": 9.515136541165708e-06,
+      "loss": 1.03,
+      "step": 2611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.620978841567955,
+      "learning_rate": 9.514691186729509e-06,
+      "loss": 0.9221,
+      "step": 2612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1422544202486944,
+      "learning_rate": 9.514245638286976e-06,
+      "loss": 0.91,
+      "step": 2613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0417797338529375,
+      "learning_rate": 9.513799895857252e-06,
+      "loss": 0.8615,
+      "step": 2614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3247386899412525,
+      "learning_rate": 9.513353959459496e-06,
+      "loss": 0.9333,
+      "step": 2615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.934416621240369,
+      "learning_rate": 9.512907829112867e-06,
+      "loss": 0.7539,
+      "step": 2616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5811846772065192,
+      "learning_rate": 9.512461504836538e-06,
+      "loss": 0.8636,
+      "step": 2617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1629740943626763,
+      "learning_rate": 9.512014986649691e-06,
+      "loss": 0.6278,
+      "step": 2618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.875509460059527,
+      "learning_rate": 9.511568274571508e-06,
+      "loss": 0.9587,
+      "step": 2619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.81035441632745,
+      "learning_rate": 9.51112136862119e-06,
+      "loss": 1.0184,
+      "step": 2620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9157842917029804,
+      "learning_rate": 9.51067426881794e-06,
+      "loss": 0.9062,
+      "step": 2621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.414397549187329,
+      "learning_rate": 9.510226975180969e-06,
+      "loss": 0.7286,
+      "step": 2622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9456627446687838,
+      "learning_rate": 9.509779487729502e-06,
+      "loss": 0.7824,
+      "step": 2623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.3373352610847755,
+      "learning_rate": 9.509331806482767e-06,
+      "loss": 0.9105,
+      "step": 2624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.49508623503877,
+      "learning_rate": 9.50888393146e-06,
+      "loss": 0.7807,
+      "step": 2625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.095951499715052,
+      "learning_rate": 9.508435862680449e-06,
+      "loss": 0.9361,
+      "step": 2626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.774027306547115,
+      "learning_rate": 9.507987600163369e-06,
+      "loss": 0.8587,
+      "step": 2627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9408652307953376,
+      "learning_rate": 9.507539143928021e-06,
+      "loss": 0.9981,
+      "step": 2628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.345832772801703,
+      "learning_rate": 9.507090493993677e-06,
+      "loss": 1.0658,
+      "step": 2629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1136138668898834,
+      "learning_rate": 9.506641650379617e-06,
+      "loss": 0.6293,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2450816538291956,
+      "learning_rate": 9.50619261310513e-06,
+      "loss": 1.0126,
+      "step": 2631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3803252747146604,
+      "learning_rate": 9.50574338218951e-06,
+      "loss": 0.9868,
+      "step": 2632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7713368425613112,
+      "learning_rate": 9.50529395765206e-06,
+      "loss": 0.9963,
+      "step": 2633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.409993040771167,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.8879,
+      "step": 2634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.800511854587184,
+      "learning_rate": 9.504394527788939e-06,
+      "loss": 0.874,
+      "step": 2635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7456201962789275,
+      "learning_rate": 9.503944522501915e-06,
+      "loss": 0.861,
+      "step": 2636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1799754801466016,
+      "learning_rate": 9.503494323670366e-06,
+      "loss": 0.7028,
+      "step": 2637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.193033824646152,
+      "learning_rate": 9.503043931313637e-06,
+      "loss": 1.1118,
+      "step": 2638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2708325588815665,
+      "learning_rate": 9.502593345451078e-06,
+      "loss": 0.6611,
+      "step": 2639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4464034082498998,
+      "learning_rate": 9.502142566102058e-06,
+      "loss": 0.7994,
+      "step": 2640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0279109917245735,
+      "learning_rate": 9.501691593285943e-06,
+      "loss": 0.6969,
+      "step": 2641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.381129873955209,
+      "learning_rate": 9.501240427022115e-06,
+      "loss": 0.7931,
+      "step": 2642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.056005597005973,
+      "learning_rate": 9.500789067329963e-06,
+      "loss": 0.7573,
+      "step": 2643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.488738616250376,
+      "learning_rate": 9.500337514228878e-06,
+      "loss": 0.9511,
+      "step": 2644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.145665838213303,
+      "learning_rate": 9.49988576773827e-06,
+      "loss": 0.7166,
+      "step": 2645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0676836711835773,
+      "learning_rate": 9.499433827877547e-06,
+      "loss": 0.7067,
+      "step": 2646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4707049960388647,
+      "learning_rate": 9.498981694666133e-06,
+      "loss": 0.7349,
+      "step": 2647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.139920446475341,
+      "learning_rate": 9.498529368123455e-06,
+      "loss": 0.7101,
+      "step": 2648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.155105117756981,
+      "learning_rate": 9.49807684826895e-06,
+      "loss": 0.6644,
+      "step": 2649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9546696311357015,
+      "learning_rate": 9.497624135122068e-06,
+      "loss": 0.9466,
+      "step": 2650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9559383691903713,
+      "learning_rate": 9.497171228702257e-06,
+      "loss": 0.7778,
+      "step": 2651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1074491319192408,
+      "learning_rate": 9.496718129028986e-06,
+      "loss": 0.7773,
+      "step": 2652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2976557649559806,
+      "learning_rate": 9.49626483612172e-06,
+      "loss": 0.9043,
+      "step": 2653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8212776681951943,
+      "learning_rate": 9.495811349999941e-06,
+      "loss": 0.9291,
+      "step": 2654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.314812915757142,
+      "learning_rate": 9.495357670683135e-06,
+      "loss": 0.9877,
+      "step": 2655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0054751595394182,
+      "learning_rate": 9.494903798190798e-06,
+      "loss": 0.9669,
+      "step": 2656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.478877316215389,
+      "learning_rate": 9.494449732542435e-06,
+      "loss": 0.8025,
+      "step": 2657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7862954830780393,
+      "learning_rate": 9.493995473757558e-06,
+      "loss": 0.843,
+      "step": 2658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1468148269082725,
+      "learning_rate": 9.493541021855685e-06,
+      "loss": 0.9325,
+      "step": 2659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1512350157461335,
+      "learning_rate": 9.493086376856346e-06,
+      "loss": 0.8009,
+      "step": 2660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0585321705525315,
+      "learning_rate": 9.492631538779082e-06,
+      "loss": 0.6758,
+      "step": 2661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.293188098977849,
+      "learning_rate": 9.492176507643433e-06,
+      "loss": 0.7722,
+      "step": 2662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4014493477080596,
+      "learning_rate": 9.491721283468954e-06,
+      "loss": 1.0372,
+      "step": 2663
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2398979332152618,
+      "learning_rate": 9.49126586627521e-06,
+      "loss": 0.6806,
+      "step": 2664
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1834668755939446,
+      "learning_rate": 9.490810256081766e-06,
+      "loss": 0.996,
+      "step": 2665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9141852869971483,
+      "learning_rate": 9.490354452908208e-06,
+      "loss": 0.7958,
+      "step": 2666
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5625865756884227,
+      "learning_rate": 9.489898456774116e-06,
+      "loss": 0.9781,
+      "step": 2667
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.24701782372567,
+      "learning_rate": 9.489442267699087e-06,
+      "loss": 0.9993,
+      "step": 2668
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4384408398632895,
+      "learning_rate": 9.488985885702728e-06,
+      "loss": 0.8816,
+      "step": 2669
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6413012479850901,
+      "learning_rate": 9.488529310804645e-06,
+      "loss": 0.8305,
+      "step": 2670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9064815181958612,
+      "learning_rate": 9.488072543024462e-06,
+      "loss": 0.8932,
+      "step": 2671
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.025388618636534,
+      "learning_rate": 9.487615582381809e-06,
+      "loss": 0.8778,
+      "step": 2672
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6864939714960516,
+      "learning_rate": 9.487158428896317e-06,
+      "loss": 0.7753,
+      "step": 2673
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9899438013432627,
+      "learning_rate": 9.486701082587635e-06,
+      "loss": 0.9845,
+      "step": 2674
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0439342547487356,
+      "learning_rate": 9.486243543475414e-06,
+      "loss": 0.7764,
+      "step": 2675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9288498046562559,
+      "learning_rate": 9.485785811579318e-06,
+      "loss": 0.9307,
+      "step": 2676
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0933951813964173,
+      "learning_rate": 9.485327886919013e-06,
+      "loss": 0.6125,
+      "step": 2677
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.678334146796131,
+      "learning_rate": 9.484869769514182e-06,
+      "loss": 1.1137,
+      "step": 2678
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.39696566385593,
+      "learning_rate": 9.484411459384508e-06,
+      "loss": 0.7089,
+      "step": 2679
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3133472193828253,
+      "learning_rate": 9.483952956549684e-06,
+      "loss": 0.7414,
+      "step": 2680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0834598825299326,
+      "learning_rate": 9.483494261029418e-06,
+      "loss": 0.8837,
+      "step": 2681
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9895827126630286,
+      "learning_rate": 9.483035372843417e-06,
+      "loss": 0.9462,
+      "step": 2682
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.933733566685967,
+      "learning_rate": 9.482576292011402e-06,
+      "loss": 0.8776,
+      "step": 2683
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3997557905756333,
+      "learning_rate": 9.482117018553101e-06,
+      "loss": 0.9798,
+      "step": 2684
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0791930829661784,
+      "learning_rate": 9.481657552488248e-06,
+      "loss": 0.688,
+      "step": 2685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.305506373550264,
+      "learning_rate": 9.481197893836591e-06,
+      "loss": 0.9118,
+      "step": 2686
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1689455160273825,
+      "learning_rate": 9.480738042617878e-06,
+      "loss": 0.7122,
+      "step": 2687
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.487058799319793,
+      "learning_rate": 9.480277998851875e-06,
+      "loss": 0.8014,
+      "step": 2688
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1708174008192014,
+      "learning_rate": 9.479817762558345e-06,
+      "loss": 0.9505,
+      "step": 2689
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8784349895967007,
+      "learning_rate": 9.479357333757072e-06,
+      "loss": 0.9135,
+      "step": 2690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.291292518511899,
+      "learning_rate": 9.478896712467836e-06,
+      "loss": 0.7703,
+      "step": 2691
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7403229402143927,
+      "learning_rate": 9.478435898710434e-06,
+      "loss": 0.6893,
+      "step": 2692
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.463546559986514,
+      "learning_rate": 9.477974892504666e-06,
+      "loss": 0.9192,
+      "step": 2693
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4402045579723124,
+      "learning_rate": 9.477513693870347e-06,
+      "loss": 0.9592,
+      "step": 2694
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.263051628495428,
+      "learning_rate": 9.47705230282729e-06,
+      "loss": 0.9606,
+      "step": 2695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9189189624324452,
+      "learning_rate": 9.476590719395324e-06,
+      "loss": 0.9902,
+      "step": 2696
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.73970821609968,
+      "learning_rate": 9.476128943594287e-06,
+      "loss": 0.7294,
+      "step": 2697
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.272952448554758,
+      "learning_rate": 9.47566697544402e-06,
+      "loss": 0.9794,
+      "step": 2698
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8921215758553316,
+      "learning_rate": 9.475204814964374e-06,
+      "loss": 0.9501,
+      "step": 2699
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2051460156284488,
+      "learning_rate": 9.474742462175212e-06,
+      "loss": 0.6593,
+      "step": 2700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3012750923946566,
+      "learning_rate": 9.4742799170964e-06,
+      "loss": 0.6754,
+      "step": 2701
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7685469090878365,
+      "learning_rate": 9.473817179747815e-06,
+      "loss": 0.9089,
+      "step": 2702
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3200268469658687,
+      "learning_rate": 9.473354250149342e-06,
+      "loss": 0.758,
+      "step": 2703
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.4661554936994,
+      "learning_rate": 9.472891128320874e-06,
+      "loss": 0.8759,
+      "step": 2704
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5065183874546824,
+      "learning_rate": 9.472427814282311e-06,
+      "loss": 0.8809,
+      "step": 2705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.244879173323187,
+      "learning_rate": 9.471964308053567e-06,
+      "loss": 0.9388,
+      "step": 2706
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.339551208162634,
+      "learning_rate": 9.471500609654556e-06,
+      "loss": 0.9399,
+      "step": 2707
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9221665682427775,
+      "learning_rate": 9.471036719105205e-06,
+      "loss": 0.8863,
+      "step": 2708
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7865676553778047,
+      "learning_rate": 9.470572636425451e-06,
+      "loss": 0.919,
+      "step": 2709
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6367547292659084,
+      "learning_rate": 9.470108361635231e-06,
+      "loss": 0.87,
+      "step": 2710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0696128493470884,
+      "learning_rate": 9.469643894754502e-06,
+      "loss": 0.6243,
+      "step": 2711
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.73191955638035,
+      "learning_rate": 9.469179235803217e-06,
+      "loss": 0.8641,
+      "step": 2712
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1932552016895985,
+      "learning_rate": 9.46871438480135e-06,
+      "loss": 0.9825,
+      "step": 2713
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.374291891245377,
+      "learning_rate": 9.46824934176887e-06,
+      "loss": 0.8028,
+      "step": 2714
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.275559743170702,
+      "learning_rate": 9.467784106725767e-06,
+      "loss": 0.9391,
+      "step": 2715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2067258442913404,
+      "learning_rate": 9.467318679692031e-06,
+      "loss": 0.8534,
+      "step": 2716
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9948259352853956,
+      "learning_rate": 9.466853060687662e-06,
+      "loss": 0.7094,
+      "step": 2717
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8657920010430624,
+      "learning_rate": 9.466387249732668e-06,
+      "loss": 0.9512,
+      "step": 2718
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3858246912864973,
+      "learning_rate": 9.465921246847067e-06,
+      "loss": 0.9419,
+      "step": 2719
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7310977775544862,
+      "learning_rate": 9.465455052050884e-06,
+      "loss": 0.7494,
+      "step": 2720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1050260243433296,
+      "learning_rate": 9.464988665364152e-06,
+      "loss": 0.9379,
+      "step": 2721
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.837415463807068,
+      "learning_rate": 9.464522086806914e-06,
+      "loss": 0.8394,
+      "step": 2722
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8873094323694612,
+      "learning_rate": 9.464055316399217e-06,
+      "loss": 0.9679,
+      "step": 2723
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5626320580421846,
+      "learning_rate": 9.463588354161122e-06,
+      "loss": 0.7527,
+      "step": 2724
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9993064127186349,
+      "learning_rate": 9.463121200112694e-06,
+      "loss": 1.0035,
+      "step": 2725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2669253838103156,
+      "learning_rate": 9.462653854274008e-06,
+      "loss": 0.9182,
+      "step": 2726
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6186589643767797,
+      "learning_rate": 9.462186316665148e-06,
+      "loss": 0.9824,
+      "step": 2727
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.464258759071543,
+      "learning_rate": 9.461718587306205e-06,
+      "loss": 1.0027,
+      "step": 2728
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5659773240018873,
+      "learning_rate": 9.461250666217277e-06,
+      "loss": 0.773,
+      "step": 2729
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1082941999716693,
+      "learning_rate": 9.46078255341847e-06,
+      "loss": 0.8417,
+      "step": 2730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8601521877648286,
+      "learning_rate": 9.460314248929903e-06,
+      "loss": 0.8964,
+      "step": 2731
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.638214182103273,
+      "learning_rate": 9.4598457527717e-06,
+      "loss": 0.9452,
+      "step": 2732
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5215911952148624,
+      "learning_rate": 9.459377064963993e-06,
+      "loss": 1.0169,
+      "step": 2733
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.13264250088916,
+      "learning_rate": 9.458908185526921e-06,
+      "loss": 0.7556,
+      "step": 2734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1752914944803017,
+      "learning_rate": 9.458439114480634e-06,
+      "loss": 0.6557,
+      "step": 2735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.569653829817035,
+      "learning_rate": 9.457969851845287e-06,
+      "loss": 0.7764,
+      "step": 2736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1399161313356627,
+      "learning_rate": 9.457500397641049e-06,
+      "loss": 1.0946,
+      "step": 2737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.227923783305132,
+      "learning_rate": 9.45703075188809e-06,
+      "loss": 1.0877,
+      "step": 2738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9049090725255677,
+      "learning_rate": 9.456560914606594e-06,
+      "loss": 0.9239,
+      "step": 2739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.103937310365811,
+      "learning_rate": 9.456090885816751e-06,
+      "loss": 0.8426,
+      "step": 2740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.637466264747535,
+      "learning_rate": 9.455620665538758e-06,
+      "loss": 0.8086,
+      "step": 2741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.77201906341018,
+      "learning_rate": 9.455150253792822e-06,
+      "loss": 0.8954,
+      "step": 2742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.942646214441931,
+      "learning_rate": 9.454679650599156e-06,
+      "loss": 0.7748,
+      "step": 2743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7499495634754907,
+      "learning_rate": 9.454208855977986e-06,
+      "loss": 0.9334,
+      "step": 2744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1385917309957714,
+      "learning_rate": 9.45373786994954e-06,
+      "loss": 0.8732,
+      "step": 2745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.116074180673772,
+      "learning_rate": 9.45326669253406e-06,
+      "loss": 0.8418,
+      "step": 2746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.038084585941397,
+      "learning_rate": 9.452795323751793e-06,
+      "loss": 0.9594,
+      "step": 2747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6799793350074923,
+      "learning_rate": 9.452323763622993e-06,
+      "loss": 0.9189,
+      "step": 2748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1818137169991878,
+      "learning_rate": 9.451852012167924e-06,
+      "loss": 0.7153,
+      "step": 2749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3329837671151843,
+      "learning_rate": 9.45138006940686e-06,
+      "loss": 0.9579,
+      "step": 2750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5816240251140945,
+      "learning_rate": 9.450907935360081e-06,
+      "loss": 1.0593,
+      "step": 2751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4616260172716173,
+      "learning_rate": 9.450435610047875e-06,
+      "loss": 0.8716,
+      "step": 2752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.178207698429367,
+      "learning_rate": 9.44996309349054e-06,
+      "loss": 0.7297,
+      "step": 2753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2196307307318155,
+      "learning_rate": 9.449490385708378e-06,
+      "loss": 0.7132,
+      "step": 2754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1458322340777802,
+      "learning_rate": 9.449017486721705e-06,
+      "loss": 0.8704,
+      "step": 2755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2419370948985784,
+      "learning_rate": 9.448544396550846e-06,
+      "loss": 0.9377,
+      "step": 2756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.30013639838421,
+      "learning_rate": 9.448071115216123e-06,
+      "loss": 0.9296,
+      "step": 2757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7882454590160735,
+      "learning_rate": 9.447597642737878e-06,
+      "loss": 0.9808,
+      "step": 2758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.834931863004433,
+      "learning_rate": 9.447123979136457e-06,
+      "loss": 0.7377,
+      "step": 2759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8581258413407764,
+      "learning_rate": 9.446650124432215e-06,
+      "loss": 0.6639,
+      "step": 2760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8976809079686008,
+      "learning_rate": 9.446176078645515e-06,
+      "loss": 1.0146,
+      "step": 2761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3043973886073044,
+      "learning_rate": 9.445701841796726e-06,
+      "loss": 0.9907,
+      "step": 2762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.271019838687002,
+      "learning_rate": 9.445227413906227e-06,
+      "loss": 0.8873,
+      "step": 2763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3012819773963784,
+      "learning_rate": 9.444752794994408e-06,
+      "loss": 0.8182,
+      "step": 2764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.999788852861631,
+      "learning_rate": 9.44427798508166e-06,
+      "loss": 1.0069,
+      "step": 2765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4626775371659018,
+      "learning_rate": 9.44380298418839e-06,
+      "loss": 0.9229,
+      "step": 2766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.089129207722837,
+      "learning_rate": 9.443327792335009e-06,
+      "loss": 0.8905,
+      "step": 2767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8407142900715598,
+      "learning_rate": 9.442852409541935e-06,
+      "loss": 0.8329,
+      "step": 2768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.093468100188462,
+      "learning_rate": 9.4423768358296e-06,
+      "loss": 0.9961,
+      "step": 2769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1982785125030408,
+      "learning_rate": 9.44190107121844e-06,
+      "loss": 0.6908,
+      "step": 2770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.003488300375799,
+      "learning_rate": 9.441425115728898e-06,
+      "loss": 0.902,
+      "step": 2771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.823170196667486,
+      "learning_rate": 9.440948969381425e-06,
+      "loss": 0.9058,
+      "step": 2772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.505838914955009,
+      "learning_rate": 9.440472632196485e-06,
+      "loss": 1.0616,
+      "step": 2773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6727909581409937,
+      "learning_rate": 9.439996104194546e-06,
+      "loss": 0.8607,
+      "step": 2774
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2022745144723475,
+      "learning_rate": 9.439519385396088e-06,
+      "loss": 1.0013,
+      "step": 2775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.435580373752283,
+      "learning_rate": 9.439042475821594e-06,
+      "loss": 0.787,
+      "step": 2776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9989194441047314,
+      "learning_rate": 9.438565375491558e-06,
+      "loss": 0.7199,
+      "step": 2777
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0178567800926177,
+      "learning_rate": 9.438088084426482e-06,
+      "loss": 0.9695,
+      "step": 2778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5325901086158327,
+      "learning_rate": 9.437610602646878e-06,
+      "loss": 0.9433,
+      "step": 2779
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1349497902301646,
+      "learning_rate": 9.437132930173264e-06,
+      "loss": 0.8706,
+      "step": 2780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.983378039962657,
+      "learning_rate": 9.436655067026164e-06,
+      "loss": 0.9922,
+      "step": 2781
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.808067841277397,
+      "learning_rate": 9.436177013226116e-06,
+      "loss": 0.9794,
+      "step": 2782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6267473163344888,
+      "learning_rate": 9.43569876879366e-06,
+      "loss": 0.7723,
+      "step": 2783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5465476378284304,
+      "learning_rate": 9.43522033374935e-06,
+      "loss": 0.8156,
+      "step": 2784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0578176332879687,
+      "learning_rate": 9.434741708113745e-06,
+      "loss": 0.9973,
+      "step": 2785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5897212075348133,
+      "learning_rate": 9.434262891907413e-06,
+      "loss": 0.8457,
+      "step": 2786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.233589026758716,
+      "learning_rate": 9.433783885150929e-06,
+      "loss": 0.7635,
+      "step": 2787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.970310494668221,
+      "learning_rate": 9.433304687864876e-06,
+      "loss": 0.9159,
+      "step": 2788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4227794977304193,
+      "learning_rate": 9.432825300069848e-06,
+      "loss": 0.9952,
+      "step": 2789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1198485536448337,
+      "learning_rate": 9.432345721786442e-06,
+      "loss": 0.9357,
+      "step": 2790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3359122437971878,
+      "learning_rate": 9.431865953035271e-06,
+      "loss": 0.7531,
+      "step": 2791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.45070418678759,
+      "learning_rate": 9.431385993836949e-06,
+      "loss": 0.8124,
+      "step": 2792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.07914482686103,
+      "learning_rate": 9.430905844212102e-06,
+      "loss": 0.7489,
+      "step": 2793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3356192165727716,
+      "learning_rate": 9.430425504181361e-06,
+      "loss": 0.8035,
+      "step": 2794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3139911447676313,
+      "learning_rate": 9.429944973765373e-06,
+      "loss": 0.9447,
+      "step": 2795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.777351001429364,
+      "learning_rate": 9.429464252984779e-06,
+      "loss": 1.0763,
+      "step": 2796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.535140029338218,
+      "learning_rate": 9.428983341860244e-06,
+      "loss": 1.0835,
+      "step": 2797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.41700908762767,
+      "learning_rate": 9.42850224041243e-06,
+      "loss": 0.7346,
+      "step": 2798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.218438961638504,
+      "learning_rate": 9.428020948662012e-06,
+      "loss": 0.9159,
+      "step": 2799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.221871199220034,
+      "learning_rate": 9.427539466629672e-06,
+      "loss": 0.765,
+      "step": 2800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.224161355073441,
+      "learning_rate": 9.427057794336101e-06,
+      "loss": 0.6313,
+      "step": 2801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1825886066308045,
+      "learning_rate": 9.426575931801996e-06,
+      "loss": 0.6803,
+      "step": 2802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1038299693937446,
+      "learning_rate": 9.426093879048065e-06,
+      "loss": 0.8434,
+      "step": 2803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2100159265766153,
+      "learning_rate": 9.425611636095023e-06,
+      "loss": 0.8387,
+      "step": 2804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.087402731313835,
+      "learning_rate": 9.425129202963592e-06,
+      "loss": 0.8895,
+      "step": 2805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2709640149302945,
+      "learning_rate": 9.424646579674504e-06,
+      "loss": 0.9113,
+      "step": 2806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.955046945272054,
+      "learning_rate": 9.424163766248499e-06,
+      "loss": 0.8666,
+      "step": 2807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0752471380030335,
+      "learning_rate": 9.423680762706323e-06,
+      "loss": 0.8039,
+      "step": 2808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.286405641391074,
+      "learning_rate": 9.423197569068733e-06,
+      "loss": 0.8125,
+      "step": 2809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.190385256672535,
+      "learning_rate": 9.422714185356491e-06,
+      "loss": 0.9298,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.792925685118603,
+      "learning_rate": 9.422230611590373e-06,
+      "loss": 0.8519,
+      "step": 2811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.740446431770417,
+      "learning_rate": 9.421746847791156e-06,
+      "loss": 0.8991,
+      "step": 2812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.139158794120214,
+      "learning_rate": 9.421262893979628e-06,
+      "loss": 0.9282,
+      "step": 2813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5714120931004856,
+      "learning_rate": 9.420778750176588e-06,
+      "loss": 0.933,
+      "step": 2814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.801948542499589,
+      "learning_rate": 9.420294416402839e-06,
+      "loss": 0.8195,
+      "step": 2815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3748315627153818,
+      "learning_rate": 9.419809892679194e-06,
+      "loss": 0.8793,
+      "step": 2816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1720551272923883,
+      "learning_rate": 9.419325179026475e-06,
+      "loss": 0.6422,
+      "step": 2817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7162507529563227,
+      "learning_rate": 9.41884027546551e-06,
+      "loss": 0.727,
+      "step": 2818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6797753851218022,
+      "learning_rate": 9.418355182017138e-06,
+      "loss": 0.7454,
+      "step": 2819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7743532741943833,
+      "learning_rate": 9.417869898702205e-06,
+      "loss": 0.7944,
+      "step": 2820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1385531372485083,
+      "learning_rate": 9.41738442554156e-06,
+      "loss": 0.745,
+      "step": 2821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7507528584611154,
+      "learning_rate": 9.41689876255607e-06,
+      "loss": 0.9456,
+      "step": 2822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3068617390331383,
+      "learning_rate": 9.416412909766604e-06,
+      "loss": 0.846,
+      "step": 2823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.088704640293213,
+      "learning_rate": 9.41592686719404e-06,
+      "loss": 0.8369,
+      "step": 2824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.570745019798633,
+      "learning_rate": 9.41544063485926e-06,
+      "loss": 0.8353,
+      "step": 2825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7698280347518762,
+      "learning_rate": 9.414954212783165e-06,
+      "loss": 0.6856,
+      "step": 2826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5936868831094366,
+      "learning_rate": 9.414467600986655e-06,
+      "loss": 0.9693,
+      "step": 2827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.524170036436239,
+      "learning_rate": 9.41398079949064e-06,
+      "loss": 0.8601,
+      "step": 2828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.348578065745812,
+      "learning_rate": 9.413493808316038e-06,
+      "loss": 0.9597,
+      "step": 2829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.075782530474753,
+      "learning_rate": 9.41300662748378e-06,
+      "loss": 0.6689,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4063304390088076,
+      "learning_rate": 9.412519257014798e-06,
+      "loss": 0.9429,
+      "step": 2831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0025800083339145,
+      "learning_rate": 9.412031696930036e-06,
+      "loss": 1.0171,
+      "step": 2832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6100999068316697,
+      "learning_rate": 9.411543947250447e-06,
+      "loss": 0.6783,
+      "step": 2833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.310727794019857,
+      "learning_rate": 9.411056007996989e-06,
+      "loss": 0.7469,
+      "step": 2834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5599741375232212,
+      "learning_rate": 9.41056787919063e-06,
+      "loss": 0.8647,
+      "step": 2835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1280115388041465,
+      "learning_rate": 9.410079560852347e-06,
+      "loss": 0.931,
+      "step": 2836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0911841641922613,
+      "learning_rate": 9.409591053003121e-06,
+      "loss": 0.8858,
+      "step": 2837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.147351871227509,
+      "learning_rate": 9.40910235566395e-06,
+      "loss": 0.6628,
+      "step": 2838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.313640253371959,
+      "learning_rate": 9.408613468855829e-06,
+      "loss": 0.8651,
+      "step": 2839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.003117077927977,
+      "learning_rate": 9.40812439259977e-06,
+      "loss": 0.9251,
+      "step": 2840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.912514442682935,
+      "learning_rate": 9.40763512691679e-06,
+      "loss": 0.828,
+      "step": 2841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.639825729444862,
+      "learning_rate": 9.407145671827909e-06,
+      "loss": 0.8624,
+      "step": 2842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.7800030996051746,
+      "learning_rate": 9.406656027354166e-06,
+      "loss": 0.8597,
+      "step": 2843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7448769948875626,
+      "learning_rate": 9.406166193516596e-06,
+      "loss": 0.8985,
+      "step": 2844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9498606048940212,
+      "learning_rate": 9.405676170336255e-06,
+      "loss": 1.1047,
+      "step": 2845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.505000124544261,
+      "learning_rate": 9.405185957834196e-06,
+      "loss": 0.8446,
+      "step": 2846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9550885647545004,
+      "learning_rate": 9.404695556031485e-06,
+      "loss": 1.0376,
+      "step": 2847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.070914832092053,
+      "learning_rate": 9.404204964949196e-06,
+      "loss": 0.6064,
+      "step": 2848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7891714046270764,
+      "learning_rate": 9.403714184608411e-06,
+      "loss": 0.8849,
+      "step": 2849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7479356527364305,
+      "learning_rate": 9.403223215030221e-06,
+      "loss": 0.8208,
+      "step": 2850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1531708739929543,
+      "learning_rate": 9.402732056235722e-06,
+      "loss": 0.8568,
+      "step": 2851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2499820790061014,
+      "learning_rate": 9.402240708246022e-06,
+      "loss": 0.8399,
+      "step": 2852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7218908403733013,
+      "learning_rate": 9.401749171082233e-06,
+      "loss": 0.8777,
+      "step": 2853
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.243373592752994,
+      "learning_rate": 9.40125744476548e-06,
+      "loss": 0.8832,
+      "step": 2854
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.391214699162161,
+      "learning_rate": 9.400765529316892e-06,
+      "loss": 0.8911,
+      "step": 2855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.744182655996322,
+      "learning_rate": 9.400273424757607e-06,
+      "loss": 0.8544,
+      "step": 2856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7841501652753284,
+      "learning_rate": 9.399781131108775e-06,
+      "loss": 0.9639,
+      "step": 2857
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.597258332844172,
+      "learning_rate": 9.399288648391548e-06,
+      "loss": 0.956,
+      "step": 2858
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1865009909611615,
+      "learning_rate": 9.398795976627091e-06,
+      "loss": 0.7262,
+      "step": 2859
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1046025331957874,
+      "learning_rate": 9.398303115836573e-06,
+      "loss": 0.8672,
+      "step": 2860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7272872469940728,
+      "learning_rate": 9.397810066041176e-06,
+      "loss": 0.9784,
+      "step": 2861
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.368688734263772,
+      "learning_rate": 9.397316827262085e-06,
+      "loss": 0.8311,
+      "step": 2862
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7255619110372886,
+      "learning_rate": 9.396823399520495e-06,
+      "loss": 0.7813,
+      "step": 2863
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1038618598429886,
+      "learning_rate": 9.396329782837614e-06,
+      "loss": 0.6791,
+      "step": 2864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9881663656678812,
+      "learning_rate": 9.39583597723465e-06,
+      "loss": 0.7852,
+      "step": 2865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3782270990899734,
+      "learning_rate": 9.395341982732825e-06,
+      "loss": 1.0081,
+      "step": 2866
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.567292420345793,
+      "learning_rate": 9.394847799353363e-06,
+      "loss": 0.7207,
+      "step": 2867
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9837047572992872,
+      "learning_rate": 9.394353427117506e-06,
+      "loss": 1.0521,
+      "step": 2868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3934226842538813,
+      "learning_rate": 9.393858866046494e-06,
+      "loss": 0.7241,
+      "step": 2869
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.193643195692098,
+      "learning_rate": 9.393364116161582e-06,
+      "loss": 0.8881,
+      "step": 2870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.856247588362727,
+      "learning_rate": 9.392869177484029e-06,
+      "loss": 0.9403,
+      "step": 2871
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8715315818181963,
+      "learning_rate": 9.392374050035104e-06,
+      "loss": 0.9176,
+      "step": 2872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2770347148227628,
+      "learning_rate": 9.391878733836085e-06,
+      "loss": 0.8838,
+      "step": 2873
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6182314909752686,
+      "learning_rate": 9.391383228908253e-06,
+      "loss": 0.9448,
+      "step": 2874
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.546985888860212,
+      "learning_rate": 9.390887535272906e-06,
+      "loss": 0.7389,
+      "step": 2875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1398788130976962,
+      "learning_rate": 9.39039165295134e-06,
+      "loss": 0.7356,
+      "step": 2876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9409700146755955,
+      "learning_rate": 9.38989558196487e-06,
+      "loss": 0.7474,
+      "step": 2877
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5956054545225773,
+      "learning_rate": 9.389399322334806e-06,
+      "loss": 0.8995,
+      "step": 2878
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.873520198190716,
+      "learning_rate": 9.388902874082482e-06,
+      "loss": 1.0127,
+      "step": 2879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1934928322305507,
+      "learning_rate": 9.388406237229223e-06,
+      "loss": 0.719,
+      "step": 2880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0486423124029307,
+      "learning_rate": 9.387909411796377e-06,
+      "loss": 1.104,
+      "step": 2881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3594676820187845,
+      "learning_rate": 9.387412397805289e-06,
+      "loss": 0.8041,
+      "step": 2882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.537981455093412,
+      "learning_rate": 9.386915195277321e-06,
+      "loss": 0.9036,
+      "step": 2883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3783383472611166,
+      "learning_rate": 9.386417804233836e-06,
+      "loss": 0.9009,
+      "step": 2884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.370549619756182,
+      "learning_rate": 9.385920224696208e-06,
+      "loss": 0.9352,
+      "step": 2885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9792821858569491,
+      "learning_rate": 9.385422456685822e-06,
+      "loss": 0.8758,
+      "step": 2886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3978557879227655,
+      "learning_rate": 9.384924500224065e-06,
+      "loss": 0.6923,
+      "step": 2887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.631512311598279,
+      "learning_rate": 9.384426355332336e-06,
+      "loss": 0.8597,
+      "step": 2888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.982153050742472,
+      "learning_rate": 9.383928022032044e-06,
+      "loss": 0.7604,
+      "step": 2889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3165168201274002,
+      "learning_rate": 9.383429500344598e-06,
+      "loss": 0.9307,
+      "step": 2890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4299686870736052,
+      "learning_rate": 9.382930790291426e-06,
+      "loss": 0.8664,
+      "step": 2891
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.064145907013079,
+      "learning_rate": 9.382431891893957e-06,
+      "loss": 0.8074,
+      "step": 2892
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.908285937295387,
+      "learning_rate": 9.381932805173628e-06,
+      "loss": 0.8948,
+      "step": 2893
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5069776344808536,
+      "learning_rate": 9.381433530151887e-06,
+      "loss": 0.897,
+      "step": 2894
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2163721465985904,
+      "learning_rate": 9.380934066850191e-06,
+      "loss": 0.7886,
+      "step": 2895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2922457857536114,
+      "learning_rate": 9.38043441529e-06,
+      "loss": 0.9363,
+      "step": 2896
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.402514726132862,
+      "learning_rate": 9.379934575492788e-06,
+      "loss": 0.667,
+      "step": 2897
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8725500120517986,
+      "learning_rate": 9.37943454748003e-06,
+      "loss": 0.84,
+      "step": 2898
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.264232361810781,
+      "learning_rate": 9.37893433127322e-06,
+      "loss": 0.6794,
+      "step": 2899
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2255980668024982,
+      "learning_rate": 9.378433926893845e-06,
+      "loss": 0.6368,
+      "step": 2900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.309047410134313,
+      "learning_rate": 9.377933334363416e-06,
+      "loss": 1.0102,
+      "step": 2901
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.216389891473012,
+      "learning_rate": 9.377432553703439e-06,
+      "loss": 0.9816,
+      "step": 2902
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.095495056233268,
+      "learning_rate": 9.376931584935439e-06,
+      "loss": 0.7248,
+      "step": 2903
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9815619535899511,
+      "learning_rate": 9.376430428080939e-06,
+      "loss": 0.8704,
+      "step": 2904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.656996307984321,
+      "learning_rate": 9.375929083161475e-06,
+      "loss": 0.9433,
+      "step": 2905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2990741516498554,
+      "learning_rate": 9.375427550198595e-06,
+      "loss": 0.6037,
+      "step": 2906
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.323890020840218,
+      "learning_rate": 9.374925829213848e-06,
+      "loss": 1.0348,
+      "step": 2907
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.49229944835517,
+      "learning_rate": 9.374423920228795e-06,
+      "loss": 0.787,
+      "step": 2908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.417640846648907,
+      "learning_rate": 9.373921823265004e-06,
+      "loss": 0.8086,
+      "step": 2909
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.079806073793901,
+      "learning_rate": 9.373419538344049e-06,
+      "loss": 1.0091,
+      "step": 2910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.764305824098615,
+      "learning_rate": 9.372917065487518e-06,
+      "loss": 0.6941,
+      "step": 2911
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6865652129520003,
+      "learning_rate": 9.372414404717001e-06,
+      "loss": 0.9306,
+      "step": 2912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2778901183465345,
+      "learning_rate": 9.3719115560541e-06,
+      "loss": 0.8932,
+      "step": 2913
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.199941386836545,
+      "learning_rate": 9.371408519520421e-06,
+      "loss": 1.0005,
+      "step": 2914
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2957402974264927,
+      "learning_rate": 9.370905295137584e-06,
+      "loss": 0.8068,
+      "step": 2915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.451858117162484,
+      "learning_rate": 9.370401882927212e-06,
+      "loss": 0.9157,
+      "step": 2916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.465525495840742,
+      "learning_rate": 9.369898282910937e-06,
+      "loss": 0.668,
+      "step": 2917
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2600312009390704,
+      "learning_rate": 9.369394495110401e-06,
+      "loss": 0.8861,
+      "step": 2918
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4985968035116377,
+      "learning_rate": 9.36889051954725e-06,
+      "loss": 1.0123,
+      "step": 2919
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.477898384655794,
+      "learning_rate": 9.368386356243148e-06,
+      "loss": 0.8705,
+      "step": 2920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5810637891888533,
+      "learning_rate": 9.367882005219752e-06,
+      "loss": 0.7711,
+      "step": 2921
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.760910743469657,
+      "learning_rate": 9.367377466498739e-06,
+      "loss": 0.8858,
+      "step": 2922
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.40133117106626,
+      "learning_rate": 9.36687274010179e-06,
+      "loss": 0.7928,
+      "step": 2923
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.054688435553047,
+      "learning_rate": 9.366367826050593e-06,
+      "loss": 0.9641,
+      "step": 2924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1935119899027928,
+      "learning_rate": 9.365862724366848e-06,
+      "loss": 0.6711,
+      "step": 2925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9085284200528834,
+      "learning_rate": 9.365357435072255e-06,
+      "loss": 0.9037,
+      "step": 2926
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.659756018721163,
+      "learning_rate": 9.364851958188534e-06,
+      "loss": 0.9305,
+      "step": 2927
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.029053231658149,
+      "learning_rate": 9.364346293737403e-06,
+      "loss": 0.933,
+      "step": 2928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.272799500996194,
+      "learning_rate": 9.36384044174059e-06,
+      "loss": 0.7234,
+      "step": 2929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4055036980467226,
+      "learning_rate": 9.363334402219836e-06,
+      "loss": 0.8465,
+      "step": 2930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1156346211735413,
+      "learning_rate": 9.362828175196884e-06,
+      "loss": 0.7269,
+      "step": 2931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1712820832310955,
+      "learning_rate": 9.362321760693489e-06,
+      "loss": 0.9448,
+      "step": 2932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4324207232287776,
+      "learning_rate": 9.361815158731413e-06,
+      "loss": 0.6733,
+      "step": 2933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0674176379529734,
+      "learning_rate": 9.361308369332426e-06,
+      "loss": 0.8544,
+      "step": 2934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6544913136594053,
+      "learning_rate": 9.360801392518303e-06,
+      "loss": 0.8991,
+      "step": 2935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8817242899465179,
+      "learning_rate": 9.360294228310834e-06,
+      "loss": 0.7861,
+      "step": 2936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.077440163848829,
+      "learning_rate": 9.359786876731811e-06,
+      "loss": 0.7734,
+      "step": 2937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1057635167086772,
+      "learning_rate": 9.359279337803036e-06,
+      "loss": 0.7336,
+      "step": 2938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5391350585038786,
+      "learning_rate": 9.358771611546319e-06,
+      "loss": 0.8889,
+      "step": 2939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7705734085262526,
+      "learning_rate": 9.358263697983479e-06,
+      "loss": 0.7793,
+      "step": 2940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.983756236293712,
+      "learning_rate": 9.35775559713634e-06,
+      "loss": 0.9192,
+      "step": 2941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.268052568832443,
+      "learning_rate": 9.357247309026738e-06,
+      "loss": 0.9931,
+      "step": 2942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9762557474068334,
+      "learning_rate": 9.356738833676517e-06,
+      "loss": 0.8458,
+      "step": 2943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.343312571328739,
+      "learning_rate": 9.356230171107524e-06,
+      "loss": 0.858,
+      "step": 2944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8749000941413494,
+      "learning_rate": 9.355721321341618e-06,
+      "loss": 0.6795,
+      "step": 2945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2487874992373262,
+      "learning_rate": 9.355212284400668e-06,
+      "loss": 0.7168,
+      "step": 2946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.391033018205487,
+      "learning_rate": 9.354703060306546e-06,
+      "loss": 0.7397,
+      "step": 2947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.202389960155977,
+      "learning_rate": 9.354193649081134e-06,
+      "loss": 0.8906,
+      "step": 2948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.64687550484835,
+      "learning_rate": 9.353684050746323e-06,
+      "loss": 0.7306,
+      "step": 2949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2974982815894025,
+      "learning_rate": 9.353174265324014e-06,
+      "loss": 1.0237,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.114386915764116,
+      "learning_rate": 9.35266429283611e-06,
+      "loss": 0.6981,
+      "step": 2951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3712244855431366,
+      "learning_rate": 9.352154133304528e-06,
+      "loss": 0.9797,
+      "step": 2952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0423535810098588,
+      "learning_rate": 9.351643786751191e-06,
+      "loss": 0.9056,
+      "step": 2953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.268743154624215,
+      "learning_rate": 9.351133253198027e-06,
+      "loss": 0.7914,
+      "step": 2954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0808841004723877,
+      "learning_rate": 9.350622532666979e-06,
+      "loss": 0.6772,
+      "step": 2955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.782917946663232,
+      "learning_rate": 9.350111625179991e-06,
+      "loss": 0.9098,
+      "step": 2956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8422236284459959,
+      "learning_rate": 9.349600530759018e-06,
+      "loss": 0.8263,
+      "step": 2957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.149394726389543,
+      "learning_rate": 9.349089249426022e-06,
+      "loss": 0.7838,
+      "step": 2958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4150065343751748,
+      "learning_rate": 9.348577781202976e-06,
+      "loss": 0.8837,
+      "step": 2959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.092609825682453,
+      "learning_rate": 9.348066126111857e-06,
+      "loss": 0.9046,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2473676369175757,
+      "learning_rate": 9.347554284174654e-06,
+      "loss": 0.6987,
+      "step": 2961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2812308790714535,
+      "learning_rate": 9.347042255413362e-06,
+      "loss": 0.7329,
+      "step": 2962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3683537903467595,
+      "learning_rate": 9.34653003984998e-06,
+      "loss": 0.8623,
+      "step": 2963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.332998854868424,
+      "learning_rate": 9.346017637506523e-06,
+      "loss": 0.8222,
+      "step": 2964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.282354213392823,
+      "learning_rate": 9.34550504840501e-06,
+      "loss": 0.9398,
+      "step": 2965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.981303269433902,
+      "learning_rate": 9.344992272567467e-06,
+      "loss": 0.7511,
+      "step": 2966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7302159168694904,
+      "learning_rate": 9.344479310015928e-06,
+      "loss": 0.732,
+      "step": 2967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2616983863575255,
+      "learning_rate": 9.343966160772438e-06,
+      "loss": 0.8429,
+      "step": 2968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7479991153708014,
+      "learning_rate": 9.343452824859048e-06,
+      "loss": 1.0182,
+      "step": 2969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.192511950165055,
+      "learning_rate": 9.342939302297816e-06,
+      "loss": 0.7558,
+      "step": 2970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3616837079521615,
+      "learning_rate": 9.34242559311081e-06,
+      "loss": 0.8822,
+      "step": 2971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.296253766476751,
+      "learning_rate": 9.341911697320105e-06,
+      "loss": 0.7205,
+      "step": 2972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.252504512874202,
+      "learning_rate": 9.341397614947785e-06,
+      "loss": 0.8395,
+      "step": 2973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9703361063996419,
+      "learning_rate": 9.340883346015941e-06,
+      "loss": 0.8024,
+      "step": 2974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.364797718265642,
+      "learning_rate": 9.340368890546672e-06,
+      "loss": 0.6902,
+      "step": 2975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3785177762233958,
+      "learning_rate": 9.339854248562084e-06,
+      "loss": 0.9213,
+      "step": 2976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.763831053258426,
+      "learning_rate": 9.339339420084295e-06,
+      "loss": 0.8838,
+      "step": 2977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0259835071171333,
+      "learning_rate": 9.338824405135425e-06,
+      "loss": 0.8333,
+      "step": 2978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5176099500954816,
+      "learning_rate": 9.338309203737609e-06,
+      "loss": 0.7288,
+      "step": 2979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.569762395707266,
+      "learning_rate": 9.337793815912984e-06,
+      "loss": 0.9377,
+      "step": 2980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2980992811859164,
+      "learning_rate": 9.337278241683698e-06,
+      "loss": 0.7468,
+      "step": 2981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2736237233084524,
+      "learning_rate": 9.336762481071906e-06,
+      "loss": 0.9624,
+      "step": 2982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2832675713333597,
+      "learning_rate": 9.336246534099772e-06,
+      "loss": 0.9153,
+      "step": 2983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5748447567814736,
+      "learning_rate": 9.335730400789466e-06,
+      "loss": 0.7845,
+      "step": 2984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5498148211251714,
+      "learning_rate": 9.335214081163171e-06,
+      "loss": 0.7702,
+      "step": 2985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9337721133910124,
+      "learning_rate": 9.33469757524307e-06,
+      "loss": 0.832,
+      "step": 2986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.881438105681414,
+      "learning_rate": 9.33418088305136e-06,
+      "loss": 0.7689,
+      "step": 2987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.069491155997431,
+      "learning_rate": 9.333664004610246e-06,
+      "loss": 0.8721,
+      "step": 2988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5134370185761647,
+      "learning_rate": 9.333146939941938e-06,
+      "loss": 0.9661,
+      "step": 2989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7595789664193435,
+      "learning_rate": 9.332629689068655e-06,
+      "loss": 0.9774,
+      "step": 2990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2055150117730764,
+      "learning_rate": 9.332112252012625e-06,
+      "loss": 1.0155,
+      "step": 2991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1456498627817728,
+      "learning_rate": 9.331594628796083e-06,
+      "loss": 0.6246,
+      "step": 2992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7036667521405278,
+      "learning_rate": 9.331076819441273e-06,
+      "loss": 0.806,
+      "step": 2993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4269633327301547,
+      "learning_rate": 9.330558823970448e-06,
+      "loss": 0.8413,
+      "step": 2994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5954700553627235,
+      "learning_rate": 9.330040642405864e-06,
+      "loss": 0.7597,
+      "step": 2995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0576597044985556,
+      "learning_rate": 9.329522274769791e-06,
+      "loss": 0.8026,
+      "step": 2996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.750192380813906,
+      "learning_rate": 9.329003721084501e-06,
+      "loss": 0.9404,
+      "step": 2997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1727522865784163,
+      "learning_rate": 9.328484981372283e-06,
+      "loss": 0.7567,
+      "step": 2998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.887387558281211,
+      "learning_rate": 9.327966055655424e-06,
+      "loss": 0.9157,
+      "step": 2999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.73083452692304,
+      "learning_rate": 9.327446943956224e-06,
+      "loss": 0.9025,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1546995018966104,
+      "learning_rate": 9.326927646296992e-06,
+      "loss": 0.9286,
+      "step": 3001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8899078478462918,
+      "learning_rate": 9.326408162700043e-06,
+      "loss": 1.0669,
+      "step": 3002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6084068543347034,
+      "learning_rate": 9.325888493187699e-06,
+      "loss": 0.7292,
+      "step": 3003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1819731465850898,
+      "learning_rate": 9.325368637782292e-06,
+      "loss": 0.6136,
+      "step": 3004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.061552638293909,
+      "learning_rate": 9.324848596506162e-06,
+      "loss": 0.91,
+      "step": 3005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.17246315842031,
+      "learning_rate": 9.324328369381654e-06,
+      "loss": 0.9628,
+      "step": 3006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5391769921254195,
+      "learning_rate": 9.323807956431128e-06,
+      "loss": 0.7429,
+      "step": 3007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1594396924985624,
+      "learning_rate": 9.323287357676944e-06,
+      "loss": 0.9848,
+      "step": 3008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5476513397416185,
+      "learning_rate": 9.322766573141473e-06,
+      "loss": 0.7968,
+      "step": 3009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8314490975670894,
+      "learning_rate": 9.322245602847094e-06,
+      "loss": 0.8502,
+      "step": 3010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.674638688115942,
+      "learning_rate": 9.321724446816197e-06,
+      "loss": 0.9852,
+      "step": 3011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2824450834946055,
+      "learning_rate": 9.321203105071176e-06,
+      "loss": 0.7407,
+      "step": 3012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1251419106509277,
+      "learning_rate": 9.320681577634432e-06,
+      "loss": 0.7616,
+      "step": 3013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0864610780653017,
+      "learning_rate": 9.320159864528378e-06,
+      "loss": 1.0028,
+      "step": 3014
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.188617883749553,
+      "learning_rate": 9.319637965775432e-06,
+      "loss": 0.8787,
+      "step": 3015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0689207920520314,
+      "learning_rate": 9.319115881398024e-06,
+      "loss": 0.7112,
+      "step": 3016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.205042039005268,
+      "learning_rate": 9.31859361141859e-06,
+      "loss": 0.9498,
+      "step": 3017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8134949336720871,
+      "learning_rate": 9.318071155859567e-06,
+      "loss": 0.8158,
+      "step": 3018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.715745543710717,
+      "learning_rate": 9.31754851474341e-06,
+      "loss": 0.9997,
+      "step": 3019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3111719578365812,
+      "learning_rate": 9.317025688092577e-06,
+      "loss": 0.7327,
+      "step": 3020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3826059117339546,
+      "learning_rate": 9.316502675929537e-06,
+      "loss": 0.7617,
+      "step": 3021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.415764895207329,
+      "learning_rate": 9.315979478276763e-06,
+      "loss": 0.7299,
+      "step": 3022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2734768227981839,
+      "learning_rate": 9.315456095156739e-06,
+      "loss": 0.6619,
+      "step": 3023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.512694399528168,
+      "learning_rate": 9.314932526591956e-06,
+      "loss": 0.8963,
+      "step": 3024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8743831482117477,
+      "learning_rate": 9.314408772604911e-06,
+      "loss": 0.9882,
+      "step": 3025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.287591039402503,
+      "learning_rate": 9.313884833218115e-06,
+      "loss": 0.8868,
+      "step": 3026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.538885622872133,
+      "learning_rate": 9.313360708454078e-06,
+      "loss": 0.8222,
+      "step": 3027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.270412917230198,
+      "learning_rate": 9.312836398335326e-06,
+      "loss": 0.9843,
+      "step": 3028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1739595432304937,
+      "learning_rate": 9.312311902884388e-06,
+      "loss": 0.7755,
+      "step": 3029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1098355428751567,
+      "learning_rate": 9.311787222123803e-06,
+      "loss": 0.8943,
+      "step": 3030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1091942024895047,
+      "learning_rate": 9.311262356076118e-06,
+      "loss": 0.8845,
+      "step": 3031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.157632588842543,
+      "learning_rate": 9.31073730476389e-06,
+      "loss": 1.0056,
+      "step": 3032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1030228619517715,
+      "learning_rate": 9.310212068209678e-06,
+      "loss": 0.6382,
+      "step": 3033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.183802238307822,
+      "learning_rate": 9.309686646436053e-06,
+      "loss": 0.5932,
+      "step": 3034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.211735027612899,
+      "learning_rate": 9.309161039465597e-06,
+      "loss": 0.8535,
+      "step": 3035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2218865200814635,
+      "learning_rate": 9.308635247320893e-06,
+      "loss": 0.6003,
+      "step": 3036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4125026403192136,
+      "learning_rate": 9.308109270024537e-06,
+      "loss": 0.8924,
+      "step": 3037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1007519356124797,
+      "learning_rate": 9.30758310759913e-06,
+      "loss": 0.7364,
+      "step": 3038
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1977770175330424,
+      "learning_rate": 9.307056760067284e-06,
+      "loss": 1.0086,
+      "step": 3039
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.782163406115625,
+      "learning_rate": 9.306530227451616e-06,
+      "loss": 0.8719,
+      "step": 3040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.50638276748778,
+      "learning_rate": 9.306003509774755e-06,
+      "loss": 0.7643,
+      "step": 3041
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4097021291888945,
+      "learning_rate": 9.305476607059332e-06,
+      "loss": 0.7316,
+      "step": 3042
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.053377204637384,
+      "learning_rate": 9.30494951932799e-06,
+      "loss": 0.8974,
+      "step": 3043
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9585325752647602,
+      "learning_rate": 9.30442224660338e-06,
+      "loss": 1.0118,
+      "step": 3044
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6881546038005304,
+      "learning_rate": 9.303894788908158e-06,
+      "loss": 1.0489,
+      "step": 3045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9757961227774983,
+      "learning_rate": 9.303367146264995e-06,
+      "loss": 0.7501,
+      "step": 3046
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1093872883642266,
+      "learning_rate": 9.302839318696558e-06,
+      "loss": 0.7373,
+      "step": 3047
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.304834423608636,
+      "learning_rate": 9.302311306225534e-06,
+      "loss": 0.72,
+      "step": 3048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9877842594999602,
+      "learning_rate": 9.301783108874611e-06,
+      "loss": 0.8912,
+      "step": 3049
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9596597253561134,
+      "learning_rate": 9.301254726666488e-06,
+      "loss": 0.8553,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9117437697018276,
+      "learning_rate": 9.300726159623869e-06,
+      "loss": 0.8494,
+      "step": 3051
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8865556169364046,
+      "learning_rate": 9.300197407769472e-06,
+      "loss": 0.8599,
+      "step": 3052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5792441625782843,
+      "learning_rate": 9.299668471126011e-06,
+      "loss": 0.9663,
+      "step": 3053
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1508451714366323,
+      "learning_rate": 9.299139349716221e-06,
+      "loss": 0.7725,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9461439189248284,
+      "learning_rate": 9.29861004356284e-06,
+      "loss": 0.8389,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4409020013751888,
+      "learning_rate": 9.29808055268861e-06,
+      "loss": 0.8939,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.323535326679325,
+      "learning_rate": 9.297550877116288e-06,
+      "loss": 1.1053,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.318079864081508,
+      "learning_rate": 9.297021016868634e-06,
+      "loss": 0.983,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.329101756824021,
+      "learning_rate": 9.296490971968416e-06,
+      "loss": 0.8886,
+      "step": 3059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.354034612227987,
+      "learning_rate": 9.295960742438413e-06,
+      "loss": 0.9249,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.594566662029526,
+      "learning_rate": 9.295430328301406e-06,
+      "loss": 0.876,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6928392688199017,
+      "learning_rate": 9.294899729580196e-06,
+      "loss": 1.0015,
+      "step": 3062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0638012502105583,
+      "learning_rate": 9.294368946297576e-06,
+      "loss": 0.9599,
+      "step": 3063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.628817665654928,
+      "learning_rate": 9.293837978476359e-06,
+      "loss": 0.8943,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.121339517187509,
+      "learning_rate": 9.293306826139361e-06,
+      "loss": 0.8463,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.137809983440236,
+      "learning_rate": 9.292775489309409e-06,
+      "loss": 0.8538,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4385359739944383,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 0.9138,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0924434892194315,
+      "learning_rate": 9.291712262261972e-06,
+      "loss": 0.9365,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0785686688712928,
+      "learning_rate": 9.291180372090178e-06,
+      "loss": 1.0299,
+      "step": 3069
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0799650120718614,
+      "learning_rate": 9.290648297516806e-06,
+      "loss": 1.0316,
+      "step": 3070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9589431263310622,
+      "learning_rate": 9.290116038564722e-06,
+      "loss": 0.7472,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4664699548945217,
+      "learning_rate": 9.2895835952568e-06,
+      "loss": 0.9355,
+      "step": 3072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3873541010712755,
+      "learning_rate": 9.289050967615914e-06,
+      "loss": 0.6563,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0417861379258073,
+      "learning_rate": 9.288518155664956e-06,
+      "loss": 0.8135,
+      "step": 3074
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.912692693498903,
+      "learning_rate": 9.28798515942682e-06,
+      "loss": 0.7543,
+      "step": 3075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2523583338532245,
+      "learning_rate": 9.287451978924415e-06,
+      "loss": 0.7912,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.877263713771907,
+      "learning_rate": 9.28691861418065e-06,
+      "loss": 1.0484,
+      "step": 3077
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6090836147048853,
+      "learning_rate": 9.286385065218442e-06,
+      "loss": 0.8387,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8104260116184343,
+      "learning_rate": 9.285851332060722e-06,
+      "loss": 0.9194,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0135272816574443,
+      "learning_rate": 9.285317414730427e-06,
+      "loss": 0.9149,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7404638169297733,
+      "learning_rate": 9.284783313250499e-06,
+      "loss": 0.6906,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7898351218962485,
+      "learning_rate": 9.284249027643887e-06,
+      "loss": 0.8902,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0068132315763547,
+      "learning_rate": 9.283714557933554e-06,
+      "loss": 0.6324,
+      "step": 3083
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7043813844293614,
+      "learning_rate": 9.283179904142465e-06,
+      "loss": 0.9488,
+      "step": 3084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1298148600597657,
+      "learning_rate": 9.282645066293597e-06,
+      "loss": 0.8834,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.398837675731012,
+      "learning_rate": 9.282110044409933e-06,
+      "loss": 0.9028,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.132039844870747,
+      "learning_rate": 9.281574838514464e-06,
+      "loss": 0.7114,
+      "step": 3087
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.078316042509498,
+      "learning_rate": 9.281039448630187e-06,
+      "loss": 0.6646,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5061317663808755,
+      "learning_rate": 9.280503874780112e-06,
+      "loss": 1.0707,
+      "step": 3089
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.107620624968619,
+      "learning_rate": 9.279968116987253e-06,
+      "loss": 0.713,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.146427095603013,
+      "learning_rate": 9.27943217527463e-06,
+      "loss": 0.877,
+      "step": 3091
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2778563613918728,
+      "learning_rate": 9.278896049665276e-06,
+      "loss": 0.6553,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.843863601750405,
+      "learning_rate": 9.27835974018223e-06,
+      "loss": 0.772,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9432911775593196,
+      "learning_rate": 9.277823246848537e-06,
+      "loss": 0.7441,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.06523677523482,
+      "learning_rate": 9.277286569687252e-06,
+      "loss": 0.7993,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1465172181168413,
+      "learning_rate": 9.276749708721438e-06,
+      "loss": 0.9183,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7768868675174252,
+      "learning_rate": 9.276212663974163e-06,
+      "loss": 1.0552,
+      "step": 3097
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4281492412735126,
+      "learning_rate": 9.275675435468505e-06,
+      "loss": 0.8241,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8336682995887392,
+      "learning_rate": 9.275138023227555e-06,
+      "loss": 0.9489,
+      "step": 3099
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.451298733408649,
+      "learning_rate": 9.274600427274399e-06,
+      "loss": 0.8779,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1350929852654548,
+      "learning_rate": 9.274062647632144e-06,
+      "loss": 0.6837,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.398040477930822,
+      "learning_rate": 9.273524684323897e-06,
+      "loss": 0.9319,
+      "step": 3102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2463963446925506,
+      "learning_rate": 9.27298653737278e-06,
+      "loss": 0.7784,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6749790478600333,
+      "learning_rate": 9.272448206801912e-06,
+      "loss": 0.9542,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6426001028034882,
+      "learning_rate": 9.271909692634431e-06,
+      "loss": 1.0737,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9098641337918776,
+      "learning_rate": 9.271370994893475e-06,
+      "loss": 0.6532,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2270653606546769,
+      "learning_rate": 9.270832113602196e-06,
+      "loss": 0.7394,
+      "step": 3107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0543635290588895,
+      "learning_rate": 9.270293048783747e-06,
+      "loss": 0.942,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0633034687192706,
+      "learning_rate": 9.269753800461299e-06,
+      "loss": 0.694,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0939423022093524,
+      "learning_rate": 9.269214368658016e-06,
+      "loss": 0.7491,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.435381878393438,
+      "learning_rate": 9.268674753397087e-06,
+      "loss": 1.2197,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9258095919399216,
+      "learning_rate": 9.268134954701697e-06,
+      "loss": 0.6953,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.062140110064109,
+      "learning_rate": 9.267594972595042e-06,
+      "loss": 0.9525,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4758249291144265,
+      "learning_rate": 9.267054807100327e-06,
+      "loss": 0.8391,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2361500156610385,
+      "learning_rate": 9.266514458240762e-06,
+      "loss": 0.7616,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8891525795254904,
+      "learning_rate": 9.26597392603957e-06,
+      "loss": 0.5758,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.685933638223758,
+      "learning_rate": 9.265433210519976e-06,
+      "loss": 0.8936,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2649416150182873,
+      "learning_rate": 9.26489231170522e-06,
+      "loss": 0.8143,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.164667051327688,
+      "learning_rate": 9.264351229618541e-06,
+      "loss": 0.8021,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4303282162776236,
+      "learning_rate": 9.263809964283192e-06,
+      "loss": 0.9052,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.321104715139511,
+      "learning_rate": 9.263268515722435e-06,
+      "loss": 0.7605,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3320801599138807,
+      "learning_rate": 9.262726883959535e-06,
+      "loss": 0.9848,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1475504236186698,
+      "learning_rate": 9.262185069017764e-06,
+      "loss": 1.0068,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.103957883560593,
+      "learning_rate": 9.261643070920409e-06,
+      "loss": 0.9667,
+      "step": 3124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4126083073987075,
+      "learning_rate": 9.261100889690763e-06,
+      "loss": 1.0358,
+      "step": 3125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0413687238101645,
+      "learning_rate": 9.26055852535212e-06,
+      "loss": 0.7515,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.540833343882218,
+      "learning_rate": 9.260015977927788e-06,
+      "loss": 0.7905,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2698230055494302,
+      "learning_rate": 9.259473247441081e-06,
+      "loss": 0.6909,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.047482230387887,
+      "learning_rate": 9.258930333915325e-06,
+      "loss": 0.8081,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6667029868475867,
+      "learning_rate": 9.258387237373845e-06,
+      "loss": 0.9366,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3459623948165724,
+      "learning_rate": 9.257843957839982e-06,
+      "loss": 0.8387,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.165563471330127,
+      "learning_rate": 9.257300495337082e-06,
+      "loss": 0.6749,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.287371436129912,
+      "learning_rate": 9.256756849888499e-06,
+      "loss": 0.7394,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2225275573954857,
+      "learning_rate": 9.256213021517593e-06,
+      "loss": 0.8595,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2678608923028207,
+      "learning_rate": 9.255669010247734e-06,
+      "loss": 0.8517,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.81168131391765,
+      "learning_rate": 9.2551248161023e-06,
+      "loss": 0.9586,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0713514240477764,
+      "learning_rate": 9.254580439104677e-06,
+      "loss": 0.6868,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1525754844333511,
+      "learning_rate": 9.254035879278255e-06,
+      "loss": 0.6873,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0040789534063617,
+      "learning_rate": 9.253491136646437e-06,
+      "loss": 0.9364,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3557537067753924,
+      "learning_rate": 9.252946211232634e-06,
+      "loss": 0.8945,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2662810568523244,
+      "learning_rate": 9.25240110306026e-06,
+      "loss": 0.9354,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0945805737848486,
+      "learning_rate": 9.25185581215274e-06,
+      "loss": 0.8668,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.119809577806173,
+      "learning_rate": 9.251310338533504e-06,
+      "loss": 0.7942,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.264230767450925,
+      "learning_rate": 9.250764682225997e-06,
+      "loss": 0.7298,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.75546707705692,
+      "learning_rate": 9.250218843253663e-06,
+      "loss": 0.9563,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.140139774243515,
+      "learning_rate": 9.24967282163996e-06,
+      "loss": 0.7914,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.621663430825466,
+      "learning_rate": 9.249126617408353e-06,
+      "loss": 0.8191,
+      "step": 3147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4178521662105377,
+      "learning_rate": 9.248580230582311e-06,
+      "loss": 0.8392,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4549236175245426,
+      "learning_rate": 9.248033661185313e-06,
+      "loss": 0.6883,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1268473650186763,
+      "learning_rate": 9.247486909240849e-06,
+      "loss": 0.7434,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.001961350782244,
+      "learning_rate": 9.246939974772413e-06,
+      "loss": 0.9213,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2685532841490264,
+      "learning_rate": 9.246392857803508e-06,
+      "loss": 0.7419,
+      "step": 3152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1823412870814247,
+      "learning_rate": 9.245845558357643e-06,
+      "loss": 0.6406,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.086698655649945,
+      "learning_rate": 9.24529807645834e-06,
+      "loss": 0.9058,
+      "step": 3154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.487035456316561,
+      "learning_rate": 9.244750412129123e-06,
+      "loss": 0.7803,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.571866476373517,
+      "learning_rate": 9.244202565393528e-06,
+      "loss": 0.7954,
+      "step": 3156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.223530337689157,
+      "learning_rate": 9.243654536275095e-06,
+      "loss": 1.0453,
+      "step": 3157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9108252756668063,
+      "learning_rate": 9.243106324797377e-06,
+      "loss": 0.8244,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.999599491605971,
+      "learning_rate": 9.24255793098393e-06,
+      "loss": 0.847,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.034536794065231,
+      "learning_rate": 9.24200935485832e-06,
+      "loss": 0.6592,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8949946472723926,
+      "learning_rate": 9.24146059644412e-06,
+      "loss": 0.8063,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2888708115265737,
+      "learning_rate": 9.240911655764912e-06,
+      "loss": 0.6856,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3972330147092227,
+      "learning_rate": 9.240362532844287e-06,
+      "loss": 0.6912,
+      "step": 3163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.844453609477797,
+      "learning_rate": 9.23981322770584e-06,
+      "loss": 0.8547,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1843967635493393,
+      "learning_rate": 9.239263740373173e-06,
+      "loss": 0.7897,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.065957769773524,
+      "learning_rate": 9.238714070869903e-06,
+      "loss": 0.9008,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3004270977978396,
+      "learning_rate": 9.238164219219649e-06,
+      "loss": 0.7672,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.742978525843792,
+      "learning_rate": 9.23761418544604e-06,
+      "loss": 0.7691,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.046004091634552,
+      "learning_rate": 9.237063969572713e-06,
+      "loss": 0.856,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.229223758555974,
+      "learning_rate": 9.23651357162331e-06,
+      "loss": 0.7553,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.278240283666766,
+      "learning_rate": 9.235962991621484e-06,
+      "loss": 0.8792,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0260316691908615,
+      "learning_rate": 9.235412229590893e-06,
+      "loss": 0.9525,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0205808757073958,
+      "learning_rate": 9.234861285555207e-06,
+      "loss": 0.8204,
+      "step": 3173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.31632198952091,
+      "learning_rate": 9.2343101595381e-06,
+      "loss": 1.0654,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0143161967645247,
+      "learning_rate": 9.233758851563254e-06,
+      "loss": 0.6504,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.829847554158129,
+      "learning_rate": 9.233207361654362e-06,
+      "loss": 0.8252,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2428639131043155,
+      "learning_rate": 9.23265568983512e-06,
+      "loss": 0.6526,
+      "step": 3177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2508839575906467,
+      "learning_rate": 9.232103836129239e-06,
+      "loss": 0.6166,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0277291574448695,
+      "learning_rate": 9.23155180056043e-06,
+      "loss": 0.8885,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1890018778997185,
+      "learning_rate": 9.230999583152413e-06,
+      "loss": 1.0171,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.248207559214574,
+      "learning_rate": 9.230447183928926e-06,
+      "loss": 0.8388,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1532012591682417,
+      "learning_rate": 9.229894602913698e-06,
+      "loss": 0.7939,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5195724708638356,
+      "learning_rate": 9.22934184013048e-06,
+      "loss": 0.8163,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3459764948046913,
+      "learning_rate": 9.228788895603024e-06,
+      "loss": 0.7416,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4020758683924806,
+      "learning_rate": 9.22823576935509e-06,
+      "loss": 0.8897,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1571508255980953,
+      "learning_rate": 9.22768246141045e-06,
+      "loss": 1.0107,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1585040531041524,
+      "learning_rate": 9.227128971792877e-06,
+      "loss": 0.8587,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.58779343936264,
+      "learning_rate": 9.226575300526159e-06,
+      "loss": 0.9522,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0540886491097932,
+      "learning_rate": 9.226021447634085e-06,
+      "loss": 0.6789,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1640529621932139,
+      "learning_rate": 9.22546741314046e-06,
+      "loss": 0.6742,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.575673711504912,
+      "learning_rate": 9.22491319706909e-06,
+      "loss": 0.9315,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3249990516952797,
+      "learning_rate": 9.224358799443791e-06,
+      "loss": 0.7997,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6466035551090896,
+      "learning_rate": 9.223804220288384e-06,
+      "loss": 0.834,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.261258638790184,
+      "learning_rate": 9.223249459626704e-06,
+      "loss": 0.8495,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2021351433463756,
+      "learning_rate": 9.22269451748259e-06,
+      "loss": 0.6719,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3592365472995698,
+      "learning_rate": 9.222139393879885e-06,
+      "loss": 1.0129,
+      "step": 3196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3880224348065957,
+      "learning_rate": 9.22158408884245e-06,
+      "loss": 0.6666,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.307563685352974,
+      "learning_rate": 9.221028602394144e-06,
+      "loss": 0.9835,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8943651309061855,
+      "learning_rate": 9.220472934558838e-06,
+      "loss": 1.0338,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5019233581242033,
+      "learning_rate": 9.21991708536041e-06,
+      "loss": 0.9794,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2364754484911855,
+      "learning_rate": 9.219361054822746e-06,
+      "loss": 0.9367,
+      "step": 3201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2704533284734194,
+      "learning_rate": 9.218804842969742e-06,
+      "loss": 1.0065,
+      "step": 3202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9802121677451474,
+      "learning_rate": 9.218248449825297e-06,
+      "loss": 0.7721,
+      "step": 3203
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6581788183045583,
+      "learning_rate": 9.217691875413323e-06,
+      "loss": 0.8738,
+      "step": 3204
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.023895750559744,
+      "learning_rate": 9.217135119757734e-06,
+      "loss": 1.0308,
+      "step": 3205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.949327044317688,
+      "learning_rate": 9.216578182882459e-06,
+      "loss": 0.8826,
+      "step": 3206
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.480186703599764,
+      "learning_rate": 9.216021064811428e-06,
+      "loss": 0.9375,
+      "step": 3207
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9787724100881925,
+      "learning_rate": 9.21546376556858e-06,
+      "loss": 0.8179,
+      "step": 3208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.238878144113274,
+      "learning_rate": 9.214906285177867e-06,
+      "loss": 0.9119,
+      "step": 3209
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4141873801224556,
+      "learning_rate": 9.214348623663244e-06,
+      "loss": 0.7741,
+      "step": 3210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.162823512040554,
+      "learning_rate": 9.213790781048677e-06,
+      "loss": 1.009,
+      "step": 3211
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9838807254970567,
+      "learning_rate": 9.213232757358133e-06,
+      "loss": 0.8248,
+      "step": 3212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6914614969378676,
+      "learning_rate": 9.212674552615594e-06,
+      "loss": 0.9125,
+      "step": 3213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.309803166268253,
+      "learning_rate": 9.212116166845048e-06,
+      "loss": 0.801,
+      "step": 3214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2189334316223115,
+      "learning_rate": 9.211557600070489e-06,
+      "loss": 0.8757,
+      "step": 3215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.204920331911715,
+      "learning_rate": 9.21099885231592e-06,
+      "loss": 1.0098,
+      "step": 3216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.620039317672361,
+      "learning_rate": 9.210439923605352e-06,
+      "loss": 0.8951,
+      "step": 3217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0691455810987547,
+      "learning_rate": 9.209880813962806e-06,
+      "loss": 0.8375,
+      "step": 3218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.98763789625787,
+      "learning_rate": 9.209321523412303e-06,
+      "loss": 0.8434,
+      "step": 3219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.363308773518401,
+      "learning_rate": 9.208762051977879e-06,
+      "loss": 0.8568,
+      "step": 3220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.509361920941755,
+      "learning_rate": 9.208202399683577e-06,
+      "loss": 0.802,
+      "step": 3221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9916089491651825,
+      "learning_rate": 9.207642566553445e-06,
+      "loss": 1.0208,
+      "step": 3222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.703128237642076,
+      "learning_rate": 9.207082552611541e-06,
+      "loss": 0.925,
+      "step": 3223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.2395740230638355,
+      "learning_rate": 9.206522357881931e-06,
+      "loss": 0.8509,
+      "step": 3224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1660277164318034,
+      "learning_rate": 9.205961982388686e-06,
+      "loss": 0.9124,
+      "step": 3225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.81052282500674,
+      "learning_rate": 9.205401426155889e-06,
+      "loss": 0.7709,
+      "step": 3226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.734589971025906,
+      "learning_rate": 9.204840689207626e-06,
+      "loss": 1.0335,
+      "step": 3227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9810349807110708,
+      "learning_rate": 9.204279771567994e-06,
+      "loss": 0.9054,
+      "step": 3228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.149123751873147,
+      "learning_rate": 9.203718673261098e-06,
+      "loss": 0.658,
+      "step": 3229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2355361970752123,
+      "learning_rate": 9.203157394311045e-06,
+      "loss": 0.8467,
+      "step": 3230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7299624972021121,
+      "learning_rate": 9.20259593474196e-06,
+      "loss": 0.8817,
+      "step": 3231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.155346463928962,
+      "learning_rate": 9.202034294577968e-06,
+      "loss": 0.8624,
+      "step": 3232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.342763397544821,
+      "learning_rate": 9.201472473843204e-06,
+      "loss": 0.7816,
+      "step": 3233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5739138314174084,
+      "learning_rate": 9.20091047256181e-06,
+      "loss": 0.9177,
+      "step": 3234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0438150702366014,
+      "learning_rate": 9.200348290757937e-06,
+      "loss": 0.6408,
+      "step": 3235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.131350688125393,
+      "learning_rate": 9.199785928455745e-06,
+      "loss": 0.9543,
+      "step": 3236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2500124566037876,
+      "learning_rate": 9.199223385679396e-06,
+      "loss": 0.8949,
+      "step": 3237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0929755145029607,
+      "learning_rate": 9.198660662453066e-06,
+      "loss": 1.0035,
+      "step": 3238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9620896500853613,
+      "learning_rate": 9.198097758800938e-06,
+      "loss": 0.9197,
+      "step": 3239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6814500120663143,
+      "learning_rate": 9.197534674747199e-06,
+      "loss": 1.0562,
+      "step": 3240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3074357628110715,
+      "learning_rate": 9.196971410316047e-06,
+      "loss": 0.6652,
+      "step": 3241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9344384907700944,
+      "learning_rate": 9.196407965531686e-06,
+      "loss": 0.5753,
+      "step": 3242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4225312496730593,
+      "learning_rate": 9.195844340418328e-06,
+      "loss": 0.9251,
+      "step": 3243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.213727002183818,
+      "learning_rate": 9.195280535000196e-06,
+      "loss": 1.0034,
+      "step": 3244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7988152236810677,
+      "learning_rate": 9.194716549301514e-06,
+      "loss": 0.7957,
+      "step": 3245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.531880370203156,
+      "learning_rate": 9.19415238334652e-06,
+      "loss": 0.9854,
+      "step": 3246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2697330608253439,
+      "learning_rate": 9.193588037159457e-06,
+      "loss": 0.5929,
+      "step": 3247
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2376417231940116,
+      "learning_rate": 9.193023510764578e-06,
+      "loss": 0.7968,
+      "step": 3248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9180784295556366,
+      "learning_rate": 9.19245880418614e-06,
+      "loss": 1.0015,
+      "step": 3249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1411602300566024,
+      "learning_rate": 9.191893917448409e-06,
+      "loss": 1.0184,
+      "step": 3250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5848311547430427,
+      "learning_rate": 9.19132885057566e-06,
+      "loss": 0.9171,
+      "step": 3251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.285287084791478,
+      "learning_rate": 9.190763603592177e-06,
+      "loss": 0.9391,
+      "step": 3252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.226996108632594,
+      "learning_rate": 9.190198176522249e-06,
+      "loss": 0.9092,
+      "step": 3253
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6108614016801015,
+      "learning_rate": 9.189632569390172e-06,
+      "loss": 0.7254,
+      "step": 3254
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1139427601361627,
+      "learning_rate": 9.189066782220253e-06,
+      "loss": 0.8949,
+      "step": 3255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0317222529103973,
+      "learning_rate": 9.188500815036806e-06,
+      "loss": 0.7609,
+      "step": 3256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.216861054322238,
+      "learning_rate": 9.18793466786415e-06,
+      "loss": 0.7397,
+      "step": 3257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2671349768095517,
+      "learning_rate": 9.187368340726615e-06,
+      "loss": 0.6543,
+      "step": 3258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1907777983647967,
+      "learning_rate": 9.186801833648535e-06,
+      "loss": 1.0772,
+      "step": 3259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.401617078441545,
+      "learning_rate": 9.186235146654257e-06,
+      "loss": 0.8028,
+      "step": 3260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6194101442847595,
+      "learning_rate": 9.18566827976813e-06,
+      "loss": 1.0575,
+      "step": 3261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2501570561741446,
+      "learning_rate": 9.185101233014516e-06,
+      "loss": 0.7671,
+      "step": 3262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9635997028423224,
+      "learning_rate": 9.184534006417782e-06,
+      "loss": 1.0013,
+      "step": 3263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9782311129044745,
+      "learning_rate": 9.183966600002301e-06,
+      "loss": 0.9732,
+      "step": 3264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5523154091152636,
+      "learning_rate": 9.183399013792457e-06,
+      "loss": 0.8528,
+      "step": 3265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.075198213076345,
+      "learning_rate": 9.18283124781264e-06,
+      "loss": 0.9389,
+      "step": 3266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8590625239756609,
+      "learning_rate": 9.18226330208725e-06,
+      "loss": 0.8654,
+      "step": 3267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2433513874563753,
+      "learning_rate": 9.181695176640693e-06,
+      "loss": 0.6937,
+      "step": 3268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8557747584931548,
+      "learning_rate": 9.181126871497378e-06,
+      "loss": 1.0514,
+      "step": 3269
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2463384870546785,
+      "learning_rate": 9.180558386681731e-06,
+      "loss": 0.784,
+      "step": 3270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0228053977879576,
+      "learning_rate": 9.179989722218179e-06,
+      "loss": 0.9257,
+      "step": 3271
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1781017631144275,
+      "learning_rate": 9.17942087813116e-06,
+      "loss": 0.9315,
+      "step": 3272
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.272132759504954,
+      "learning_rate": 9.178851854445118e-06,
+      "loss": 0.937,
+      "step": 3273
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2293300528093227,
+      "learning_rate": 9.178282651184506e-06,
+      "loss": 0.7196,
+      "step": 3274
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.337840374835757,
+      "learning_rate": 9.177713268373781e-06,
+      "loss": 0.7935,
+      "step": 3275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6340271659094636,
+      "learning_rate": 9.177143706037411e-06,
+      "loss": 0.8333,
+      "step": 3276
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.875243356267475,
+      "learning_rate": 9.176573964199875e-06,
+      "loss": 0.9899,
+      "step": 3277
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7289262655101263,
+      "learning_rate": 9.176004042885653e-06,
+      "loss": 0.9404,
+      "step": 3278
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.342480957327379,
+      "learning_rate": 9.175433942119238e-06,
+      "loss": 0.8783,
+      "step": 3279
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.212821764782417,
+      "learning_rate": 9.174863661925128e-06,
+      "loss": 0.9121,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8601733140502232,
+      "learning_rate": 9.174293202327828e-06,
+      "loss": 0.9143,
+      "step": 3281
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.382848293309861,
+      "learning_rate": 9.173722563351852e-06,
+      "loss": 0.8808,
+      "step": 3282
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.160935403157398,
+      "learning_rate": 9.173151745021722e-06,
+      "loss": 0.9246,
+      "step": 3283
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0948901299269915,
+      "learning_rate": 9.172580747361968e-06,
+      "loss": 0.7764,
+      "step": 3284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0912242121050126,
+      "learning_rate": 9.172009570397124e-06,
+      "loss": 0.8715,
+      "step": 3285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4002550809299983,
+      "learning_rate": 9.17143821415174e-06,
+      "loss": 0.7305,
+      "step": 3286
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4701718732282054,
+      "learning_rate": 9.170866678650366e-06,
+      "loss": 0.8156,
+      "step": 3287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3769899763594076,
+      "learning_rate": 9.17029496391756e-06,
+      "loss": 0.9752,
+      "step": 3288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.286265900507053,
+      "learning_rate": 9.169723069977892e-06,
+      "loss": 0.9123,
+      "step": 3289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2709293424114034,
+      "learning_rate": 9.169150996855939e-06,
+      "loss": 0.769,
+      "step": 3290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4634418310839576,
+      "learning_rate": 9.16857874457628e-06,
+      "loss": 0.8951,
+      "step": 3291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.638835316206576,
+      "learning_rate": 9.16800631316351e-06,
+      "loss": 1.0762,
+      "step": 3292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0291404005373206,
+      "learning_rate": 9.167433702642224e-06,
+      "loss": 0.8668,
+      "step": 3293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3047538886730443,
+      "learning_rate": 9.166860913037032e-06,
+      "loss": 0.7882,
+      "step": 3294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.400180446723336,
+      "learning_rate": 9.166287944372546e-06,
+      "loss": 0.7652,
+      "step": 3295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7383261266170256,
+      "learning_rate": 9.165714796673389e-06,
+      "loss": 0.8837,
+      "step": 3296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.334114194338,
+      "learning_rate": 9.16514146996419e-06,
+      "loss": 0.8949,
+      "step": 3297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.657241372617888,
+      "learning_rate": 9.164567964269583e-06,
+      "loss": 0.8203,
+      "step": 3298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.240659370834829,
+      "learning_rate": 9.163994279614218e-06,
+      "loss": 0.6797,
+      "step": 3299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9909140079662708,
+      "learning_rate": 9.163420416022745e-06,
+      "loss": 0.731,
+      "step": 3300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1787606278487868,
+      "learning_rate": 9.162846373519824e-06,
+      "loss": 0.7766,
+      "step": 3301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1843158293117173,
+      "learning_rate": 9.162272152130123e-06,
+      "loss": 0.6615,
+      "step": 3302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1421678511450652,
+      "learning_rate": 9.161697751878317e-06,
+      "loss": 0.8057,
+      "step": 3303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0896599658656068,
+      "learning_rate": 9.161123172789091e-06,
+      "loss": 0.6097,
+      "step": 3304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.805083285184337,
+      "learning_rate": 9.160548414887135e-06,
+      "loss": 0.8676,
+      "step": 3305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.112230701703866,
+      "learning_rate": 9.159973478197148e-06,
+      "loss": 0.9024,
+      "step": 3306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1806149924529934,
+      "learning_rate": 9.159398362743834e-06,
+      "loss": 0.7713,
+      "step": 3307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.162175344852935,
+      "learning_rate": 9.15882306855191e-06,
+      "loss": 0.6973,
+      "step": 3308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4252964170608386,
+      "learning_rate": 9.158247595646098e-06,
+      "loss": 0.9256,
+      "step": 3309
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.962082174213007,
+      "learning_rate": 9.157671944051125e-06,
+      "loss": 0.668,
+      "step": 3310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9633791795228484,
+      "learning_rate": 9.157096113791727e-06,
+      "loss": 0.8803,
+      "step": 3311
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.142344184849117,
+      "learning_rate": 9.156520104892653e-06,
+      "loss": 0.7669,
+      "step": 3312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8674384137135474,
+      "learning_rate": 9.155943917378652e-06,
+      "loss": 0.8144,
+      "step": 3313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8682046639063554,
+      "learning_rate": 9.155367551274485e-06,
+      "loss": 0.8895,
+      "step": 3314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.136809323516493,
+      "learning_rate": 9.15479100660492e-06,
+      "loss": 0.9706,
+      "step": 3315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.345384947633705,
+      "learning_rate": 9.154214283394733e-06,
+      "loss": 0.8833,
+      "step": 3316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4599895611519536,
+      "learning_rate": 9.153637381668705e-06,
+      "loss": 0.8739,
+      "step": 3317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1199010197728887,
+      "learning_rate": 9.153060301451629e-06,
+      "loss": 1.0536,
+      "step": 3318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0569393254480712,
+      "learning_rate": 9.152483042768302e-06,
+      "loss": 0.6046,
+      "step": 3319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3016856686269804,
+      "learning_rate": 9.151905605643531e-06,
+      "loss": 0.8822,
+      "step": 3320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0721496910944697,
+      "learning_rate": 9.151327990102129e-06,
+      "loss": 0.8536,
+      "step": 3321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8555856520212888,
+      "learning_rate": 9.150750196168918e-06,
+      "loss": 0.9354,
+      "step": 3322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2114504515821833,
+      "learning_rate": 9.150172223868727e-06,
+      "loss": 0.8096,
+      "step": 3323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.035522415675031,
+      "learning_rate": 9.149594073226391e-06,
+      "loss": 0.8146,
+      "step": 3324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.181198939746774,
+      "learning_rate": 9.149015744266759e-06,
+      "loss": 0.799,
+      "step": 3325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.134625335755381,
+      "learning_rate": 9.148437237014677e-06,
+      "loss": 1.0709,
+      "step": 3326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.104274522320897,
+      "learning_rate": 9.14785855149501e-06,
+      "loss": 0.8442,
+      "step": 3327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8414596909399614,
+      "learning_rate": 9.147279687732622e-06,
+      "loss": 0.8517,
+      "step": 3328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1588492029095045,
+      "learning_rate": 9.14670064575239e-06,
+      "loss": 0.8338,
+      "step": 3329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.951289963404498,
+      "learning_rate": 9.146121425579197e-06,
+      "loss": 1.105,
+      "step": 3330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1361264097796726,
+      "learning_rate": 9.14554202723793e-06,
+      "loss": 0.702,
+      "step": 3331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0638504539662117,
+      "learning_rate": 9.144962450753491e-06,
+      "loss": 0.7534,
+      "step": 3332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.424557039128996,
+      "learning_rate": 9.144382696150785e-06,
+      "loss": 0.955,
+      "step": 3333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2797264970996105,
+      "learning_rate": 9.143802763454723e-06,
+      "loss": 0.6867,
+      "step": 3334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.079523553832919,
+      "learning_rate": 9.143222652690227e-06,
+      "loss": 0.7825,
+      "step": 3335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2952239550131137,
+      "learning_rate": 9.142642363882228e-06,
+      "loss": 0.5372,
+      "step": 3336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3486175735951145,
+      "learning_rate": 9.14206189705566e-06,
+      "loss": 0.8454,
+      "step": 3337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9598021348450858,
+      "learning_rate": 9.141481252235467e-06,
+      "loss": 0.7641,
+      "step": 3338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2485550825504843,
+      "learning_rate": 9.140900429446601e-06,
+      "loss": 0.694,
+      "step": 3339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1405986128699976,
+      "learning_rate": 9.140319428714022e-06,
+      "loss": 0.6132,
+      "step": 3340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.120362019022197,
+      "learning_rate": 9.139738250062695e-06,
+      "loss": 0.774,
+      "step": 3341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2788275159551095,
+      "learning_rate": 9.139156893517595e-06,
+      "loss": 0.9187,
+      "step": 3342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4391331760765254,
+      "learning_rate": 9.138575359103707e-06,
+      "loss": 0.7629,
+      "step": 3343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1029620158882416,
+      "learning_rate": 9.137993646846018e-06,
+      "loss": 0.8682,
+      "step": 3344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3982037318820923,
+      "learning_rate": 9.137411756769525e-06,
+      "loss": 0.8516,
+      "step": 3345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.192913104984407,
+      "learning_rate": 9.136829688899236e-06,
+      "loss": 0.7116,
+      "step": 3346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.228960137318865,
+      "learning_rate": 9.136247443260162e-06,
+      "loss": 0.7754,
+      "step": 3347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9827490821008251,
+      "learning_rate": 9.135665019877323e-06,
+      "loss": 0.8984,
+      "step": 3348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.479961509509215,
+      "learning_rate": 9.135082418775746e-06,
+      "loss": 0.995,
+      "step": 3349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.206323681317111,
+      "learning_rate": 9.134499639980469e-06,
+      "loss": 0.9066,
+      "step": 3350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1764977426199639,
+      "learning_rate": 9.133916683516536e-06,
+      "loss": 0.5965,
+      "step": 3351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6643524411896897,
+      "learning_rate": 9.133333549408997e-06,
+      "loss": 0.7517,
+      "step": 3352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.363434875394637,
+      "learning_rate": 9.132750237682907e-06,
+      "loss": 0.8448,
+      "step": 3353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0266218934732043,
+      "learning_rate": 9.132166748363335e-06,
+      "loss": 0.772,
+      "step": 3354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.931207507625117,
+      "learning_rate": 9.131583081475356e-06,
+      "loss": 0.8455,
+      "step": 3355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5934682410797842,
+      "learning_rate": 9.130999237044052e-06,
+      "loss": 0.8649,
+      "step": 3356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.309189354137096,
+      "learning_rate": 9.13041521509451e-06,
+      "loss": 0.9101,
+      "step": 3357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2712800893241296,
+      "learning_rate": 9.129831015651827e-06,
+      "loss": 0.7254,
+      "step": 3358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8031108960477673,
+      "learning_rate": 9.129246638741108e-06,
+      "loss": 0.9007,
+      "step": 3359
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4776207455429633,
+      "learning_rate": 9.128662084387462e-06,
+      "loss": 0.7824,
+      "step": 3360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.027640554568897,
+      "learning_rate": 9.128077352616013e-06,
+      "loss": 0.8865,
+      "step": 3361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3389575423639064,
+      "learning_rate": 9.127492443451887e-06,
+      "loss": 0.9317,
+      "step": 3362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.301146456223987,
+      "learning_rate": 9.126907356920219e-06,
+      "loss": 0.8745,
+      "step": 3363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.105600959455103,
+      "learning_rate": 9.126322093046149e-06,
+      "loss": 0.8051,
+      "step": 3364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9304759415331079,
+      "learning_rate": 9.12573665185483e-06,
+      "loss": 0.809,
+      "step": 3365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.560523947484186,
+      "learning_rate": 9.125151033371417e-06,
+      "loss": 0.8987,
+      "step": 3366
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9180782499098308,
+      "learning_rate": 9.12456523762108e-06,
+      "loss": 0.8761,
+      "step": 3367
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0030504229580264,
+      "learning_rate": 9.123979264628985e-06,
+      "loss": 0.8725,
+      "step": 3368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2284418557718957,
+      "learning_rate": 9.123393114420318e-06,
+      "loss": 0.5976,
+      "step": 3369
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.384945609025221,
+      "learning_rate": 9.122806787020266e-06,
+      "loss": 0.9429,
+      "step": 3370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.155689818450961,
+      "learning_rate": 9.122220282454024e-06,
+      "loss": 0.7185,
+      "step": 3371
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4345490820792537,
+      "learning_rate": 9.121633600746796e-06,
+      "loss": 0.8091,
+      "step": 3372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5068035839576017,
+      "learning_rate": 9.121046741923792e-06,
+      "loss": 0.7679,
+      "step": 3373
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7352203845182927,
+      "learning_rate": 9.120459706010233e-06,
+      "loss": 0.897,
+      "step": 3374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.762371620338535,
+      "learning_rate": 9.119872493031343e-06,
+      "loss": 0.8447,
+      "step": 3375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.00253960965454,
+      "learning_rate": 9.119285103012356e-06,
+      "loss": 0.6365,
+      "step": 3376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2091463071460717,
+      "learning_rate": 9.118697535978513e-06,
+      "loss": 0.8793,
+      "step": 3377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6287255941269203,
+      "learning_rate": 9.118109791955067e-06,
+      "loss": 0.9752,
+      "step": 3378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2885014882087384,
+      "learning_rate": 9.11752187096727e-06,
+      "loss": 0.897,
+      "step": 3379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8725658727584278,
+      "learning_rate": 9.116933773040389e-06,
+      "loss": 0.6656,
+      "step": 3380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7296708937920697,
+      "learning_rate": 9.116345498199693e-06,
+      "loss": 0.7468,
+      "step": 3381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4982868650063037,
+      "learning_rate": 9.115757046470464e-06,
+      "loss": 0.8067,
+      "step": 3382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.004014905557072,
+      "learning_rate": 9.115168417877988e-06,
+      "loss": 1.0102,
+      "step": 3383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.046052750554042,
+      "learning_rate": 9.114579612447562e-06,
+      "loss": 0.9605,
+      "step": 3384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8468631036815024,
+      "learning_rate": 9.113990630204485e-06,
+      "loss": 0.7674,
+      "step": 3385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3071529936993125,
+      "learning_rate": 9.113401471174068e-06,
+      "loss": 0.8401,
+      "step": 3386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.941275800235633,
+      "learning_rate": 9.11281213538163e-06,
+      "loss": 1.0116,
+      "step": 3387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8358366926329055,
+      "learning_rate": 9.112222622852494e-06,
+      "loss": 0.9185,
+      "step": 3388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2340155324647308,
+      "learning_rate": 9.111632933611993e-06,
+      "loss": 0.7214,
+      "step": 3389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2400127245235355,
+      "learning_rate": 9.111043067685468e-06,
+      "loss": 0.6522,
+      "step": 3390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.24315572387043,
+      "learning_rate": 9.110453025098267e-06,
+      "loss": 0.8097,
+      "step": 3391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.279558762569489,
+      "learning_rate": 9.109862805875745e-06,
+      "loss": 0.9755,
+      "step": 3392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3574571305022527,
+      "learning_rate": 9.109272410043265e-06,
+      "loss": 0.9154,
+      "step": 3393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1282900756947836,
+      "learning_rate": 9.108681837626199e-06,
+      "loss": 0.9423,
+      "step": 3394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.943101756590441,
+      "learning_rate": 9.108091088649922e-06,
+      "loss": 0.9344,
+      "step": 3395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.109486606072231,
+      "learning_rate": 9.107500163139822e-06,
+      "loss": 0.5568,
+      "step": 3396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0978908572062633,
+      "learning_rate": 9.106909061121291e-06,
+      "loss": 0.983,
+      "step": 3397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.384836665859532,
+      "learning_rate": 9.106317782619733e-06,
+      "loss": 0.9028,
+      "step": 3398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4429215522056067,
+      "learning_rate": 9.105726327660556e-06,
+      "loss": 0.7729,
+      "step": 3399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.074108870314416,
+      "learning_rate": 9.105134696269172e-06,
+      "loss": 0.8237,
+      "step": 3400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8801365586703938,
+      "learning_rate": 9.10454288847101e-06,
+      "loss": 0.8677,
+      "step": 3401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3546183963110776,
+      "learning_rate": 9.103950904291496e-06,
+      "loss": 0.8572,
+      "step": 3402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1597191586821283,
+      "learning_rate": 9.103358743756072e-06,
+      "loss": 0.9196,
+      "step": 3403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.080575342222885,
+      "learning_rate": 9.102766406890185e-06,
+      "loss": 1.1403,
+      "step": 3404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.883536716453644,
+      "learning_rate": 9.102173893719288e-06,
+      "loss": 1.0718,
+      "step": 3405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1744935876044629,
+      "learning_rate": 9.101581204268843e-06,
+      "loss": 0.7766,
+      "step": 3406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5038538657603115,
+      "learning_rate": 9.100988338564318e-06,
+      "loss": 0.6855,
+      "step": 3407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.246904995279873,
+      "learning_rate": 9.100395296631192e-06,
+      "loss": 0.9273,
+      "step": 3408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0201889491572764,
+      "learning_rate": 9.099802078494947e-06,
+      "loss": 0.687,
+      "step": 3409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.008228882230867,
+      "learning_rate": 9.099208684181078e-06,
+      "loss": 0.6757,
+      "step": 3410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.315578144930983,
+      "learning_rate": 9.098615113715078e-06,
+      "loss": 0.8728,
+      "step": 3411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4184528754675996,
+      "learning_rate": 9.098021367122462e-06,
+      "loss": 0.9538,
+      "step": 3412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7602561383356856,
+      "learning_rate": 9.097427444428742e-06,
+      "loss": 0.9287,
+      "step": 3413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.262523228337465,
+      "learning_rate": 9.096833345659437e-06,
+      "loss": 0.7937,
+      "step": 3414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1644057024930463,
+      "learning_rate": 9.096239070840079e-06,
+      "loss": 0.9067,
+      "step": 3415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2938316163153765,
+      "learning_rate": 9.095644619996206e-06,
+      "loss": 0.9944,
+      "step": 3416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3230923769351044,
+      "learning_rate": 9.095049993153362e-06,
+      "loss": 0.8745,
+      "step": 3417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.274423653436821,
+      "learning_rate": 9.0944551903371e-06,
+      "loss": 0.9249,
+      "step": 3418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.413313289987743,
+      "learning_rate": 9.09386021157298e-06,
+      "loss": 0.8048,
+      "step": 3419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.079281523924938,
+      "learning_rate": 9.09326505688657e-06,
+      "loss": 0.963,
+      "step": 3420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.171629089509474,
+      "learning_rate": 9.092669726303441e-06,
+      "loss": 0.7981,
+      "step": 3421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3289311719634367,
+      "learning_rate": 9.092074219849181e-06,
+      "loss": 0.7484,
+      "step": 3422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0685414561138296,
+      "learning_rate": 9.09147853754938e-06,
+      "loss": 0.9713,
+      "step": 3423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9473764956851531,
+      "learning_rate": 9.09088267942963e-06,
+      "loss": 0.8656,
+      "step": 3424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9451196233816064,
+      "learning_rate": 9.090286645515542e-06,
+      "loss": 0.8901,
+      "step": 3425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9403133993277484,
+      "learning_rate": 9.089690435832729e-06,
+      "loss": 0.8217,
+      "step": 3426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8980441994909465,
+      "learning_rate": 9.089094050406808e-06,
+      "loss": 0.5754,
+      "step": 3427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0962234778410487,
+      "learning_rate": 9.088497489263411e-06,
+      "loss": 0.7891,
+      "step": 3428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.002214828279123,
+      "learning_rate": 9.087900752428168e-06,
+      "loss": 0.8874,
+      "step": 3429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3655960314268563,
+      "learning_rate": 9.087303839926727e-06,
+      "loss": 0.7836,
+      "step": 3430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.265930964815593,
+      "learning_rate": 9.086706751784738e-06,
+      "loss": 0.9311,
+      "step": 3431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4528111512277277,
+      "learning_rate": 9.086109488027857e-06,
+      "loss": 0.6004,
+      "step": 3432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.623765329896358,
+      "learning_rate": 9.08551204868175e-06,
+      "loss": 0.8093,
+      "step": 3433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.619419779392002,
+      "learning_rate": 9.084914433772094e-06,
+      "loss": 0.9286,
+      "step": 3434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.332164811174805,
+      "learning_rate": 9.084316643324567e-06,
+      "loss": 0.809,
+      "step": 3435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5421608921292456,
+      "learning_rate": 9.083718677364855e-06,
+      "loss": 0.8656,
+      "step": 3436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9608258296314431,
+      "learning_rate": 9.08312053591866e-06,
+      "loss": 0.6664,
+      "step": 3437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2475903916367113,
+      "learning_rate": 9.082522219011678e-06,
+      "loss": 0.969,
+      "step": 3438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.93175992143812,
+      "learning_rate": 9.081923726669626e-06,
+      "loss": 0.9386,
+      "step": 3439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5203137082006966,
+      "learning_rate": 9.081325058918222e-06,
+      "loss": 0.9767,
+      "step": 3440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2421273451061783,
+      "learning_rate": 9.080726215783189e-06,
+      "loss": 0.7031,
+      "step": 3441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0865957947668226,
+      "learning_rate": 9.080127197290262e-06,
+      "loss": 0.8886,
+      "step": 3442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2621709871236937,
+      "learning_rate": 9.079528003465183e-06,
+      "loss": 0.8221,
+      "step": 3443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2866111297417144,
+      "learning_rate": 9.0789286343337e-06,
+      "loss": 0.7983,
+      "step": 3444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3447383914299826,
+      "learning_rate": 9.07832908992157e-06,
+      "loss": 0.809,
+      "step": 3445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0408819693801068,
+      "learning_rate": 9.077729370254555e-06,
+      "loss": 0.6802,
+      "step": 3446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2583688058854627,
+      "learning_rate": 9.077129475358426e-06,
+      "loss": 0.7604,
+      "step": 3447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1361470443003479,
+      "learning_rate": 9.076529405258966e-06,
+      "loss": 0.6312,
+      "step": 3448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0734296256780413,
+      "learning_rate": 9.075929159981957e-06,
+      "loss": 0.9123,
+      "step": 3449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2512621035545806,
+      "learning_rate": 9.075328739553195e-06,
+      "loss": 1.0349,
+      "step": 3450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.723577785955099,
+      "learning_rate": 9.07472814399848e-06,
+      "loss": 0.9706,
+      "step": 3451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.448884910650896,
+      "learning_rate": 9.074127373343623e-06,
+      "loss": 0.936,
+      "step": 3452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.347049537560534,
+      "learning_rate": 9.07352642761444e-06,
+      "loss": 1.0443,
+      "step": 3453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8110021872106241,
+      "learning_rate": 9.072925306836751e-06,
+      "loss": 0.8288,
+      "step": 3454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.062966809440395,
+      "learning_rate": 9.072324011036392e-06,
+      "loss": 0.9125,
+      "step": 3455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2258363047063872,
+      "learning_rate": 9.071722540239203e-06,
+      "loss": 0.7644,
+      "step": 3456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8261955450834904,
+      "learning_rate": 9.071120894471026e-06,
+      "loss": 1.0175,
+      "step": 3457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2848851436185793,
+      "learning_rate": 9.070519073757717e-06,
+      "loss": 0.6858,
+      "step": 3458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3280445195860624,
+      "learning_rate": 9.06991707812514e-06,
+      "loss": 0.9573,
+      "step": 3459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3633084156965416,
+      "learning_rate": 9.069314907599159e-06,
+      "loss": 0.6493,
+      "step": 3460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.811133861831706,
+      "learning_rate": 9.068712562205655e-06,
+      "loss": 0.8573,
+      "step": 3461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8662240089946096,
+      "learning_rate": 9.06811004197051e-06,
+      "loss": 0.8293,
+      "step": 3462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8087594214539784,
+      "learning_rate": 9.067507346919618e-06,
+      "loss": 0.9745,
+      "step": 3463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1836357966805477,
+      "learning_rate": 9.066904477078875e-06,
+      "loss": 0.7024,
+      "step": 3464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9765409433753125,
+      "learning_rate": 9.06630143247419e-06,
+      "loss": 0.6211,
+      "step": 3465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0566384847003538,
+      "learning_rate": 9.065698213131477e-06,
+      "loss": 0.7089,
+      "step": 3466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0213612428735837,
+      "learning_rate": 9.065094819076655e-06,
+      "loss": 0.8359,
+      "step": 3467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.300140206435464,
+      "learning_rate": 9.064491250335656e-06,
+      "loss": 0.6225,
+      "step": 3468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0080122292796556,
+      "learning_rate": 9.063887506934417e-06,
+      "loss": 0.9996,
+      "step": 3469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.068487000176446,
+      "learning_rate": 9.06328358889888e-06,
+      "loss": 0.8951,
+      "step": 3470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6829638787436965,
+      "learning_rate": 9.062679496254997e-06,
+      "loss": 0.977,
+      "step": 3471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.461703345772984,
+      "learning_rate": 9.062075229028728e-06,
+      "loss": 0.8131,
+      "step": 3472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1587064724245812,
+      "learning_rate": 9.061470787246042e-06,
+      "loss": 0.759,
+      "step": 3473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2321764423196595,
+      "learning_rate": 9.06086617093291e-06,
+      "loss": 0.8737,
+      "step": 3474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.048281382983206,
+      "learning_rate": 9.060261380115314e-06,
+      "loss": 0.7388,
+      "step": 3475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.261362209584742,
+      "learning_rate": 9.059656414819243e-06,
+      "loss": 1.0208,
+      "step": 3476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8761572177215298,
+      "learning_rate": 9.059051275070696e-06,
+      "loss": 0.7389,
+      "step": 3477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5202080520149908,
+      "learning_rate": 9.058445960895674e-06,
+      "loss": 0.7218,
+      "step": 3478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3978096428401017,
+      "learning_rate": 9.057840472320192e-06,
+      "loss": 0.7961,
+      "step": 3479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1888233779587982,
+      "learning_rate": 9.057234809370266e-06,
+      "loss": 0.6731,
+      "step": 3480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6738824584669714,
+      "learning_rate": 9.056628972071926e-06,
+      "loss": 0.8032,
+      "step": 3481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.248841754024754,
+      "learning_rate": 9.056022960451204e-06,
+      "loss": 0.7974,
+      "step": 3482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7414006300611073,
+      "learning_rate": 9.055416774534142e-06,
+      "loss": 0.8255,
+      "step": 3483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5754031214102335,
+      "learning_rate": 9.054810414346789e-06,
+      "loss": 0.8812,
+      "step": 3484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4019346200867964,
+      "learning_rate": 9.054203879915202e-06,
+      "loss": 0.9614,
+      "step": 3485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2244523917399524,
+      "learning_rate": 9.053597171265447e-06,
+      "loss": 0.8529,
+      "step": 3486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0135293336169036,
+      "learning_rate": 9.052990288423591e-06,
+      "loss": 0.6053,
+      "step": 3487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.375788848461228,
+      "learning_rate": 9.052383231415717e-06,
+      "loss": 0.9174,
+      "step": 3488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0578240806033286,
+      "learning_rate": 9.05177600026791e-06,
+      "loss": 0.9075,
+      "step": 3489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3054079177416915,
+      "learning_rate": 9.051168595006264e-06,
+      "loss": 0.596,
+      "step": 3490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9587462063421086,
+      "learning_rate": 9.050561015656882e-06,
+      "loss": 0.8324,
+      "step": 3491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.047595626630741,
+      "learning_rate": 9.049953262245872e-06,
+      "loss": 0.7422,
+      "step": 3492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8071814848824104,
+      "learning_rate": 9.04934533479935e-06,
+      "loss": 0.7135,
+      "step": 3493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.486555925389927,
+      "learning_rate": 9.048737233343442e-06,
+      "loss": 0.7525,
+      "step": 3494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0192874522057473,
+      "learning_rate": 9.048128957904277e-06,
+      "loss": 0.8619,
+      "step": 3495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2502711682311547,
+      "learning_rate": 9.047520508507994e-06,
+      "loss": 0.7217,
+      "step": 3496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.217832809923705,
+      "learning_rate": 9.046911885180743e-06,
+      "loss": 0.8588,
+      "step": 3497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9822764608430015,
+      "learning_rate": 9.046303087948674e-06,
+      "loss": 0.8654,
+      "step": 3498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0250375100395464,
+      "learning_rate": 9.045694116837948e-06,
+      "loss": 0.9101,
+      "step": 3499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.711047369526188,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.9715,
+      "step": 3500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.004403735885255,
+      "learning_rate": 9.044475653085218e-06,
+      "loss": 0.8131,
+      "step": 3501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.954400734614963,
+      "learning_rate": 9.04386616049557e-06,
+      "loss": 0.9343,
+      "step": 3502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3918995542095611,
+      "learning_rate": 9.043256494131987e-06,
+      "loss": 0.7612,
+      "step": 3503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.495266219912447,
+      "learning_rate": 9.042646654020667e-06,
+      "loss": 1.089,
+      "step": 3504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5580562303678491,
+      "learning_rate": 9.04203664018782e-06,
+      "loss": 0.752,
+      "step": 3505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1251553550101403,
+      "learning_rate": 9.041426452659654e-06,
+      "loss": 0.6077,
+      "step": 3506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.132155038368516,
+      "learning_rate": 9.040816091462393e-06,
+      "loss": 0.7713,
+      "step": 3507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4563304939925956,
+      "learning_rate": 9.040205556622266e-06,
+      "loss": 0.9538,
+      "step": 3508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1937864476047375,
+      "learning_rate": 9.039594848165507e-06,
+      "loss": 0.8283,
+      "step": 3509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2717125362958854,
+      "learning_rate": 9.038983966118359e-06,
+      "loss": 0.8238,
+      "step": 3510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1218753525079144,
+      "learning_rate": 9.038372910507079e-06,
+      "loss": 0.7058,
+      "step": 3511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.057151499062998,
+      "learning_rate": 9.03776168135792e-06,
+      "loss": 0.602,
+      "step": 3512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9323750572278935,
+      "learning_rate": 9.037150278697147e-06,
+      "loss": 0.5053,
+      "step": 3513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.223035853348879,
+      "learning_rate": 9.036538702551037e-06,
+      "loss": 1.0361,
+      "step": 3514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1965082157698448,
+      "learning_rate": 9.035926952945868e-06,
+      "loss": 0.6728,
+      "step": 3515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9615985029343075,
+      "learning_rate": 9.035315029907928e-06,
+      "loss": 0.914,
+      "step": 3516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4084682106543327,
+      "learning_rate": 9.034702933463516e-06,
+      "loss": 0.932,
+      "step": 3517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.122790628109082,
+      "learning_rate": 9.034090663638933e-06,
+      "loss": 0.8526,
+      "step": 3518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.239611194072638,
+      "learning_rate": 9.03347822046049e-06,
+      "loss": 0.8027,
+      "step": 3519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4936533508291205,
+      "learning_rate": 9.032865603954504e-06,
+      "loss": 0.9455,
+      "step": 3520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2576656545916602,
+      "learning_rate": 9.032252814147302e-06,
+      "loss": 1.0566,
+      "step": 3521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.865072095949122,
+      "learning_rate": 9.031639851065217e-06,
+      "loss": 0.8641,
+      "step": 3522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7980609734432176,
+      "learning_rate": 9.031026714734588e-06,
+      "loss": 0.6387,
+      "step": 3523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.293439500168352,
+      "learning_rate": 9.03041340518176e-06,
+      "loss": 0.7774,
+      "step": 3524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0784832246456997,
+      "learning_rate": 9.029799922433098e-06,
+      "loss": 0.8319,
+      "step": 3525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0471646835069737,
+      "learning_rate": 9.029186266514956e-06,
+      "loss": 0.8989,
+      "step": 3526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2893438970315314,
+      "learning_rate": 9.028572437453705e-06,
+      "loss": 0.6669,
+      "step": 3527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.35445796745486,
+      "learning_rate": 9.027958435275726e-06,
+      "loss": 0.8767,
+      "step": 3528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.304809630312567,
+      "learning_rate": 9.027344260007401e-06,
+      "loss": 0.8498,
+      "step": 3529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8813519094377869,
+      "learning_rate": 9.026729911675124e-06,
+      "loss": 0.8883,
+      "step": 3530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2277797801506884,
+      "learning_rate": 9.026115390305295e-06,
+      "loss": 0.7702,
+      "step": 3531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8088912958716024,
+      "learning_rate": 9.025500695924321e-06,
+      "loss": 0.6602,
+      "step": 3532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.859676667706023,
+      "learning_rate": 9.024885828558618e-06,
+      "loss": 0.7519,
+      "step": 3533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.66393963273552,
+      "learning_rate": 9.024270788234606e-06,
+      "loss": 0.7667,
+      "step": 3534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1555366807719283,
+      "learning_rate": 9.023655574978716e-06,
+      "loss": 0.8496,
+      "step": 3535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.996376145740077,
+      "learning_rate": 9.023040188817385e-06,
+      "loss": 0.7508,
+      "step": 3536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2529479920111923,
+      "learning_rate": 9.022424629777056e-06,
+      "loss": 0.6872,
+      "step": 3537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0618238386701562,
+      "learning_rate": 9.021808897884183e-06,
+      "loss": 0.6447,
+      "step": 3538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.620990491346329,
+      "learning_rate": 9.021192993165224e-06,
+      "loss": 1.0455,
+      "step": 3539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1481480878687162,
+      "learning_rate": 9.020576915646649e-06,
+      "loss": 0.94,
+      "step": 3540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0337342836451353,
+      "learning_rate": 9.019960665354926e-06,
+      "loss": 0.6342,
+      "step": 3541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4578961143625713,
+      "learning_rate": 9.019344242316542e-06,
+      "loss": 0.8781,
+      "step": 3542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1034137778017707,
+      "learning_rate": 9.018727646557985e-06,
+      "loss": 1.0745,
+      "step": 3543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5965919478651096,
+      "learning_rate": 9.01811087810575e-06,
+      "loss": 0.8484,
+      "step": 3544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0279872515162274,
+      "learning_rate": 9.01749393698634e-06,
+      "loss": 0.9488,
+      "step": 3545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0940383209365914,
+      "learning_rate": 9.01687682322627e-06,
+      "loss": 0.6826,
+      "step": 3546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.033472388444839,
+      "learning_rate": 9.016259536852059e-06,
+      "loss": 0.8048,
+      "step": 3547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.702469268338486,
+      "learning_rate": 9.015642077890228e-06,
+      "loss": 0.6018,
+      "step": 3548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.796742379852457,
+      "learning_rate": 9.015024446367315e-06,
+      "loss": 1.1501,
+      "step": 3549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9319178212932266,
+      "learning_rate": 9.014406642309859e-06,
+      "loss": 1.0048,
+      "step": 3550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5988187605928412,
+      "learning_rate": 9.013788665744409e-06,
+      "loss": 0.7809,
+      "step": 3551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2681694993274797,
+      "learning_rate": 9.01317051669752e-06,
+      "loss": 0.831,
+      "step": 3552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3871823161717725,
+      "learning_rate": 9.012552195195759e-06,
+      "loss": 0.688,
+      "step": 3553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8607331547603747,
+      "learning_rate": 9.01193370126569e-06,
+      "loss": 0.6905,
+      "step": 3554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1814022902719823,
+      "learning_rate": 9.011315034933896e-06,
+      "loss": 0.8255,
+      "step": 3555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8804978891382051,
+      "learning_rate": 9.010696196226963e-06,
+      "loss": 0.9114,
+      "step": 3556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0336671795302577,
+      "learning_rate": 9.010077185171481e-06,
+      "loss": 0.6224,
+      "step": 3557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.574091001131123,
+      "learning_rate": 9.009458001794052e-06,
+      "loss": 0.7126,
+      "step": 3558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2047552268945076,
+      "learning_rate": 9.008838646121282e-06,
+      "loss": 0.9908,
+      "step": 3559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7544565791438758,
+      "learning_rate": 9.008219118179787e-06,
+      "loss": 0.967,
+      "step": 3560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.690261515989758,
+      "learning_rate": 9.007599417996191e-06,
+      "loss": 0.6994,
+      "step": 3561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4010296558828608,
+      "learning_rate": 9.006979545597123e-06,
+      "loss": 0.9125,
+      "step": 3562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2738957828627164,
+      "learning_rate": 9.00635950100922e-06,
+      "loss": 0.828,
+      "step": 3563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5676065332735862,
+      "learning_rate": 9.005739284259123e-06,
+      "loss": 0.9211,
+      "step": 3564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0214639175104523,
+      "learning_rate": 9.00511889537349e-06,
+      "loss": 0.6759,
+      "step": 3565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9668179533513912,
+      "learning_rate": 9.00449833437898e-06,
+      "loss": 0.7935,
+      "step": 3566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.592815333035239,
+      "learning_rate": 9.003877601302256e-06,
+      "loss": 0.6705,
+      "step": 3567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3359040106096876,
+      "learning_rate": 9.003256696169995e-06,
+      "loss": 0.8631,
+      "step": 3568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0384424212561942,
+      "learning_rate": 9.002635619008877e-06,
+      "loss": 0.8593,
+      "step": 3569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0088514724505635,
+      "learning_rate": 9.002014369845592e-06,
+      "loss": 0.8216,
+      "step": 3570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1496509340748313,
+      "learning_rate": 9.00139294870684e-06,
+      "loss": 0.7263,
+      "step": 3571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.010285932569476,
+      "learning_rate": 9.000771355619318e-06,
+      "loss": 0.8338,
+      "step": 3572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1470433371131117,
+      "learning_rate": 9.000149590609741e-06,
+      "loss": 0.8509,
+      "step": 3573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.145429288538416,
+      "learning_rate": 8.999527653704829e-06,
+      "loss": 0.7996,
+      "step": 3574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4870508865225407,
+      "learning_rate": 8.998905544931304e-06,
+      "loss": 1.0187,
+      "step": 3575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0512731281082175,
+      "learning_rate": 8.998283264315903e-06,
+      "loss": 0.836,
+      "step": 3576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3272641979119206,
+      "learning_rate": 8.997660811885367e-06,
+      "loss": 0.8881,
+      "step": 3577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.375815585952761,
+      "learning_rate": 8.99703818766644e-06,
+      "loss": 0.6186,
+      "step": 3578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8089649018559535,
+      "learning_rate": 8.996415391685882e-06,
+      "loss": 0.9344,
+      "step": 3579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2047557190622922,
+      "learning_rate": 8.995792423970454e-06,
+      "loss": 0.7078,
+      "step": 3580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3325414853997413,
+      "learning_rate": 8.995169284546926e-06,
+      "loss": 0.9836,
+      "step": 3581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.464027967929417,
+      "learning_rate": 8.994545973442076e-06,
+      "loss": 0.8514,
+      "step": 3582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8569950137963978,
+      "learning_rate": 8.993922490682687e-06,
+      "loss": 0.8614,
+      "step": 3583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3790674711766258,
+      "learning_rate": 8.993298836295556e-06,
+      "loss": 0.8667,
+      "step": 3584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.795893423408693,
+      "learning_rate": 8.99267501030748e-06,
+      "loss": 0.9258,
+      "step": 3585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1733490121970291,
+      "learning_rate": 8.992051012745268e-06,
+      "loss": 0.7738,
+      "step": 3586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1941096153077821,
+      "learning_rate": 8.991426843635731e-06,
+      "loss": 0.7453,
+      "step": 3587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2235009966672754,
+      "learning_rate": 8.990802503005695e-06,
+      "loss": 0.7059,
+      "step": 3588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1233515498679754,
+      "learning_rate": 8.990177990881986e-06,
+      "loss": 0.6472,
+      "step": 3589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0668779119694185,
+      "learning_rate": 8.989553307291443e-06,
+      "loss": 0.665,
+      "step": 3590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5862694653567093,
+      "learning_rate": 8.988928452260909e-06,
+      "loss": 0.844,
+      "step": 3591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0016597207640063,
+      "learning_rate": 8.988303425817235e-06,
+      "loss": 1.0127,
+      "step": 3592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.464309472239494,
+      "learning_rate": 8.987678227987282e-06,
+      "loss": 0.75,
+      "step": 3593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1726810788609479,
+      "learning_rate": 8.987052858797914e-06,
+      "loss": 0.6918,
+      "step": 3594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2357971437985205,
+      "learning_rate": 8.986427318276003e-06,
+      "loss": 0.8152,
+      "step": 3595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.246672063159511,
+      "learning_rate": 8.985801606448435e-06,
+      "loss": 1.0144,
+      "step": 3596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3126757421747826,
+      "learning_rate": 8.985175723342093e-06,
+      "loss": 1.0083,
+      "step": 3597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0804953576007366,
+      "learning_rate": 8.984549668983875e-06,
+      "loss": 0.7665,
+      "step": 3598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0334332010639216,
+      "learning_rate": 8.983923443400682e-06,
+      "loss": 0.9838,
+      "step": 3599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.545171387768477,
+      "learning_rate": 8.983297046619427e-06,
+      "loss": 0.8631,
+      "step": 3600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3800008179425487,
+      "learning_rate": 8.982670478667029e-06,
+      "loss": 0.8321,
+      "step": 3601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1122877211779976,
+      "learning_rate": 8.98204373957041e-06,
+      "loss": 0.7642,
+      "step": 3602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.110524702212672,
+      "learning_rate": 8.981416829356502e-06,
+      "loss": 0.8853,
+      "step": 3603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1121101549081933,
+      "learning_rate": 8.980789748052245e-06,
+      "loss": 1.0177,
+      "step": 3604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3209544431628317,
+      "learning_rate": 8.980162495684587e-06,
+      "loss": 0.7945,
+      "step": 3605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9786263305911729,
+      "learning_rate": 8.979535072280483e-06,
+      "loss": 0.8831,
+      "step": 3606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7135242644558857,
+      "learning_rate": 8.978907477866894e-06,
+      "loss": 0.9762,
+      "step": 3607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.264775434243341,
+      "learning_rate": 8.978279712470789e-06,
+      "loss": 0.8761,
+      "step": 3608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.271792672759413,
+      "learning_rate": 8.977651776119145e-06,
+      "loss": 1.0153,
+      "step": 3609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1366750702566457,
+      "learning_rate": 8.977023668838945e-06,
+      "loss": 0.6876,
+      "step": 3610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2170564688060734,
+      "learning_rate": 8.976395390657181e-06,
+      "loss": 0.747,
+      "step": 3611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9431453127061311,
+      "learning_rate": 8.975766941600852e-06,
+      "loss": 0.8004,
+      "step": 3612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0657548077239802,
+      "learning_rate": 8.975138321696961e-06,
+      "loss": 0.913,
+      "step": 3613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3442944368662575,
+      "learning_rate": 8.974509530972523e-06,
+      "loss": 0.6877,
+      "step": 3614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2813208719308364,
+      "learning_rate": 8.973880569454561e-06,
+      "loss": 0.8659,
+      "step": 3615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.348213931949922,
+      "learning_rate": 8.9732514371701e-06,
+      "loss": 0.9928,
+      "step": 3616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3682601451054563,
+      "learning_rate": 8.972622134146174e-06,
+      "loss": 0.834,
+      "step": 3617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.150566293246871,
+      "learning_rate": 8.97199266040983e-06,
+      "loss": 0.9609,
+      "step": 3618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.34619824036059,
+      "learning_rate": 8.971363015988115e-06,
+      "loss": 0.7985,
+      "step": 3619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2949599411476695,
+      "learning_rate": 8.970733200908084e-06,
+      "loss": 0.8802,
+      "step": 3620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.418349801407155,
+      "learning_rate": 8.970103215196807e-06,
+      "loss": 0.6796,
+      "step": 3621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6515034071397716,
+      "learning_rate": 8.96947305888135e-06,
+      "loss": 0.807,
+      "step": 3622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1801980858145686,
+      "learning_rate": 8.968842731988798e-06,
+      "loss": 0.7042,
+      "step": 3623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.314005654735003,
+      "learning_rate": 8.968212234546235e-06,
+      "loss": 0.6941,
+      "step": 3624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.508791469673464,
+      "learning_rate": 8.967581566580752e-06,
+      "loss": 0.954,
+      "step": 3625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.149564080083841,
+      "learning_rate": 8.966950728119453e-06,
+      "loss": 0.6926,
+      "step": 3626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3281175597853903,
+      "learning_rate": 8.966319719189447e-06,
+      "loss": 0.6871,
+      "step": 3627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.305942821618736,
+      "learning_rate": 8.96568853981785e-06,
+      "loss": 0.9113,
+      "step": 3628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1999759292425616,
+      "learning_rate": 8.965057190031785e-06,
+      "loss": 0.7082,
+      "step": 3629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7090867355723356,
+      "learning_rate": 8.96442566985838e-06,
+      "loss": 0.6832,
+      "step": 3630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6279346159291896,
+      "learning_rate": 8.963793979324775e-06,
+      "loss": 0.9618,
+      "step": 3631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.525228720416207,
+      "learning_rate": 8.963162118458115e-06,
+      "loss": 0.9332,
+      "step": 3632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0669962564550233,
+      "learning_rate": 8.962530087285552e-06,
+      "loss": 0.6656,
+      "step": 3633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0091787198648445,
+      "learning_rate": 8.961897885834247e-06,
+      "loss": 0.7609,
+      "step": 3634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9213463656484913,
+      "learning_rate": 8.961265514131368e-06,
+      "loss": 0.7871,
+      "step": 3635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8124547137751457,
+      "learning_rate": 8.960632972204084e-06,
+      "loss": 0.8332,
+      "step": 3636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.150178978077755,
+      "learning_rate": 8.960000260079582e-06,
+      "loss": 0.8594,
+      "step": 3637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.004852026102709,
+      "learning_rate": 8.95936737778505e-06,
+      "loss": 0.7915,
+      "step": 3638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1377336110778475,
+      "learning_rate": 8.958734325347684e-06,
+      "loss": 0.6093,
+      "step": 3639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5772160088013734,
+      "learning_rate": 8.958101102794686e-06,
+      "loss": 0.9395,
+      "step": 3640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.661371654212696,
+      "learning_rate": 8.95746771015327e-06,
+      "loss": 0.8978,
+      "step": 3641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9968943293116899,
+      "learning_rate": 8.95683414745065e-06,
+      "loss": 0.806,
+      "step": 3642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3092552984366144,
+      "learning_rate": 8.956200414714057e-06,
+      "loss": 0.6197,
+      "step": 3643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0613691019842104,
+      "learning_rate": 8.955566511970721e-06,
+      "loss": 0.6136,
+      "step": 3644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.584656555232214,
+      "learning_rate": 8.954932439247882e-06,
+      "loss": 0.8595,
+      "step": 3645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1496167184267752,
+      "learning_rate": 8.954298196572789e-06,
+      "loss": 0.6888,
+      "step": 3646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5938056904280153,
+      "learning_rate": 8.953663783972692e-06,
+      "loss": 0.9349,
+      "step": 3647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3784387660983772,
+      "learning_rate": 8.95302920147486e-06,
+      "loss": 0.9792,
+      "step": 3648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8815946913583719,
+      "learning_rate": 8.95239444910656e-06,
+      "loss": 0.8112,
+      "step": 3649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.133101382924274,
+      "learning_rate": 8.951759526895067e-06,
+      "loss": 0.9933,
+      "step": 3650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4218155857674497,
+      "learning_rate": 8.951124434867666e-06,
+      "loss": 0.978,
+      "step": 3651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9852078801578774,
+      "learning_rate": 8.950489173051648e-06,
+      "loss": 0.8691,
+      "step": 3652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2418229883907292,
+      "learning_rate": 8.949853741474312e-06,
+      "loss": 0.8603,
+      "step": 3653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.019894465358218,
+      "learning_rate": 8.949218140162965e-06,
+      "loss": 0.7628,
+      "step": 3654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.289636040029631,
+      "learning_rate": 8.948582369144919e-06,
+      "loss": 0.7429,
+      "step": 3655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0260130470260456,
+      "learning_rate": 8.947946428447495e-06,
+      "loss": 0.8764,
+      "step": 3656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3085414929203574,
+      "learning_rate": 8.94731031809802e-06,
+      "loss": 0.6465,
+      "step": 3657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.903709253231394,
+      "learning_rate": 8.946674038123829e-06,
+      "loss": 0.81,
+      "step": 3658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.058766704479871,
+      "learning_rate": 8.946037588552266e-06,
+      "loss": 0.6755,
+      "step": 3659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3922733033833707,
+      "learning_rate": 8.945400969410678e-06,
+      "loss": 0.9592,
+      "step": 3660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.796983906299395,
+      "learning_rate": 8.944764180726423e-06,
+      "loss": 0.7843,
+      "step": 3661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6980792594782996,
+      "learning_rate": 8.944127222526867e-06,
+      "loss": 0.9403,
+      "step": 3662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9785305533589694,
+      "learning_rate": 8.94349009483938e-06,
+      "loss": 0.974,
+      "step": 3663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8698634966524192,
+      "learning_rate": 8.94285279769134e-06,
+      "loss": 0.8568,
+      "step": 3664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8361663514105666,
+      "learning_rate": 8.942215331110136e-06,
+      "loss": 0.81,
+      "step": 3665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.496337704865172,
+      "learning_rate": 8.941577695123158e-06,
+      "loss": 0.8824,
+      "step": 3666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1540308850095826,
+      "learning_rate": 8.940939889757809e-06,
+      "loss": 0.8936,
+      "step": 3667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9566736661534196,
+      "learning_rate": 8.940301915041496e-06,
+      "loss": 1.0548,
+      "step": 3668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9253560489617414,
+      "learning_rate": 8.939663771001632e-06,
+      "loss": 0.9621,
+      "step": 3669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5978618903532507,
+      "learning_rate": 8.939025457665644e-06,
+      "loss": 0.7838,
+      "step": 3670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.128702377873588,
+      "learning_rate": 8.93838697506096e-06,
+      "loss": 0.9227,
+      "step": 3671
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4320812195936696,
+      "learning_rate": 8.937748323215013e-06,
+      "loss": 0.8821,
+      "step": 3672
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9817238415474325,
+      "learning_rate": 8.937109502155254e-06,
+      "loss": 0.8062,
+      "step": 3673
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9157464363355663,
+      "learning_rate": 8.93647051190913e-06,
+      "loss": 0.9282,
+      "step": 3674
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2820337698536894,
+      "learning_rate": 8.935831352504103e-06,
+      "loss": 0.7663,
+      "step": 3675
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.33528422346256,
+      "learning_rate": 8.935192023967634e-06,
+      "loss": 0.8827,
+      "step": 3676
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.003647055512618,
+      "learning_rate": 8.934552526327201e-06,
+      "loss": 0.9021,
+      "step": 3677
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.127499045583687,
+      "learning_rate": 8.933912859610284e-06,
+      "loss": 0.7252,
+      "step": 3678
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.856071431917304,
+      "learning_rate": 8.93327302384437e-06,
+      "loss": 0.8163,
+      "step": 3679
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.432079687605383,
+      "learning_rate": 8.932633019056954e-06,
+      "loss": 0.8853,
+      "step": 3680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.224813981935746,
+      "learning_rate": 8.931992845275539e-06,
+      "loss": 0.9969,
+      "step": 3681
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5008911092744204,
+      "learning_rate": 8.931352502527633e-06,
+      "loss": 0.8299,
+      "step": 3682
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.063370954165471,
+      "learning_rate": 8.930711990840755e-06,
+      "loss": 0.7643,
+      "step": 3683
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.732748727359249,
+      "learning_rate": 8.930071310242429e-06,
+      "loss": 0.9998,
+      "step": 3684
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8790511460432169,
+      "learning_rate": 8.929430460760187e-06,
+      "loss": 1.154,
+      "step": 3685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.863160601636231,
+      "learning_rate": 8.928789442421564e-06,
+      "loss": 0.8291,
+      "step": 3686
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8724856677895334,
+      "learning_rate": 8.928148255254112e-06,
+      "loss": 0.7287,
+      "step": 3687
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.930241402668977,
+      "learning_rate": 8.927506899285378e-06,
+      "loss": 0.7317,
+      "step": 3688
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3333072764362996,
+      "learning_rate": 8.926865374542928e-06,
+      "loss": 0.9404,
+      "step": 3689
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.62994429158127,
+      "learning_rate": 8.926223681054326e-06,
+      "loss": 0.7799,
+      "step": 3690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.069503031180634,
+      "learning_rate": 8.92558181884715e-06,
+      "loss": 0.8255,
+      "step": 3691
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.201612211905422,
+      "learning_rate": 8.924939787948977e-06,
+      "loss": 0.8379,
+      "step": 3692
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8153036491910854,
+      "learning_rate": 8.924297588387404e-06,
+      "loss": 0.8478,
+      "step": 3693
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1308366210732785,
+      "learning_rate": 8.92365522019002e-06,
+      "loss": 0.6914,
+      "step": 3694
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.238723178042403,
+      "learning_rate": 8.923012683384436e-06,
+      "loss": 0.8367,
+      "step": 3695
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9643594303968808,
+      "learning_rate": 8.922369977998257e-06,
+      "loss": 0.8296,
+      "step": 3696
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.577810287175034,
+      "learning_rate": 8.921727104059107e-06,
+      "loss": 0.7445,
+      "step": 3697
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5046851368121166,
+      "learning_rate": 8.921084061594608e-06,
+      "loss": 0.8227,
+      "step": 3698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9468751674735802,
+      "learning_rate": 8.920440850632395e-06,
+      "loss": 1.107,
+      "step": 3699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.074467639466453,
+      "learning_rate": 8.919797471200106e-06,
+      "loss": 0.7335,
+      "step": 3700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8826266261537317,
+      "learning_rate": 8.91915392332539e-06,
+      "loss": 0.9675,
+      "step": 3701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.746470152893695,
+      "learning_rate": 8.918510207035901e-06,
+      "loss": 0.9161,
+      "step": 3702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.708925832050352,
+      "learning_rate": 8.917866322359303e-06,
+      "loss": 0.9453,
+      "step": 3703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.048545858298353,
+      "learning_rate": 8.917222269323263e-06,
+      "loss": 0.5402,
+      "step": 3704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0210010144194652,
+      "learning_rate": 8.916578047955458e-06,
+      "loss": 0.7754,
+      "step": 3705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0985027697111307,
+      "learning_rate": 8.915933658283572e-06,
+      "loss": 0.906,
+      "step": 3706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.078192825290193,
+      "learning_rate": 8.915289100335295e-06,
+      "loss": 1.1657,
+      "step": 3707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.235753915977002,
+      "learning_rate": 8.914644374138325e-06,
+      "loss": 0.8032,
+      "step": 3708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.141806179143029,
+      "learning_rate": 8.91399947972037e-06,
+      "loss": 0.6724,
+      "step": 3709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.207844460186117,
+      "learning_rate": 8.913354417109136e-06,
+      "loss": 0.7908,
+      "step": 3710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2441899733630972,
+      "learning_rate": 8.91270918633235e-06,
+      "loss": 0.9951,
+      "step": 3711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5556175096926825,
+      "learning_rate": 8.912063787417738e-06,
+      "loss": 0.9429,
+      "step": 3712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.3547955160718015,
+      "learning_rate": 8.911418220393031e-06,
+      "loss": 0.7543,
+      "step": 3713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.052126753421225,
+      "learning_rate": 8.91077248528597e-06,
+      "loss": 0.8419,
+      "step": 3714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6492554962314228,
+      "learning_rate": 8.910126582124307e-06,
+      "loss": 0.9398,
+      "step": 3715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6928522324074335,
+      "learning_rate": 8.909480510935797e-06,
+      "loss": 0.9468,
+      "step": 3716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.028195582513883,
+      "learning_rate": 8.908834271748202e-06,
+      "loss": 0.6144,
+      "step": 3717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.26763869589723,
+      "learning_rate": 8.908187864589292e-06,
+      "loss": 0.8455,
+      "step": 3718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2870453229279324,
+      "learning_rate": 8.907541289486847e-06,
+      "loss": 1.0025,
+      "step": 3719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3127200172847218,
+      "learning_rate": 8.90689454646865e-06,
+      "loss": 0.6864,
+      "step": 3720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9512224213856386,
+      "learning_rate": 8.906247635562493e-06,
+      "loss": 0.8136,
+      "step": 3721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.510854923663712,
+      "learning_rate": 8.905600556796176e-06,
+      "loss": 0.8396,
+      "step": 3722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.204694531563819,
+      "learning_rate": 8.904953310197504e-06,
+      "loss": 1.0156,
+      "step": 3723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.226701393050224,
+      "learning_rate": 8.904305895794292e-06,
+      "loss": 0.913,
+      "step": 3724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8719697721921027,
+      "learning_rate": 8.903658313614358e-06,
+      "loss": 0.916,
+      "step": 3725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0251077710536873,
+      "learning_rate": 8.903010563685534e-06,
+      "loss": 0.9581,
+      "step": 3726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7364203925519908,
+      "learning_rate": 8.902362646035655e-06,
+      "loss": 0.919,
+      "step": 3727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.192370113412689,
+      "learning_rate": 8.90171456069256e-06,
+      "loss": 0.8452,
+      "step": 3728
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.040509468662177,
+      "learning_rate": 8.901066307684102e-06,
+      "loss": 0.954,
+      "step": 3729
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5858506772248515,
+      "learning_rate": 8.900417887038136e-06,
+      "loss": 0.6368,
+      "step": 3730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1411157827306386,
+      "learning_rate": 8.899769298782528e-06,
+      "loss": 0.5945,
+      "step": 3731
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9745657298883843,
+      "learning_rate": 8.899120542945147e-06,
+      "loss": 0.8345,
+      "step": 3732
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8612213616996252,
+      "learning_rate": 8.898471619553872e-06,
+      "loss": 0.9494,
+      "step": 3733
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1099225952810015,
+      "learning_rate": 8.89782252863659e-06,
+      "loss": 0.7208,
+      "step": 3734
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.625691100590587,
+      "learning_rate": 8.89717327022119e-06,
+      "loss": 0.8941,
+      "step": 3735
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.152052335102718,
+      "learning_rate": 8.89652384433558e-06,
+      "loss": 0.8414,
+      "step": 3736
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7592102645790493,
+      "learning_rate": 8.89587425100766e-06,
+      "loss": 0.8124,
+      "step": 3737
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.41278604660796,
+      "learning_rate": 8.895224490265346e-06,
+      "loss": 1.0251,
+      "step": 3738
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7330342347311127,
+      "learning_rate": 8.894574562136561e-06,
+      "loss": 1.0672,
+      "step": 3739
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1401669180441534,
+      "learning_rate": 8.893924466649235e-06,
+      "loss": 0.9305,
+      "step": 3740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.323258300951865,
+      "learning_rate": 8.893274203831302e-06,
+      "loss": 0.9037,
+      "step": 3741
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.46629013591146,
+      "learning_rate": 8.892623773710704e-06,
+      "loss": 0.9154,
+      "step": 3742
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.315635778259215,
+      "learning_rate": 8.891973176315393e-06,
+      "loss": 0.8644,
+      "step": 3743
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0405635845140098,
+      "learning_rate": 8.89132241167333e-06,
+      "loss": 0.9093,
+      "step": 3744
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.25586156163197,
+      "learning_rate": 8.890671479812472e-06,
+      "loss": 0.8977,
+      "step": 3745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.539501751125677,
+      "learning_rate": 8.890020380760798e-06,
+      "loss": 0.9641,
+      "step": 3746
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9774563857013776,
+      "learning_rate": 8.889369114546285e-06,
+      "loss": 0.7713,
+      "step": 3747
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0825975798283327,
+      "learning_rate": 8.888717681196918e-06,
+      "loss": 0.8516,
+      "step": 3748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6243363652384355,
+      "learning_rate": 8.888066080740692e-06,
+      "loss": 0.9047,
+      "step": 3749
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2328386907572013,
+      "learning_rate": 8.887414313205608e-06,
+      "loss": 1.0669,
+      "step": 3750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0274291109083835,
+      "learning_rate": 8.886762378619674e-06,
+      "loss": 0.7762,
+      "step": 3751
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.120450833061051,
+      "learning_rate": 8.886110277010902e-06,
+      "loss": 0.8802,
+      "step": 3752
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.205814951116257,
+      "learning_rate": 8.885458008407317e-06,
+      "loss": 0.9211,
+      "step": 3753
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.367747548432093,
+      "learning_rate": 8.88480557283695e-06,
+      "loss": 0.7294,
+      "step": 3754
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2830846163091154,
+      "learning_rate": 8.884152970327833e-06,
+      "loss": 0.7355,
+      "step": 3755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.717655988835119,
+      "learning_rate": 8.883500200908013e-06,
+      "loss": 0.9207,
+      "step": 3756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0683424108891475,
+      "learning_rate": 8.882847264605542e-06,
+      "loss": 0.9853,
+      "step": 3757
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9128502086907915,
+      "learning_rate": 8.882194161448476e-06,
+      "loss": 0.8991,
+      "step": 3758
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9521198677791711,
+      "learning_rate": 8.88154089146488e-06,
+      "loss": 0.8022,
+      "step": 3759
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.286922321833568,
+      "learning_rate": 8.880887454682827e-06,
+      "loss": 0.9194,
+      "step": 3760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.195159339352561,
+      "learning_rate": 8.880233851130397e-06,
+      "loss": 0.9495,
+      "step": 3761
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.151145550681024,
+      "learning_rate": 8.879580080835677e-06,
+      "loss": 0.8693,
+      "step": 3762
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0262608308399463,
+      "learning_rate": 8.87892614382676e-06,
+      "loss": 0.8097,
+      "step": 3763
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.149797985018559,
+      "learning_rate": 8.878272040131748e-06,
+      "loss": 0.8697,
+      "step": 3764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4338469267598075,
+      "learning_rate": 8.877617769778751e-06,
+      "loss": 0.9131,
+      "step": 3765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3252886162109825,
+      "learning_rate": 8.876963332795881e-06,
+      "loss": 0.753,
+      "step": 3766
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.031874699585823,
+      "learning_rate": 8.876308729211261e-06,
+      "loss": 0.7839,
+      "step": 3767
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2486607933527853,
+      "learning_rate": 8.875653959053023e-06,
+      "loss": 0.5873,
+      "step": 3768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.940681901049866,
+      "learning_rate": 8.874999022349303e-06,
+      "loss": 0.6111,
+      "step": 3769
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0797968893449537,
+      "learning_rate": 8.874343919128245e-06,
+      "loss": 0.9194,
+      "step": 3770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9685850157236027,
+      "learning_rate": 8.873688649417999e-06,
+      "loss": 0.9415,
+      "step": 3771
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.958665463666576,
+      "learning_rate": 8.873033213246726e-06,
+      "loss": 0.996,
+      "step": 3772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2646094396117884,
+      "learning_rate": 8.87237761064259e-06,
+      "loss": 0.8445,
+      "step": 3773
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.931752943770893,
+      "learning_rate": 8.871721841633762e-06,
+      "loss": 0.659,
+      "step": 3774
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1771565084132547,
+      "learning_rate": 8.871065906248426e-06,
+      "loss": 0.9924,
+      "step": 3775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2065811720999906,
+      "learning_rate": 8.870409804514765e-06,
+      "loss": 0.642,
+      "step": 3776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.236915792091447,
+      "learning_rate": 8.869753536460976e-06,
+      "loss": 0.8242,
+      "step": 3777
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.345432446111317,
+      "learning_rate": 8.86909710211526e-06,
+      "loss": 0.8541,
+      "step": 3778
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.084266383569708,
+      "learning_rate": 8.868440501505822e-06,
+      "loss": 0.9417,
+      "step": 3779
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9338238419018683,
+      "learning_rate": 8.867783734660883e-06,
+      "loss": 0.8174,
+      "step": 3780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.672351763197171,
+      "learning_rate": 8.86712680160866e-06,
+      "loss": 0.8388,
+      "step": 3781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.079974961042275,
+      "learning_rate": 8.866469702377386e-06,
+      "loss": 0.7341,
+      "step": 3782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.732907037196,
+      "learning_rate": 8.8658124369953e-06,
+      "loss": 0.751,
+      "step": 3783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9927867089628573,
+      "learning_rate": 8.865155005490643e-06,
+      "loss": 0.805,
+      "step": 3784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.739269748717133,
+      "learning_rate": 8.864497407891668e-06,
+      "loss": 0.7467,
+      "step": 3785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1208662228023898,
+      "learning_rate": 8.86383964422663e-06,
+      "loss": 0.6865,
+      "step": 3786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1567517586326146,
+      "learning_rate": 8.8631817145238e-06,
+      "loss": 0.7167,
+      "step": 3787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.119413823874302,
+      "learning_rate": 8.862523618811448e-06,
+      "loss": 0.6918,
+      "step": 3788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0702845849601434,
+      "learning_rate": 8.861865357117852e-06,
+      "loss": 0.8068,
+      "step": 3789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.990105576921987,
+      "learning_rate": 8.861206929471302e-06,
+      "loss": 1.1408,
+      "step": 3790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0602086829546515,
+      "learning_rate": 8.86054833590009e-06,
+      "loss": 0.6917,
+      "step": 3791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.38365700631875,
+      "learning_rate": 8.859889576432519e-06,
+      "loss": 0.8418,
+      "step": 3792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2297389903037965,
+      "learning_rate": 8.859230651096896e-06,
+      "loss": 0.8811,
+      "step": 3793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.296416039750523,
+      "learning_rate": 8.858571559921539e-06,
+      "loss": 0.8826,
+      "step": 3794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0524246486078317,
+      "learning_rate": 8.857912302934766e-06,
+      "loss": 0.7301,
+      "step": 3795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1779498399139667,
+      "learning_rate": 8.857252880164908e-06,
+      "loss": 0.773,
+      "step": 3796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.187962354051819,
+      "learning_rate": 8.856593291640305e-06,
+      "loss": 0.709,
+      "step": 3797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.04878049122168,
+      "learning_rate": 8.8559335373893e-06,
+      "loss": 0.7967,
+      "step": 3798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6888639234806373,
+      "learning_rate": 8.855273617440243e-06,
+      "loss": 0.8972,
+      "step": 3799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4752507914113524,
+      "learning_rate": 8.854613531821491e-06,
+      "loss": 0.8701,
+      "step": 3800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9791479707813897,
+      "learning_rate": 8.853953280561412e-06,
+      "loss": 0.6853,
+      "step": 3801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3296584640440345,
+      "learning_rate": 8.853292863688378e-06,
+      "loss": 0.7969,
+      "step": 3802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8581423482582924,
+      "learning_rate": 8.852632281230766e-06,
+      "loss": 0.9339,
+      "step": 3803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.657026175190554,
+      "learning_rate": 8.851971533216968e-06,
+      "loss": 0.8,
+      "step": 3804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0883360758617449,
+      "learning_rate": 8.85131061967537e-06,
+      "loss": 0.6589,
+      "step": 3805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8107865266089216,
+      "learning_rate": 8.85064954063438e-06,
+      "loss": 0.9036,
+      "step": 3806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.462742826457438,
+      "learning_rate": 8.849988296122402e-06,
+      "loss": 0.9809,
+      "step": 3807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9497018821229346,
+      "learning_rate": 8.849326886167854e-06,
+      "loss": 0.8126,
+      "step": 3808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.88386645641212,
+      "learning_rate": 8.848665310799156e-06,
+      "loss": 0.9167,
+      "step": 3809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6098423956911896,
+      "learning_rate": 8.84800357004474e-06,
+      "loss": 0.8761,
+      "step": 3810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8371808354810026,
+      "learning_rate": 8.847341663933038e-06,
+      "loss": 0.938,
+      "step": 3811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0253404025146877,
+      "learning_rate": 8.846679592492498e-06,
+      "loss": 0.6617,
+      "step": 3812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.203138350146886,
+      "learning_rate": 8.846017355751567e-06,
+      "loss": 1.0135,
+      "step": 3813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.273834366508432,
+      "learning_rate": 8.845354953738706e-06,
+      "loss": 0.9309,
+      "step": 3814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4748499569701954,
+      "learning_rate": 8.844692386482379e-06,
+      "loss": 0.6469,
+      "step": 3815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.710171974957078,
+      "learning_rate": 8.844029654011055e-06,
+      "loss": 0.928,
+      "step": 3816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3600339969053103,
+      "learning_rate": 8.84336675635322e-06,
+      "loss": 0.825,
+      "step": 3817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2366640210234705,
+      "learning_rate": 8.842703693537352e-06,
+      "loss": 0.6799,
+      "step": 3818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.320814980509048,
+      "learning_rate": 8.84204046559195e-06,
+      "loss": 0.8885,
+      "step": 3819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7507149401468105,
+      "learning_rate": 8.841377072545513e-06,
+      "loss": 0.8793,
+      "step": 3820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2573026991369622,
+      "learning_rate": 8.840713514426545e-06,
+      "loss": 0.7571,
+      "step": 3821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1528870101655897,
+      "learning_rate": 8.840049791263567e-06,
+      "loss": 0.7104,
+      "step": 3822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3073252842883853,
+      "learning_rate": 8.839385903085096e-06,
+      "loss": 0.7761,
+      "step": 3823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0104362430107003,
+      "learning_rate": 8.83872184991966e-06,
+      "loss": 0.9447,
+      "step": 3824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.568637741944457,
+      "learning_rate": 8.8380576317958e-06,
+      "loss": 0.8904,
+      "step": 3825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.680234311975607,
+      "learning_rate": 8.837393248742054e-06,
+      "loss": 0.752,
+      "step": 3826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0809704128153315,
+      "learning_rate": 8.836728700786974e-06,
+      "loss": 0.5373,
+      "step": 3827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.490460182158252,
+      "learning_rate": 8.836063987959116e-06,
+      "loss": 0.9713,
+      "step": 3828
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.793992340796723,
+      "learning_rate": 8.835399110287046e-06,
+      "loss": 0.8589,
+      "step": 3829
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9169791254287893,
+      "learning_rate": 8.834734067799336e-06,
+      "loss": 0.7832,
+      "step": 3830
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3602512837778067,
+      "learning_rate": 8.83406886052456e-06,
+      "loss": 0.6674,
+      "step": 3831
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1221171677822386,
+      "learning_rate": 8.833403488491308e-06,
+      "loss": 0.8034,
+      "step": 3832
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.627767059260879,
+      "learning_rate": 8.83273795172817e-06,
+      "loss": 0.789,
+      "step": 3833
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9647191928696779,
+      "learning_rate": 8.832072250263746e-06,
+      "loss": 1.0181,
+      "step": 3834
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9470955443850082,
+      "learning_rate": 8.831406384126643e-06,
+      "loss": 0.7823,
+      "step": 3835
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0983516230377246,
+      "learning_rate": 8.830740353345475e-06,
+      "loss": 0.933,
+      "step": 3836
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.017278079584237,
+      "learning_rate": 8.830074157948862e-06,
+      "loss": 0.8668,
+      "step": 3837
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1403412944371,
+      "learning_rate": 8.829407797965433e-06,
+      "loss": 0.7743,
+      "step": 3838
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.645024803764373,
+      "learning_rate": 8.82874127342382e-06,
+      "loss": 0.8487,
+      "step": 3839
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.446602110878232,
+      "learning_rate": 8.82807458435267e-06,
+      "loss": 0.8627,
+      "step": 3840
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.473913836450249,
+      "learning_rate": 8.82740773078063e-06,
+      "loss": 0.8697,
+      "step": 3841
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1158102473375155,
+      "learning_rate": 8.826740712736355e-06,
+      "loss": 0.7571,
+      "step": 3842
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0668930271220605,
+      "learning_rate": 8.826073530248508e-06,
+      "loss": 0.8826,
+      "step": 3843
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2956686077516837,
+      "learning_rate": 8.82540618334576e-06,
+      "loss": 0.902,
+      "step": 3844
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2898579123457314,
+      "learning_rate": 8.82473867205679e-06,
+      "loss": 0.6335,
+      "step": 3845
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0889840807788853,
+      "learning_rate": 8.824070996410281e-06,
+      "loss": 0.9352,
+      "step": 3846
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9735080075785407,
+      "learning_rate": 8.823403156434924e-06,
+      "loss": 0.8929,
+      "step": 3847
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9708237472928734,
+      "learning_rate": 8.822735152159418e-06,
+      "loss": 1.1327,
+      "step": 3848
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.251308803255633,
+      "learning_rate": 8.82206698361247e-06,
+      "loss": 0.8238,
+      "step": 3849
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.438656224083202,
+      "learning_rate": 8.82139865082279e-06,
+      "loss": 0.6799,
+      "step": 3850
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3283908182966986,
+      "learning_rate": 8.8207301538191e-06,
+      "loss": 0.954,
+      "step": 3851
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9013629228057773,
+      "learning_rate": 8.820061492630127e-06,
+      "loss": 0.8921,
+      "step": 3852
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8996009206550086,
+      "learning_rate": 8.819392667284603e-06,
+      "loss": 0.7276,
+      "step": 3853
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.881431675490626,
+      "learning_rate": 8.818723677811269e-06,
+      "loss": 0.9695,
+      "step": 3854
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1743627124890046,
+      "learning_rate": 8.818054524238874e-06,
+      "loss": 0.8084,
+      "step": 3855
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2062887581413986,
+      "learning_rate": 8.817385206596174e-06,
+      "loss": 0.6776,
+      "step": 3856
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9625786827572294,
+      "learning_rate": 8.81671572491193e-06,
+      "loss": 0.9033,
+      "step": 3857
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2157397884128334,
+      "learning_rate": 8.81604607921491e-06,
+      "loss": 0.6726,
+      "step": 3858
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4660165058771475,
+      "learning_rate": 8.815376269533893e-06,
+      "loss": 0.6795,
+      "step": 3859
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.305714082705911,
+      "learning_rate": 8.814706295897659e-06,
+      "loss": 0.9024,
+      "step": 3860
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.367792968845254,
+      "learning_rate": 8.814036158335e-06,
+      "loss": 0.8497,
+      "step": 3861
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1114708914804436,
+      "learning_rate": 8.813365856874711e-06,
+      "loss": 0.6402,
+      "step": 3862
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.222575984978837,
+      "learning_rate": 8.8126953915456e-06,
+      "loss": 0.8864,
+      "step": 3863
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.45558651295376,
+      "learning_rate": 8.812024762376477e-06,
+      "loss": 0.8513,
+      "step": 3864
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4959812700569435,
+      "learning_rate": 8.81135396939616e-06,
+      "loss": 0.8756,
+      "step": 3865
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2730922594824645,
+      "learning_rate": 8.810683012633475e-06,
+      "loss": 0.8698,
+      "step": 3866
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.49304447536743,
+      "learning_rate": 8.810011892117254e-06,
+      "loss": 0.8992,
+      "step": 3867
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1678463559512178,
+      "learning_rate": 8.809340607876335e-06,
+      "loss": 0.9024,
+      "step": 3868
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0527367107564336,
+      "learning_rate": 8.808669159939568e-06,
+      "loss": 0.8349,
+      "step": 3869
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8118280419352801,
+      "learning_rate": 8.807997548335806e-06,
+      "loss": 0.9231,
+      "step": 3870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7645406759498137,
+      "learning_rate": 8.807325773093904e-06,
+      "loss": 0.8163,
+      "step": 3871
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1024202590734054,
+      "learning_rate": 8.806653834242738e-06,
+      "loss": 0.8672,
+      "step": 3872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.055296171450371,
+      "learning_rate": 8.805981731811177e-06,
+      "loss": 0.945,
+      "step": 3873
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9145882120599331,
+      "learning_rate": 8.805309465828105e-06,
+      "loss": 0.7431,
+      "step": 3874
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2278357924198104,
+      "learning_rate": 8.804637036322411e-06,
+      "loss": 1.0232,
+      "step": 3875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8180820335233574,
+      "learning_rate": 8.80396444332299e-06,
+      "loss": 0.8716,
+      "step": 3876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5177358897575433,
+      "learning_rate": 8.803291686858745e-06,
+      "loss": 1.0077,
+      "step": 3877
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1681711910863326,
+      "learning_rate": 8.802618766958586e-06,
+      "loss": 0.901,
+      "step": 3878
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1172743217423777,
+      "learning_rate": 8.80194568365143e-06,
+      "loss": 0.6736,
+      "step": 3879
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8143226275172972,
+      "learning_rate": 8.8012724369662e-06,
+      "loss": 0.7895,
+      "step": 3880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3125228410017606,
+      "learning_rate": 8.800599026931827e-06,
+      "loss": 0.6835,
+      "step": 3881
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6540798448030474,
+      "learning_rate": 8.79992545357725e-06,
+      "loss": 0.9076,
+      "step": 3882
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0112127481842803,
+      "learning_rate": 8.799251716931412e-06,
+      "loss": 0.8267,
+      "step": 3883
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.617208885301534,
+      "learning_rate": 8.798577817023269e-06,
+      "loss": 0.6525,
+      "step": 3884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3421644993496984,
+      "learning_rate": 8.797903753881775e-06,
+      "loss": 0.6949,
+      "step": 3885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0602638499957466,
+      "learning_rate": 8.7972295275359e-06,
+      "loss": 0.9579,
+      "step": 3886
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4195349756735522,
+      "learning_rate": 8.796555138014613e-06,
+      "loss": 0.9137,
+      "step": 3887
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7132762288812713,
+      "learning_rate": 8.795880585346899e-06,
+      "loss": 0.8432,
+      "step": 3888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3948401586950068,
+      "learning_rate": 8.795205869561742e-06,
+      "loss": 0.9223,
+      "step": 3889
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.23467659570787,
+      "learning_rate": 8.794530990688135e-06,
+      "loss": 0.917,
+      "step": 3890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9648319907631535,
+      "learning_rate": 8.793855948755081e-06,
+      "loss": 0.9971,
+      "step": 3891
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.476303732432018,
+      "learning_rate": 8.793180743791587e-06,
+      "loss": 0.899,
+      "step": 3892
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.32691972455966,
+      "learning_rate": 8.79250537582667e-06,
+      "loss": 0.9405,
+      "step": 3893
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1454724704917527,
+      "learning_rate": 8.79182984488935e-06,
+      "loss": 0.6674,
+      "step": 3894
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6856590786039087,
+      "learning_rate": 8.791154151008657e-06,
+      "loss": 0.8993,
+      "step": 3895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0492305394270858,
+      "learning_rate": 8.790478294213626e-06,
+      "loss": 1.0251,
+      "step": 3896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0731005879457531,
+      "learning_rate": 8.789802274533301e-06,
+      "loss": 0.6295,
+      "step": 3897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3634838200055874,
+      "learning_rate": 8.789126091996731e-06,
+      "loss": 0.8356,
+      "step": 3898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4550608407218633,
+      "learning_rate": 8.788449746632976e-06,
+      "loss": 1.006,
+      "step": 3899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3462425491672994,
+      "learning_rate": 8.787773238471098e-06,
+      "loss": 0.9204,
+      "step": 3900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4765153251915883,
+      "learning_rate": 8.787096567540166e-06,
+      "loss": 1.0164,
+      "step": 3901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0170789364404933,
+      "learning_rate": 8.786419733869262e-06,
+      "loss": 0.8772,
+      "step": 3902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.052340132329809,
+      "learning_rate": 8.785742737487468e-06,
+      "loss": 0.7478,
+      "step": 3903
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.135287642024043,
+      "learning_rate": 8.78506557842388e-06,
+      "loss": 0.6366,
+      "step": 3904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1247320670914105,
+      "learning_rate": 8.78438825670759e-06,
+      "loss": 0.7292,
+      "step": 3905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0588988721056807,
+      "learning_rate": 8.78371077236771e-06,
+      "loss": 0.9852,
+      "step": 3906
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9865257281853637,
+      "learning_rate": 8.783033125433354e-06,
+      "loss": 0.7982,
+      "step": 3907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.125945420095094,
+      "learning_rate": 8.782355315933638e-06,
+      "loss": 0.9966,
+      "step": 3908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.007375204600878,
+      "learning_rate": 8.781677343897687e-06,
+      "loss": 0.9404,
+      "step": 3909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8377345026167724,
+      "learning_rate": 8.780999209354642e-06,
+      "loss": 0.6994,
+      "step": 3910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.861198208395082,
+      "learning_rate": 8.780320912333639e-06,
+      "loss": 0.8587,
+      "step": 3911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0936837897495053,
+      "learning_rate": 8.779642452863827e-06,
+      "loss": 0.7417,
+      "step": 3912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.654864699072151,
+      "learning_rate": 8.778963830974362e-06,
+      "loss": 0.8099,
+      "step": 3913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9287127536605972,
+      "learning_rate": 8.778285046694403e-06,
+      "loss": 0.7336,
+      "step": 3914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.072632151457663,
+      "learning_rate": 8.777606100053123e-06,
+      "loss": 0.8043,
+      "step": 3915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1633410857605126,
+      "learning_rate": 8.776926991079696e-06,
+      "loss": 0.7385,
+      "step": 3916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3191795674898152,
+      "learning_rate": 8.7762477198033e-06,
+      "loss": 0.7548,
+      "step": 3917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9459743676844703,
+      "learning_rate": 8.775568286253135e-06,
+      "loss": 0.8935,
+      "step": 3918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.066127309646664,
+      "learning_rate": 8.77488869045839e-06,
+      "loss": 0.6954,
+      "step": 3919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.259269143704889,
+      "learning_rate": 8.77420893244827e-06,
+      "loss": 0.8297,
+      "step": 3920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9507701223975835,
+      "learning_rate": 8.773529012251987e-06,
+      "loss": 0.8183,
+      "step": 3921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.50934360645919,
+      "learning_rate": 8.772848929898758e-06,
+      "loss": 0.9145,
+      "step": 3922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.228949453946301,
+      "learning_rate": 8.77216868541781e-06,
+      "loss": 1.0497,
+      "step": 3923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.174694309144655,
+      "learning_rate": 8.771488278838368e-06,
+      "loss": 1.0212,
+      "step": 3924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.982663568085766,
+      "learning_rate": 8.770807710189679e-06,
+      "loss": 0.8195,
+      "step": 3925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9271327233948468,
+      "learning_rate": 8.770126979500984e-06,
+      "loss": 0.7712,
+      "step": 3926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.177128560744698,
+      "learning_rate": 8.769446086801536e-06,
+      "loss": 0.7863,
+      "step": 3927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.698148922379841,
+      "learning_rate": 8.768765032120595e-06,
+      "loss": 0.8242,
+      "step": 3928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.607940870390153,
+      "learning_rate": 8.768083815487428e-06,
+      "loss": 0.8768,
+      "step": 3929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.829698167018722,
+      "learning_rate": 8.767402436931304e-06,
+      "loss": 0.8714,
+      "step": 3930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.242801956219165,
+      "learning_rate": 8.76672089648151e-06,
+      "loss": 0.8507,
+      "step": 3931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5850118577091927,
+      "learning_rate": 8.766039194167328e-06,
+      "loss": 0.836,
+      "step": 3932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4654786667303266,
+      "learning_rate": 8.765357330018056e-06,
+      "loss": 0.7399,
+      "step": 3933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.366880609403527,
+      "learning_rate": 8.764675304062992e-06,
+      "loss": 0.935,
+      "step": 3934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.227773839249988,
+      "learning_rate": 8.763993116331448e-06,
+      "loss": 0.9928,
+      "step": 3935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9457458109975807,
+      "learning_rate": 8.763310766852736e-06,
+      "loss": 0.9853,
+      "step": 3936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1022435922401272,
+      "learning_rate": 8.762628255656178e-06,
+      "loss": 0.5897,
+      "step": 3937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.411922363726163,
+      "learning_rate": 8.761945582771104e-06,
+      "loss": 0.9414,
+      "step": 3938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.320210392484668,
+      "learning_rate": 8.76126274822685e-06,
+      "loss": 0.7262,
+      "step": 3939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6906105740298933,
+      "learning_rate": 8.76057975205276e-06,
+      "loss": 0.8058,
+      "step": 3940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.30933258548634,
+      "learning_rate": 8.759896594278183e-06,
+      "loss": 0.7764,
+      "step": 3941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.439908166327285,
+      "learning_rate": 8.759213274932476e-06,
+      "loss": 0.694,
+      "step": 3942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.323831865139149,
+      "learning_rate": 8.758529794045e-06,
+      "loss": 0.7514,
+      "step": 3943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5945986378893675,
+      "learning_rate": 8.75784615164513e-06,
+      "loss": 0.8696,
+      "step": 3944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0505207653920139,
+      "learning_rate": 8.757162347762242e-06,
+      "loss": 0.7538,
+      "step": 3945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.391313440763793,
+      "learning_rate": 8.75647838242572e-06,
+      "loss": 0.8909,
+      "step": 3946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.304830124883328,
+      "learning_rate": 8.755794255664954e-06,
+      "loss": 0.9511,
+      "step": 3947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3282048209381734,
+      "learning_rate": 8.755109967509345e-06,
+      "loss": 0.8136,
+      "step": 3948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9837291238891974,
+      "learning_rate": 8.754425517988298e-06,
+      "loss": 0.6505,
+      "step": 3949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.347997097681044,
+      "learning_rate": 8.753740907131226e-06,
+      "loss": 0.7481,
+      "step": 3950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.237123895675599,
+      "learning_rate": 8.753056134967545e-06,
+      "loss": 1.0265,
+      "step": 3951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.472160872751201,
+      "learning_rate": 8.752371201526685e-06,
+      "loss": 0.9628,
+      "step": 3952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.16633315055691,
+      "learning_rate": 8.751686106838078e-06,
+      "loss": 0.859,
+      "step": 3953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.837703879397773,
+      "learning_rate": 8.751000850931162e-06,
+      "loss": 0.8837,
+      "step": 3954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5771679124338203,
+      "learning_rate": 8.750315433835387e-06,
+      "loss": 1.0149,
+      "step": 3955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.076963188305742,
+      "learning_rate": 8.749629855580204e-06,
+      "loss": 1.0985,
+      "step": 3956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4128002167089497,
+      "learning_rate": 8.748944116195076e-06,
+      "loss": 0.8768,
+      "step": 3957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1096848459807425,
+      "learning_rate": 8.74825821570947e-06,
+      "loss": 1.0062,
+      "step": 3958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.951383606456695,
+      "learning_rate": 8.74757215415286e-06,
+      "loss": 0.8024,
+      "step": 3959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.214434417282453,
+      "learning_rate": 8.74688593155473e-06,
+      "loss": 0.9913,
+      "step": 3960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.281452991589645,
+      "learning_rate": 8.746199547944565e-06,
+      "loss": 0.9275,
+      "step": 3961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.093522509304149,
+      "learning_rate": 8.745513003351862e-06,
+      "loss": 0.6878,
+      "step": 3962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.106651536836003,
+      "learning_rate": 8.744826297806124e-06,
+      "loss": 0.7582,
+      "step": 3963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1904000388184732,
+      "learning_rate": 8.74413943133686e-06,
+      "loss": 1.0219,
+      "step": 3964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2874318063745993,
+      "learning_rate": 8.743452403973586e-06,
+      "loss": 0.7767,
+      "step": 3965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3368874126911576,
+      "learning_rate": 8.742765215745825e-06,
+      "loss": 0.898,
+      "step": 3966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2503849439153694,
+      "learning_rate": 8.742077866683108e-06,
+      "loss": 0.6228,
+      "step": 3967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7331408868602702,
+      "learning_rate": 8.74139035681497e-06,
+      "loss": 0.7821,
+      "step": 3968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2039921098050588,
+      "learning_rate": 8.740702686170955e-06,
+      "loss": 0.8513,
+      "step": 3969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0907458874627562,
+      "learning_rate": 8.740014854780616e-06,
+      "loss": 0.7912,
+      "step": 3970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4807665802916534,
+      "learning_rate": 8.739326862673508e-06,
+      "loss": 0.8882,
+      "step": 3971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1307406988729474,
+      "learning_rate": 8.738638709879198e-06,
+      "loss": 0.6988,
+      "step": 3972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0729058908885674,
+      "learning_rate": 8.737950396427255e-06,
+      "loss": 0.6651,
+      "step": 3973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4140170129498735,
+      "learning_rate": 8.73726192234726e-06,
+      "loss": 0.6437,
+      "step": 3974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4037573295922425,
+      "learning_rate": 8.736573287668797e-06,
+      "loss": 1.0452,
+      "step": 3975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3436476038570015,
+      "learning_rate": 8.735884492421457e-06,
+      "loss": 0.8667,
+      "step": 3976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9803813296440023,
+      "learning_rate": 8.73519553663484e-06,
+      "loss": 1.0039,
+      "step": 3977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8590374098647646,
+      "learning_rate": 8.734506420338554e-06,
+      "loss": 0.814,
+      "step": 3978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.323347136158537,
+      "learning_rate": 8.733817143562207e-06,
+      "loss": 0.7903,
+      "step": 3979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.688493791845497,
+      "learning_rate": 8.733127706335423e-06,
+      "loss": 0.8654,
+      "step": 3980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2056309730914623,
+      "learning_rate": 8.732438108687829e-06,
+      "loss": 0.8874,
+      "step": 3981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.301917225557125,
+      "learning_rate": 8.731748350649054e-06,
+      "loss": 0.864,
+      "step": 3982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2253701844971334,
+      "learning_rate": 8.731058432248743e-06,
+      "loss": 0.8271,
+      "step": 3983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.861649774466908,
+      "learning_rate": 8.73036835351654e-06,
+      "loss": 0.7873,
+      "step": 3984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.365529664116272,
+      "learning_rate": 8.729678114482104e-06,
+      "loss": 0.6817,
+      "step": 3985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.302561723632632,
+      "learning_rate": 8.72898771517509e-06,
+      "loss": 1.1006,
+      "step": 3986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.621933299165559,
+      "learning_rate": 8.728297155625171e-06,
+      "loss": 0.7066,
+      "step": 3987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2264016306234446,
+      "learning_rate": 8.727606435862018e-06,
+      "loss": 0.7192,
+      "step": 3988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0709100706452888,
+      "learning_rate": 8.726915555915317e-06,
+      "loss": 0.9032,
+      "step": 3989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9370117222795555,
+      "learning_rate": 8.726224515814752e-06,
+      "loss": 0.8977,
+      "step": 3990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4243795982454115,
+      "learning_rate": 8.72553331559002e-06,
+      "loss": 0.8683,
+      "step": 3991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0583991710726433,
+      "learning_rate": 8.724841955270827e-06,
+      "loss": 0.921,
+      "step": 3992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.118098445174474,
+      "learning_rate": 8.724150434886878e-06,
+      "loss": 0.8442,
+      "step": 3993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9493851234400443,
+      "learning_rate": 8.723458754467893e-06,
+      "loss": 0.8897,
+      "step": 3994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3619079691041076,
+      "learning_rate": 8.72276691404359e-06,
+      "loss": 0.9001,
+      "step": 3995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4936977161226177,
+      "learning_rate": 8.722074913643703e-06,
+      "loss": 0.7461,
+      "step": 3996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9543116140452692,
+      "learning_rate": 8.721382753297967e-06,
+      "loss": 0.8229,
+      "step": 3997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2592517075252654,
+      "learning_rate": 8.720690433036125e-06,
+      "loss": 0.9058,
+      "step": 3998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4475986993040726,
+      "learning_rate": 8.719997952887932e-06,
+      "loss": 0.9259,
+      "step": 3999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2400352127453735,
+      "learning_rate": 8.719305312883137e-06,
+      "loss": 0.7536,
+      "step": 4000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2617966769483404,
+      "learning_rate": 8.718612513051513e-06,
+      "loss": 0.7176,
+      "step": 4001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7975800236785253,
+      "learning_rate": 8.717919553422827e-06,
+      "loss": 0.9221,
+      "step": 4002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2042000572265152,
+      "learning_rate": 8.717226434026858e-06,
+      "loss": 0.7138,
+      "step": 4003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.353543271683173,
+      "learning_rate": 8.71653315489339e-06,
+      "loss": 0.96,
+      "step": 4004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2258928832728992,
+      "learning_rate": 8.715839716052215e-06,
+      "loss": 1.0527,
+      "step": 4005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9514648435889934,
+      "learning_rate": 8.715146117533131e-06,
+      "loss": 0.8676,
+      "step": 4006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9108837619445243,
+      "learning_rate": 8.714452359365943e-06,
+      "loss": 0.6817,
+      "step": 4007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.080345781567984,
+      "learning_rate": 8.713758441580467e-06,
+      "loss": 0.6817,
+      "step": 4008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2010417872904922,
+      "learning_rate": 8.71306436420652e-06,
+      "loss": 0.6488,
+      "step": 4009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.191656700899036,
+      "learning_rate": 8.712370127273927e-06,
+      "loss": 0.8581,
+      "step": 4010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9642266360064449,
+      "learning_rate": 8.711675730812522e-06,
+      "loss": 0.9181,
+      "step": 4011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2145044516862598,
+      "learning_rate": 8.710981174852144e-06,
+      "loss": 0.663,
+      "step": 4012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9380891582937994,
+      "learning_rate": 8.710286459422641e-06,
+      "loss": 0.8026,
+      "step": 4013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.943079639408462,
+      "learning_rate": 8.709591584553865e-06,
+      "loss": 0.7928,
+      "step": 4014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0697208503690576,
+      "learning_rate": 8.708896550275675e-06,
+      "loss": 0.9841,
+      "step": 4015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9023771940803538,
+      "learning_rate": 8.708201356617945e-06,
+      "loss": 0.8983,
+      "step": 4016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6536194649829763,
+      "learning_rate": 8.70750600361054e-06,
+      "loss": 0.7182,
+      "step": 4017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4568017653102023,
+      "learning_rate": 8.706810491283346e-06,
+      "loss": 0.8397,
+      "step": 4018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4809562229890885,
+      "learning_rate": 8.706114819666249e-06,
+      "loss": 0.8537,
+      "step": 4019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1395971386420216,
+      "learning_rate": 8.705418988789146e-06,
+      "loss": 1.0131,
+      "step": 4020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.226460919672092,
+      "learning_rate": 8.704722998681937e-06,
+      "loss": 0.7385,
+      "step": 4021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.263168815282028,
+      "learning_rate": 8.704026849374527e-06,
+      "loss": 0.9889,
+      "step": 4022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.313704165631213,
+      "learning_rate": 8.703330540896836e-06,
+      "loss": 0.881,
+      "step": 4023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.345297476547644,
+      "learning_rate": 8.702634073278784e-06,
+      "loss": 1.1171,
+      "step": 4024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3557908660334614,
+      "learning_rate": 8.701937446550298e-06,
+      "loss": 0.6384,
+      "step": 4025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.229766920252198,
+      "learning_rate": 8.701240660741317e-06,
+      "loss": 0.9624,
+      "step": 4026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5548514588146722,
+      "learning_rate": 8.700543715881781e-06,
+      "loss": 0.8525,
+      "step": 4027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.443121555195883,
+      "learning_rate": 8.699846612001638e-06,
+      "loss": 0.9772,
+      "step": 4028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.043767868427078,
+      "learning_rate": 8.699149349130848e-06,
+      "loss": 0.7982,
+      "step": 4029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0835057889567867,
+      "learning_rate": 8.698451927299374e-06,
+      "loss": 0.8094,
+      "step": 4030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2590313214647053,
+      "learning_rate": 8.697754346537182e-06,
+      "loss": 0.7377,
+      "step": 4031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9802868548883985,
+      "learning_rate": 8.69705660687425e-06,
+      "loss": 0.7321,
+      "step": 4032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1734165739027755,
+      "learning_rate": 8.696358708340562e-06,
+      "loss": 0.9636,
+      "step": 4033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1173602704299914,
+      "learning_rate": 8.695660650966109e-06,
+      "loss": 0.8654,
+      "step": 4034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0918590854176549,
+      "learning_rate": 8.694962434780885e-06,
+      "loss": 0.5834,
+      "step": 4035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2929746541906137,
+      "learning_rate": 8.694264059814897e-06,
+      "loss": 0.8682,
+      "step": 4036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6722953605221136,
+      "learning_rate": 8.693565526098156e-06,
+      "loss": 0.9986,
+      "step": 4037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2821818818310438,
+      "learning_rate": 8.692866833660679e-06,
+      "loss": 0.7784,
+      "step": 4038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3291381137729052,
+      "learning_rate": 8.692167982532487e-06,
+      "loss": 0.9963,
+      "step": 4039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1356675322454866,
+      "learning_rate": 8.691468972743615e-06,
+      "loss": 0.8611,
+      "step": 4040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2980814000562892,
+      "learning_rate": 8.6907698043241e-06,
+      "loss": 0.8685,
+      "step": 4041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9433107255543693,
+      "learning_rate": 8.690070477303987e-06,
+      "loss": 0.7563,
+      "step": 4042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4130799568994608,
+      "learning_rate": 8.689370991713327e-06,
+      "loss": 0.8664,
+      "step": 4043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1679199613864357,
+      "learning_rate": 8.688671347582178e-06,
+      "loss": 0.8396,
+      "step": 4044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9934354617521997,
+      "learning_rate": 8.687971544940608e-06,
+      "loss": 0.813,
+      "step": 4045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.050822399422949,
+      "learning_rate": 8.687271583818687e-06,
+      "loss": 0.8386,
+      "step": 4046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.072710100998607,
+      "learning_rate": 8.686571464246491e-06,
+      "loss": 0.8099,
+      "step": 4047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.070297070361979,
+      "learning_rate": 8.685871186254112e-06,
+      "loss": 0.8757,
+      "step": 4048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0458137282782753,
+      "learning_rate": 8.685170749871638e-06,
+      "loss": 0.8014,
+      "step": 4049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.749764981736724,
+      "learning_rate": 8.68447015512917e-06,
+      "loss": 0.9095,
+      "step": 4050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8063354953200264,
+      "learning_rate": 8.683769402056814e-06,
+      "loss": 0.9733,
+      "step": 4051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1956724685004576,
+      "learning_rate": 8.683068490684681e-06,
+      "loss": 0.8665,
+      "step": 4052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3019612357643497,
+      "learning_rate": 8.682367421042895e-06,
+      "loss": 0.9565,
+      "step": 4053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8790161328834079,
+      "learning_rate": 8.681666193161578e-06,
+      "loss": 0.9089,
+      "step": 4054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3132810193913997,
+      "learning_rate": 8.680964807070865e-06,
+      "loss": 0.7667,
+      "step": 4055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.332697795602213,
+      "learning_rate": 8.680263262800897e-06,
+      "loss": 0.8544,
+      "step": 4056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.43799872950135,
+      "learning_rate": 8.679561560381818e-06,
+      "loss": 0.734,
+      "step": 4057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9464492408837122,
+      "learning_rate": 8.678859699843787e-06,
+      "loss": 0.8597,
+      "step": 4058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9978216045122459,
+      "learning_rate": 8.67815768121696e-06,
+      "loss": 0.8696,
+      "step": 4059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.0690879189523415,
+      "learning_rate": 8.677455504531507e-06,
+      "loss": 0.7199,
+      "step": 4060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9806715237968134,
+      "learning_rate": 8.676753169817598e-06,
+      "loss": 0.6963,
+      "step": 4061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2690878156446948,
+      "learning_rate": 8.676050677105419e-06,
+      "loss": 0.7915,
+      "step": 4062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3014294350466504,
+      "learning_rate": 8.675348026425154e-06,
+      "loss": 0.8639,
+      "step": 4063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3526399611012647,
+      "learning_rate": 8.674645217807e-06,
+      "loss": 0.9077,
+      "step": 4064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.463289512067439,
+      "learning_rate": 8.673942251281158e-06,
+      "loss": 0.7387,
+      "step": 4065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0005548893564016,
+      "learning_rate": 8.673239126877835e-06,
+      "loss": 0.7397,
+      "step": 4066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0647352791237898,
+      "learning_rate": 8.672535844627243e-06,
+      "loss": 0.6168,
+      "step": 4067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2564204089235225,
+      "learning_rate": 8.67183240455961e-06,
+      "loss": 0.7316,
+      "step": 4068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6937623278288823,
+      "learning_rate": 8.671128806705159e-06,
+      "loss": 1.0746,
+      "step": 4069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9495765542933392,
+      "learning_rate": 8.670425051094128e-06,
+      "loss": 0.7775,
+      "step": 4070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1471504976869586,
+      "learning_rate": 8.669721137756756e-06,
+      "loss": 0.8946,
+      "step": 4071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.492589899709145,
+      "learning_rate": 8.669017066723297e-06,
+      "loss": 0.7718,
+      "step": 4072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9957551399014662,
+      "learning_rate": 8.668312838024002e-06,
+      "loss": 0.6498,
+      "step": 4073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5976399253216567,
+      "learning_rate": 8.667608451689135e-06,
+      "loss": 0.8382,
+      "step": 4074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8854104116266175,
+      "learning_rate": 8.666903907748963e-06,
+      "loss": 0.7945,
+      "step": 4075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.104756915595179,
+      "learning_rate": 8.666199206233765e-06,
+      "loss": 0.9281,
+      "step": 4076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9112708223133372,
+      "learning_rate": 8.665494347173822e-06,
+      "loss": 0.9938,
+      "step": 4077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6242594306379887,
+      "learning_rate": 8.664789330599423e-06,
+      "loss": 0.9018,
+      "step": 4078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5460618430397166,
+      "learning_rate": 8.664084156540864e-06,
+      "loss": 0.8217,
+      "step": 4079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1477644413198975,
+      "learning_rate": 8.66337882502845e-06,
+      "loss": 0.7628,
+      "step": 4080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6866156074820775,
+      "learning_rate": 8.662673336092487e-06,
+      "loss": 0.8169,
+      "step": 4081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0698474187845877,
+      "learning_rate": 8.661967689763296e-06,
+      "loss": 0.9052,
+      "step": 4082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1218794783755315,
+      "learning_rate": 8.661261886071194e-06,
+      "loss": 0.9881,
+      "step": 4083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.184967305408809,
+      "learning_rate": 8.660555925046518e-06,
+      "loss": 0.8216,
+      "step": 4084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.612574528441582,
+      "learning_rate": 8.659849806719602e-06,
+      "loss": 1.0204,
+      "step": 4085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1644805550053228,
+      "learning_rate": 8.659143531120785e-06,
+      "loss": 0.9898,
+      "step": 4086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1435069530711153,
+      "learning_rate": 8.658437098280426e-06,
+      "loss": 0.9167,
+      "step": 4087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8301961880088293,
+      "learning_rate": 8.657730508228874e-06,
+      "loss": 0.8768,
+      "step": 4088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0356817190664342,
+      "learning_rate": 8.657023760996497e-06,
+      "loss": 0.7444,
+      "step": 4089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.062743720892345,
+      "learning_rate": 8.656316856613662e-06,
+      "loss": 0.7136,
+      "step": 4090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2876861285843955,
+      "learning_rate": 8.655609795110751e-06,
+      "loss": 0.8824,
+      "step": 4091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2758842941920574,
+      "learning_rate": 8.654902576518145e-06,
+      "loss": 0.8906,
+      "step": 4092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.910555997222323,
+      "learning_rate": 8.654195200866236e-06,
+      "loss": 0.9527,
+      "step": 4093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8596011933503513,
+      "learning_rate": 8.653487668185419e-06,
+      "loss": 0.9175,
+      "step": 4094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.061020330801834,
+      "learning_rate": 8.652779978506103e-06,
+      "loss": 0.8405,
+      "step": 4095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0636322881539972,
+      "learning_rate": 8.652072131858694e-06,
+      "loss": 0.8406,
+      "step": 4096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5557850979391663,
+      "learning_rate": 8.651364128273612e-06,
+      "loss": 1.0096,
+      "step": 4097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0384123972999704,
+      "learning_rate": 8.650655967781282e-06,
+      "loss": 0.7929,
+      "step": 4098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8223046514813104,
+      "learning_rate": 8.649947650412135e-06,
+      "loss": 0.9372,
+      "step": 4099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.002276668701185,
+      "learning_rate": 8.649239176196609e-06,
+      "loss": 0.8403,
+      "step": 4100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.330507421018653,
+      "learning_rate": 8.648530545165146e-06,
+      "loss": 0.6934,
+      "step": 4101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.104149385194072,
+      "learning_rate": 8.647821757348202e-06,
+      "loss": 0.9531,
+      "step": 4102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.42862570746783,
+      "learning_rate": 8.647112812776231e-06,
+      "loss": 0.7074,
+      "step": 4103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.274580803202587,
+      "learning_rate": 8.646403711479702e-06,
+      "loss": 0.6676,
+      "step": 4104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9660851969108355,
+      "learning_rate": 8.645694453489085e-06,
+      "loss": 1.0125,
+      "step": 4105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2166431560202264,
+      "learning_rate": 8.644985038834855e-06,
+      "loss": 0.8796,
+      "step": 4106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.765654767765466,
+      "learning_rate": 8.644275467547502e-06,
+      "loss": 0.737,
+      "step": 4107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.242934810836512,
+      "learning_rate": 8.643565739657515e-06,
+      "loss": 0.9537,
+      "step": 4108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.149955618153788,
+      "learning_rate": 8.642855855195394e-06,
+      "loss": 0.9344,
+      "step": 4109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9997011388140442,
+      "learning_rate": 8.642145814191643e-06,
+      "loss": 1.0197,
+      "step": 4110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8389430043992274,
+      "learning_rate": 8.641435616676776e-06,
+      "loss": 0.8948,
+      "step": 4111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1903620120032812,
+      "learning_rate": 8.64072526268131e-06,
+      "loss": 0.7226,
+      "step": 4112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9618742292946698,
+      "learning_rate": 8.640014752235773e-06,
+      "loss": 1.0166,
+      "step": 4113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9721121329340863,
+      "learning_rate": 8.639304085370692e-06,
+      "loss": 0.9502,
+      "step": 4114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.896761601203051,
+      "learning_rate": 8.638593262116612e-06,
+      "loss": 1.0781,
+      "step": 4115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.02010453348063,
+      "learning_rate": 8.637882282504075e-06,
+      "loss": 0.7609,
+      "step": 4116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.154938027849684,
+      "learning_rate": 8.637171146563634e-06,
+      "loss": 0.94,
+      "step": 4117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.016764407978776,
+      "learning_rate": 8.636459854325849e-06,
+      "loss": 0.9617,
+      "step": 4118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0952227401924493,
+      "learning_rate": 8.635748405821285e-06,
+      "loss": 1.1098,
+      "step": 4119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3259636236171275,
+      "learning_rate": 8.635036801080513e-06,
+      "loss": 0.9491,
+      "step": 4120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.780681388879135,
+      "learning_rate": 8.634325040134117e-06,
+      "loss": 0.9528,
+      "step": 4121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2637518157787953,
+      "learning_rate": 8.63361312301268e-06,
+      "loss": 1.1406,
+      "step": 4122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1464338859005156,
+      "learning_rate": 8.632901049746793e-06,
+      "loss": 0.9327,
+      "step": 4123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2350616535412637,
+      "learning_rate": 8.632188820367056e-06,
+      "loss": 0.8572,
+      "step": 4124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5694971958076254,
+      "learning_rate": 8.631476434904077e-06,
+      "loss": 0.9471,
+      "step": 4125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8502076307550164,
+      "learning_rate": 8.63076389338847e-06,
+      "loss": 0.948,
+      "step": 4126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.118533398810898,
+      "learning_rate": 8.630051195850851e-06,
+      "loss": 0.6698,
+      "step": 4127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2887646582566337,
+      "learning_rate": 8.629338342321846e-06,
+      "loss": 0.9773,
+      "step": 4128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1883714762577093,
+      "learning_rate": 8.62862533283209e-06,
+      "loss": 0.8503,
+      "step": 4129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3874166096824867,
+      "learning_rate": 8.627912167412222e-06,
+      "loss": 1.0251,
+      "step": 4130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2378130056381313,
+      "learning_rate": 8.62719884609289e-06,
+      "loss": 0.892,
+      "step": 4131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8988053521709563,
+      "learning_rate": 8.626485368904744e-06,
+      "loss": 0.9226,
+      "step": 4132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.630484566707074,
+      "learning_rate": 8.625771735878445e-06,
+      "loss": 0.7312,
+      "step": 4133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4660991393628942,
+      "learning_rate": 8.625057947044662e-06,
+      "loss": 0.6385,
+      "step": 4134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.02197811406415,
+      "learning_rate": 8.624344002434061e-06,
+      "loss": 0.7747,
+      "step": 4135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0432023930366128,
+      "learning_rate": 8.62362990207733e-06,
+      "loss": 0.7255,
+      "step": 4136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.107614899610059,
+      "learning_rate": 8.622915646005152e-06,
+      "loss": 0.6925,
+      "step": 4137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.137902134996767,
+      "learning_rate": 8.622201234248218e-06,
+      "loss": 0.7449,
+      "step": 4138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.885635779610183,
+      "learning_rate": 8.62148666683723e-06,
+      "loss": 0.8956,
+      "step": 4139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1598071528389142,
+      "learning_rate": 8.620771943802895e-06,
+      "loss": 0.9462,
+      "step": 4140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.884022200031417,
+      "learning_rate": 8.620057065175926e-06,
+      "loss": 0.7171,
+      "step": 4141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2496964830141402,
+      "learning_rate": 8.619342030987044e-06,
+      "loss": 0.7356,
+      "step": 4142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3070335228148773,
+      "learning_rate": 8.618626841266972e-06,
+      "loss": 0.7296,
+      "step": 4143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.392024600822441,
+      "learning_rate": 8.617911496046446e-06,
+      "loss": 0.9776,
+      "step": 4144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0039324247567776,
+      "learning_rate": 8.617195995356206e-06,
+      "loss": 0.6023,
+      "step": 4145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4949425298334607,
+      "learning_rate": 8.616480339226998e-06,
+      "loss": 0.7978,
+      "step": 4146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.716164530823717,
+      "learning_rate": 8.615764527689574e-06,
+      "loss": 0.839,
+      "step": 4147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1654976454321067,
+      "learning_rate": 8.615048560774698e-06,
+      "loss": 0.9448,
+      "step": 4148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8766550585553707,
+      "learning_rate": 8.614332438513132e-06,
+      "loss": 0.7926,
+      "step": 4149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2038587151964655,
+      "learning_rate": 8.613616160935652e-06,
+      "loss": 0.7153,
+      "step": 4150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9404774131165983,
+      "learning_rate": 8.612899728073039e-06,
+      "loss": 0.8396,
+      "step": 4151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.157834507079328,
+      "learning_rate": 8.612183139956078e-06,
+      "loss": 0.8218,
+      "step": 4152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9092685327027925,
+      "learning_rate": 8.611466396615562e-06,
+      "loss": 0.8927,
+      "step": 4153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.025954746460297,
+      "learning_rate": 8.610749498082291e-06,
+      "loss": 0.8618,
+      "step": 4154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.037277964863612,
+      "learning_rate": 8.610032444387074e-06,
+      "loss": 0.8653,
+      "step": 4155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0173266545537,
+      "learning_rate": 8.609315235560722e-06,
+      "loss": 0.9346,
+      "step": 4156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3939101973267776,
+      "learning_rate": 8.608597871634056e-06,
+      "loss": 0.9103,
+      "step": 4157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1947114079218704,
+      "learning_rate": 8.607880352637905e-06,
+      "loss": 0.7959,
+      "step": 4158
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1943224256817437,
+      "learning_rate": 8.607162678603097e-06,
+      "loss": 0.9861,
+      "step": 4159
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1081652370994817,
+      "learning_rate": 8.606444849560476e-06,
+      "loss": 0.5759,
+      "step": 4160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4482762650190435,
+      "learning_rate": 8.605726865540889e-06,
+      "loss": 0.8044,
+      "step": 4161
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9332157256827731,
+      "learning_rate": 8.605008726575186e-06,
+      "loss": 0.8037,
+      "step": 4162
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4863670065490404,
+      "learning_rate": 8.60429043269423e-06,
+      "loss": 0.7327,
+      "step": 4163
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.0576587311492345,
+      "learning_rate": 8.603571983928888e-06,
+      "loss": 0.8584,
+      "step": 4164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4895258759900893,
+      "learning_rate": 8.602853380310033e-06,
+      "loss": 0.836,
+      "step": 4165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.052462841423401,
+      "learning_rate": 8.602134621868542e-06,
+      "loss": 0.798,
+      "step": 4166
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.04775167601586,
+      "learning_rate": 8.601415708635304e-06,
+      "loss": 0.9331,
+      "step": 4167
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0903437192974395,
+      "learning_rate": 8.600696640641213e-06,
+      "loss": 0.6106,
+      "step": 4168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7650449122956233,
+      "learning_rate": 8.599977417917169e-06,
+      "loss": 0.8236,
+      "step": 4169
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3319862960611566,
+      "learning_rate": 8.599258040494078e-06,
+      "loss": 0.8978,
+      "step": 4170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0256042908379954,
+      "learning_rate": 8.59853850840285e-06,
+      "loss": 0.8744,
+      "step": 4171
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5140342574922654,
+      "learning_rate": 8.59781882167441e-06,
+      "loss": 0.8575,
+      "step": 4172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9430693562071657,
+      "learning_rate": 8.597098980339683e-06,
+      "loss": 0.9343,
+      "step": 4173
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3124352290068946,
+      "learning_rate": 8.5963789844296e-06,
+      "loss": 0.8645,
+      "step": 4174
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.207981197224037,
+      "learning_rate": 8.595658833975104e-06,
+      "loss": 0.735,
+      "step": 4175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0238018130490354,
+      "learning_rate": 8.59493852900714e-06,
+      "loss": 0.7773,
+      "step": 4176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8926979064388707,
+      "learning_rate": 8.59421806955666e-06,
+      "loss": 1.095,
+      "step": 4177
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2703168055839362,
+      "learning_rate": 8.593497455654627e-06,
+      "loss": 0.7873,
+      "step": 4178
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0350701471344217,
+      "learning_rate": 8.592776687332003e-06,
+      "loss": 1.1065,
+      "step": 4179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9176122486351923,
+      "learning_rate": 8.592055764619762e-06,
+      "loss": 0.7391,
+      "step": 4180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.163639163927914,
+      "learning_rate": 8.591334687548888e-06,
+      "loss": 0.9008,
+      "step": 4181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.328602949263451,
+      "learning_rate": 8.590613456150364e-06,
+      "loss": 0.8642,
+      "step": 4182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4219223228402385,
+      "learning_rate": 8.58989207045518e-06,
+      "loss": 0.8605,
+      "step": 4183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5333501998049432,
+      "learning_rate": 8.58917053049434e-06,
+      "loss": 0.9814,
+      "step": 4184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.476871860421355,
+      "learning_rate": 8.58844883629885e-06,
+      "loss": 0.9151,
+      "step": 4185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0531695100262033,
+      "learning_rate": 8.58772698789972e-06,
+      "loss": 0.6328,
+      "step": 4186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.132546386570587,
+      "learning_rate": 8.587004985327971e-06,
+      "loss": 1.0176,
+      "step": 4187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4634409291959156,
+      "learning_rate": 8.586282828614631e-06,
+      "loss": 0.8893,
+      "step": 4188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0303468048735516,
+      "learning_rate": 8.58556051779073e-06,
+      "loss": 0.9711,
+      "step": 4189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3351078181253975,
+      "learning_rate": 8.584838052887308e-06,
+      "loss": 0.864,
+      "step": 4190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.779705276621271,
+      "learning_rate": 8.58411543393541e-06,
+      "loss": 0.855,
+      "step": 4191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.086231158537961,
+      "learning_rate": 8.58339266096609e-06,
+      "loss": 0.6353,
+      "step": 4192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.088285493776697,
+      "learning_rate": 8.582669734010407e-06,
+      "loss": 0.8079,
+      "step": 4193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9415376808057199,
+      "learning_rate": 8.581946653099427e-06,
+      "loss": 0.7985,
+      "step": 4194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.773818862930254,
+      "learning_rate": 8.581223418264222e-06,
+      "loss": 1.0388,
+      "step": 4195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.017908625273765,
+      "learning_rate": 8.58050002953587e-06,
+      "loss": 1.0093,
+      "step": 4196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2423865274867283,
+      "learning_rate": 8.579776486945457e-06,
+      "loss": 0.9607,
+      "step": 4197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0847179211461477,
+      "learning_rate": 8.579052790524077e-06,
+      "loss": 1.071,
+      "step": 4198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4049361214729896,
+      "learning_rate": 8.578328940302827e-06,
+      "loss": 0.898,
+      "step": 4199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.991486485845755,
+      "learning_rate": 8.577604936312813e-06,
+      "loss": 0.9437,
+      "step": 4200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0195513657706425,
+      "learning_rate": 8.576880778585148e-06,
+      "loss": 0.6263,
+      "step": 4201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0808976448255587,
+      "learning_rate": 8.57615646715095e-06,
+      "loss": 0.7372,
+      "step": 4202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.94400938377132,
+      "learning_rate": 8.575432002041341e-06,
+      "loss": 0.9148,
+      "step": 4203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0314998404755933,
+      "learning_rate": 8.574707383287459e-06,
+      "loss": 0.7166,
+      "step": 4204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.265196019458165,
+      "learning_rate": 8.57398261092044e-06,
+      "loss": 0.8901,
+      "step": 4205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2360460513862903,
+      "learning_rate": 8.573257684971425e-06,
+      "loss": 0.8165,
+      "step": 4206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.864407865721783,
+      "learning_rate": 8.572532605471572e-06,
+      "loss": 0.8647,
+      "step": 4207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4474299825047896,
+      "learning_rate": 8.571807372452036e-06,
+      "loss": 0.9826,
+      "step": 4208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.627561192994462,
+      "learning_rate": 8.571081985943984e-06,
+      "loss": 1.0022,
+      "step": 4209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1343457661386953,
+      "learning_rate": 8.570356445978583e-06,
+      "loss": 0.932,
+      "step": 4210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8996193813319593,
+      "learning_rate": 8.569630752587014e-06,
+      "loss": 0.8479,
+      "step": 4211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3399094066907518,
+      "learning_rate": 8.568904905800464e-06,
+      "loss": 0.7455,
+      "step": 4212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4529150676823734,
+      "learning_rate": 8.56817890565012e-06,
+      "loss": 0.6952,
+      "step": 4213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.164591223651343,
+      "learning_rate": 8.567452752167183e-06,
+      "loss": 0.6841,
+      "step": 4214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.138917665197672,
+      "learning_rate": 8.566726445382854e-06,
+      "loss": 0.6611,
+      "step": 4215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8814471224971654,
+      "learning_rate": 8.565999985328348e-06,
+      "loss": 0.8345,
+      "step": 4216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1298412094774046,
+      "learning_rate": 8.565273372034879e-06,
+      "loss": 0.9542,
+      "step": 4217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8799116225722816,
+      "learning_rate": 8.564546605533674e-06,
+      "loss": 0.8963,
+      "step": 4218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2740725691296837,
+      "learning_rate": 8.563819685855963e-06,
+      "loss": 0.7339,
+      "step": 4219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1138850281539363,
+      "learning_rate": 8.563092613032981e-06,
+      "loss": 0.8358,
+      "step": 4220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2982067167599247,
+      "learning_rate": 8.562365387095977e-06,
+      "loss": 0.6464,
+      "step": 4221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1766550969848848,
+      "learning_rate": 8.561638008076197e-06,
+      "loss": 0.6796,
+      "step": 4222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.38972770203917,
+      "learning_rate": 8.5609104760049e-06,
+      "loss": 0.637,
+      "step": 4223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0816339018572756,
+      "learning_rate": 8.560182790913349e-06,
+      "loss": 0.8362,
+      "step": 4224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.411668505920211,
+      "learning_rate": 8.559454952832815e-06,
+      "loss": 0.9504,
+      "step": 4225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7781491215221426,
+      "learning_rate": 8.558726961794573e-06,
+      "loss": 0.8628,
+      "step": 4226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0602598588504974,
+      "learning_rate": 8.557998817829909e-06,
+      "loss": 0.6495,
+      "step": 4227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5854668947080643,
+      "learning_rate": 8.557270520970111e-06,
+      "loss": 0.8746,
+      "step": 4228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2770201383696533,
+      "learning_rate": 8.556542071246476e-06,
+      "loss": 0.9307,
+      "step": 4229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.429569449042359,
+      "learning_rate": 8.555813468690309e-06,
+      "loss": 0.9161,
+      "step": 4230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.456421897559217,
+      "learning_rate": 8.555084713332917e-06,
+      "loss": 0.9081,
+      "step": 4231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5010283519944543,
+      "learning_rate": 8.554355805205616e-06,
+      "loss": 0.8065,
+      "step": 4232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9833524735980432,
+      "learning_rate": 8.553626744339732e-06,
+      "loss": 0.6287,
+      "step": 4233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0970556301525884,
+      "learning_rate": 8.552897530766592e-06,
+      "loss": 0.8691,
+      "step": 4234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8382205468191026,
+      "learning_rate": 8.552168164517532e-06,
+      "loss": 0.9398,
+      "step": 4235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3718470824085816,
+      "learning_rate": 8.551438645623896e-06,
+      "loss": 0.8634,
+      "step": 4236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.068209483362169,
+      "learning_rate": 8.55070897411703e-06,
+      "loss": 0.8267,
+      "step": 4237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1402249382147855,
+      "learning_rate": 8.549979150028292e-06,
+      "loss": 0.7157,
+      "step": 4238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.315245129423356,
+      "learning_rate": 8.549249173389045e-06,
+      "loss": 0.8437,
+      "step": 4239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1414364547596163,
+      "learning_rate": 8.548519044230654e-06,
+      "loss": 0.6957,
+      "step": 4240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7569378831596287,
+      "learning_rate": 8.5477887625845e-06,
+      "loss": 0.82,
+      "step": 4241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4387307390483637,
+      "learning_rate": 8.547058328481959e-06,
+      "loss": 0.669,
+      "step": 4242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1242000542673103,
+      "learning_rate": 8.546327741954423e-06,
+      "loss": 0.7878,
+      "step": 4243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1177146383145438,
+      "learning_rate": 8.545597003033286e-06,
+      "loss": 0.8948,
+      "step": 4244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9272929154867742,
+      "learning_rate": 8.544866111749948e-06,
+      "loss": 0.8271,
+      "step": 4245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.510582900386849,
+      "learning_rate": 8.544135068135819e-06,
+      "loss": 0.8364,
+      "step": 4246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2524688681314524,
+      "learning_rate": 8.543403872222313e-06,
+      "loss": 0.7738,
+      "step": 4247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.043711643355393,
+      "learning_rate": 8.54267252404085e-06,
+      "loss": 0.9761,
+      "step": 4248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.399273414960344,
+      "learning_rate": 8.54194102362286e-06,
+      "loss": 0.8588,
+      "step": 4249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.991158434414303,
+      "learning_rate": 8.541209370999777e-06,
+      "loss": 0.8217,
+      "step": 4250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.862751785215958,
+      "learning_rate": 8.540477566203039e-06,
+      "loss": 0.8877,
+      "step": 4251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1338931065795284,
+      "learning_rate": 8.539745609264094e-06,
+      "loss": 0.7684,
+      "step": 4252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.189905848918212,
+      "learning_rate": 8.539013500214399e-06,
+      "loss": 0.9748,
+      "step": 4253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6631690794378904,
+      "learning_rate": 8.538281239085411e-06,
+      "loss": 0.7752,
+      "step": 4254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3786385551510816,
+      "learning_rate": 8.537548825908597e-06,
+      "loss": 0.9775,
+      "step": 4255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5182373527711217,
+      "learning_rate": 8.536816260715433e-06,
+      "loss": 0.7184,
+      "step": 4256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1385788514239126,
+      "learning_rate": 8.536083543537396e-06,
+      "loss": 1.0153,
+      "step": 4257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0650625747918845,
+      "learning_rate": 8.535350674405975e-06,
+      "loss": 0.9053,
+      "step": 4258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.09026759918533,
+      "learning_rate": 8.534617653352661e-06,
+      "loss": 0.9145,
+      "step": 4259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1834805527028796,
+      "learning_rate": 8.533884480408955e-06,
+      "loss": 0.6457,
+      "step": 4260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2861683877454326,
+      "learning_rate": 8.533151155606364e-06,
+      "loss": 0.686,
+      "step": 4261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.409259222062061,
+      "learning_rate": 8.532417678976398e-06,
+      "loss": 0.9233,
+      "step": 4262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0576234113596126,
+      "learning_rate": 8.531684050550575e-06,
+      "loss": 0.8782,
+      "step": 4263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.3051331745359755,
+      "learning_rate": 8.530950270360425e-06,
+      "loss": 0.769,
+      "step": 4264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5613811613993933,
+      "learning_rate": 8.530216338437478e-06,
+      "loss": 0.7849,
+      "step": 4265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8664455117855174,
+      "learning_rate": 8.529482254813272e-06,
+      "loss": 0.8115,
+      "step": 4266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5600114870499504,
+      "learning_rate": 8.528748019519352e-06,
+      "loss": 0.6446,
+      "step": 4267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.416821405468037,
+      "learning_rate": 8.528013632587272e-06,
+      "loss": 0.7815,
+      "step": 4268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.21103687342019,
+      "learning_rate": 8.52727909404859e-06,
+      "loss": 0.8724,
+      "step": 4269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.283948941639288,
+      "learning_rate": 8.526544403934868e-06,
+      "loss": 0.966,
+      "step": 4270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8618203132831814,
+      "learning_rate": 8.52580956227768e-06,
+      "loss": 0.7498,
+      "step": 4271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.69031781349236,
+      "learning_rate": 8.525074569108603e-06,
+      "loss": 0.9681,
+      "step": 4272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2943763299121867,
+      "learning_rate": 8.524339424459219e-06,
+      "loss": 0.8512,
+      "step": 4273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4560164076117315,
+      "learning_rate": 8.523604128361123e-06,
+      "loss": 0.7178,
+      "step": 4274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8884383911476796,
+      "learning_rate": 8.522868680845908e-06,
+      "loss": 0.6854,
+      "step": 4275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.04124909345954,
+      "learning_rate": 8.52213308194518e-06,
+      "loss": 0.6409,
+      "step": 4276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9664762542860859,
+      "learning_rate": 8.521397331690551e-06,
+      "loss": 0.7963,
+      "step": 4277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.618626957996641,
+      "learning_rate": 8.520661430113637e-06,
+      "loss": 1.0816,
+      "step": 4278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1954569909725734,
+      "learning_rate": 8.519925377246057e-06,
+      "loss": 0.9465,
+      "step": 4279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2727252414538746,
+      "learning_rate": 8.519189173119446e-06,
+      "loss": 0.9044,
+      "step": 4280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.28345470663054,
+      "learning_rate": 8.518452817765438e-06,
+      "loss": 0.9637,
+      "step": 4281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.161304242004154,
+      "learning_rate": 8.517716311215678e-06,
+      "loss": 0.8062,
+      "step": 4282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.929983174912129,
+      "learning_rate": 8.516979653501813e-06,
+      "loss": 0.7853,
+      "step": 4283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.063087774740537,
+      "learning_rate": 8.516242844655498e-06,
+      "loss": 0.8375,
+      "step": 4284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.200002062628562,
+      "learning_rate": 8.515505884708399e-06,
+      "loss": 0.8349,
+      "step": 4285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.977493166039945,
+      "learning_rate": 8.514768773692182e-06,
+      "loss": 0.7752,
+      "step": 4286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2335759979552443,
+      "learning_rate": 8.514031511638524e-06,
+      "loss": 0.9561,
+      "step": 4287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1730123302326882,
+      "learning_rate": 8.513294098579103e-06,
+      "loss": 1.0634,
+      "step": 4288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3262524871942105,
+      "learning_rate": 8.512556534545612e-06,
+      "loss": 0.9509,
+      "step": 4289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9395846992310861,
+      "learning_rate": 8.511818819569743e-06,
+      "loss": 0.8002,
+      "step": 4290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.35478543460504,
+      "learning_rate": 8.5110809536832e-06,
+      "loss": 0.7743,
+      "step": 4291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.241232071766614,
+      "learning_rate": 8.510342936917685e-06,
+      "loss": 1.0179,
+      "step": 4292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.974074163071248,
+      "learning_rate": 8.50960476930492e-06,
+      "loss": 0.8126,
+      "step": 4293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.931507992783669,
+      "learning_rate": 8.50886645087662e-06,
+      "loss": 0.8069,
+      "step": 4294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1119043743040629,
+      "learning_rate": 8.508127981664514e-06,
+      "loss": 0.7039,
+      "step": 4295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9143479643132362,
+      "learning_rate": 8.507389361700335e-06,
+      "loss": 0.96,
+      "step": 4296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0991178266875712,
+      "learning_rate": 8.506650591015825e-06,
+      "loss": 0.6697,
+      "step": 4297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8992124509103265,
+      "learning_rate": 8.50591166964273e-06,
+      "loss": 0.8345,
+      "step": 4298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1748272357455294,
+      "learning_rate": 8.5051725976128e-06,
+      "loss": 0.8782,
+      "step": 4299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.130796968458989,
+      "learning_rate": 8.504433374957799e-06,
+      "loss": 0.8382,
+      "step": 4300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.07060670894323,
+      "learning_rate": 8.50369400170949e-06,
+      "loss": 0.9397,
+      "step": 4301
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.021420521336371,
+      "learning_rate": 8.502954477899647e-06,
+      "loss": 0.7329,
+      "step": 4302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.71103850251842,
+      "learning_rate": 8.50221480356005e-06,
+      "loss": 0.8225,
+      "step": 4303
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.390439153150531,
+      "learning_rate": 8.50147497872248e-06,
+      "loss": 0.9052,
+      "step": 4304
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.353411855744115,
+      "learning_rate": 8.500735003418734e-06,
+      "loss": 0.8958,
+      "step": 4305
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2766705865824624,
+      "learning_rate": 8.499994877680609e-06,
+      "loss": 0.6544,
+      "step": 4306
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0100654896523664,
+      "learning_rate": 8.499254601539908e-06,
+      "loss": 0.6857,
+      "step": 4307
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.773921104618756,
+      "learning_rate": 8.498514175028442e-06,
+      "loss": 0.6613,
+      "step": 4308
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6030692640900535,
+      "learning_rate": 8.497773598178033e-06,
+      "loss": 0.7853,
+      "step": 4309
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.029582289971877,
+      "learning_rate": 8.497032871020501e-06,
+      "loss": 0.7491,
+      "step": 4310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3880023891604782,
+      "learning_rate": 8.49629199358768e-06,
+      "loss": 0.8174,
+      "step": 4311
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0039877505470995,
+      "learning_rate": 8.495550965911403e-06,
+      "loss": 0.8531,
+      "step": 4312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.235421817548423,
+      "learning_rate": 8.494809788023518e-06,
+      "loss": 0.8259,
+      "step": 4313
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8554039997953624,
+      "learning_rate": 8.494068459955871e-06,
+      "loss": 0.9328,
+      "step": 4314
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0497974163546828,
+      "learning_rate": 8.493326981740322e-06,
+      "loss": 0.6576,
+      "step": 4315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.153025592591107,
+      "learning_rate": 8.492585353408732e-06,
+      "loss": 0.8854,
+      "step": 4316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4563324084555864,
+      "learning_rate": 8.491843574992971e-06,
+      "loss": 1.0508,
+      "step": 4317
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7883241615119432,
+      "learning_rate": 8.491101646524916e-06,
+      "loss": 0.9476,
+      "step": 4318
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1400508324192633,
+      "learning_rate": 8.490359568036446e-06,
+      "loss": 0.9132,
+      "step": 4319
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.018625199467526,
+      "learning_rate": 8.489617339559455e-06,
+      "loss": 0.6647,
+      "step": 4320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.050947434318041,
+      "learning_rate": 8.488874961125832e-06,
+      "loss": 0.884,
+      "step": 4321
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2426610413578985,
+      "learning_rate": 8.488132432767483e-06,
+      "loss": 0.7545,
+      "step": 4322
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8762017806676314,
+      "learning_rate": 8.487389754516315e-06,
+      "loss": 0.7592,
+      "step": 4323
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1800594550997099,
+      "learning_rate": 8.486646926404243e-06,
+      "loss": 0.6135,
+      "step": 4324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.16813100057715,
+      "learning_rate": 8.485903948463185e-06,
+      "loss": 0.8502,
+      "step": 4325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.086028187153477,
+      "learning_rate": 8.485160820725073e-06,
+      "loss": 0.8022,
+      "step": 4326
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1838426959681272,
+      "learning_rate": 8.484417543221839e-06,
+      "loss": 0.7185,
+      "step": 4327
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7580756510592224,
+      "learning_rate": 8.483674115985421e-06,
+      "loss": 0.8574,
+      "step": 4328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8441924674968988,
+      "learning_rate": 8.48293053904777e-06,
+      "loss": 0.7662,
+      "step": 4329
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.018120319993391,
+      "learning_rate": 8.482186812440836e-06,
+      "loss": 0.8644,
+      "step": 4330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.023599055321104,
+      "learning_rate": 8.481442936196578e-06,
+      "loss": 0.8817,
+      "step": 4331
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0874442958559112,
+      "learning_rate": 8.480698910346965e-06,
+      "loss": 0.5996,
+      "step": 4332
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1773172990767233,
+      "learning_rate": 8.479954734923967e-06,
+      "loss": 0.9087,
+      "step": 4333
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3612987699413983,
+      "learning_rate": 8.479210409959565e-06,
+      "loss": 0.7172,
+      "step": 4334
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3530094126319607,
+      "learning_rate": 8.478465935485741e-06,
+      "loss": 0.8807,
+      "step": 4335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3500281388111572,
+      "learning_rate": 8.47772131153449e-06,
+      "loss": 0.9083,
+      "step": 4336
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4109794581418353,
+      "learning_rate": 8.476976538137809e-06,
+      "loss": 0.8397,
+      "step": 4337
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0638171894022963,
+      "learning_rate": 8.476231615327703e-06,
+      "loss": 0.873,
+      "step": 4338
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0673728161336,
+      "learning_rate": 8.475486543136181e-06,
+      "loss": 1.0203,
+      "step": 4339
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.664822300540005,
+      "learning_rate": 8.474741321595263e-06,
+      "loss": 0.882,
+      "step": 4340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.049554929125958,
+      "learning_rate": 8.473995950736973e-06,
+      "loss": 0.6814,
+      "step": 4341
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1647728051906467,
+      "learning_rate": 8.473250430593338e-06,
+      "loss": 0.8802,
+      "step": 4342
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1519387462960005,
+      "learning_rate": 8.472504761196397e-06,
+      "loss": 0.717,
+      "step": 4343
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3427913483164646,
+      "learning_rate": 8.471758942578193e-06,
+      "loss": 0.7848,
+      "step": 4344
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.901319286606905,
+      "learning_rate": 8.471012974770776e-06,
+      "loss": 0.778,
+      "step": 4345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9096996897831477,
+      "learning_rate": 8.4702668578062e-06,
+      "loss": 0.8167,
+      "step": 4346
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.522793081399012,
+      "learning_rate": 8.46952059171653e-06,
+      "loss": 0.8471,
+      "step": 4347
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1515909875715575,
+      "learning_rate": 8.468774176533831e-06,
+      "loss": 0.8691,
+      "step": 4348
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.059531506513637,
+      "learning_rate": 8.46802761229018e-06,
+      "loss": 0.6619,
+      "step": 4349
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3294648637157245,
+      "learning_rate": 8.46728089901766e-06,
+      "loss": 1.0043,
+      "step": 4350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9522284101466933,
+      "learning_rate": 8.466534036748359e-06,
+      "loss": 0.732,
+      "step": 4351
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.809849450214524,
+      "learning_rate": 8.465787025514368e-06,
+      "loss": 0.8013,
+      "step": 4352
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6278515996226375,
+      "learning_rate": 8.465039865347791e-06,
+      "loss": 0.7176,
+      "step": 4353
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.019539654171552,
+      "learning_rate": 8.464292556280734e-06,
+      "loss": 0.7088,
+      "step": 4354
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3009627340760863,
+      "learning_rate": 8.463545098345308e-06,
+      "loss": 0.7757,
+      "step": 4355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9906178425175456,
+      "learning_rate": 8.462797491573638e-06,
+      "loss": 0.7621,
+      "step": 4356
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.182586185377908,
+      "learning_rate": 8.462049735997848e-06,
+      "loss": 0.974,
+      "step": 4357
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2840373333410886,
+      "learning_rate": 8.46130183165007e-06,
+      "loss": 0.7978,
+      "step": 4358
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4297985746354174,
+      "learning_rate": 8.46055377856244e-06,
+      "loss": 0.9612,
+      "step": 4359
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2339108652669863,
+      "learning_rate": 8.459805576767111e-06,
+      "loss": 0.9005,
+      "step": 4360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4474874888522513,
+      "learning_rate": 8.459057226296232e-06,
+      "loss": 0.9174,
+      "step": 4361
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.381318532272321,
+      "learning_rate": 8.458308727181956e-06,
+      "loss": 0.956,
+      "step": 4362
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.168082663571989,
+      "learning_rate": 8.457560079456455e-06,
+      "loss": 0.6464,
+      "step": 4363
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6758822838140732,
+      "learning_rate": 8.456811283151896e-06,
+      "loss": 0.9157,
+      "step": 4364
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3591055352853623,
+      "learning_rate": 8.456062338300458e-06,
+      "loss": 0.8666,
+      "step": 4365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0949952139251273,
+      "learning_rate": 8.455313244934324e-06,
+      "loss": 0.6861,
+      "step": 4366
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1214075945397526,
+      "learning_rate": 8.454564003085685e-06,
+      "loss": 0.9529,
+      "step": 4367
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.389862532471913,
+      "learning_rate": 8.453814612786736e-06,
+      "loss": 0.7604,
+      "step": 4368
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0849048337657554,
+      "learning_rate": 8.453065074069682e-06,
+      "loss": 0.9221,
+      "step": 4369
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.707644043480254,
+      "learning_rate": 8.45231538696673e-06,
+      "loss": 0.9254,
+      "step": 4370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4626575197916036,
+      "learning_rate": 8.451565551510097e-06,
+      "loss": 0.8316,
+      "step": 4371
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.278603868568913,
+      "learning_rate": 8.450815567732007e-06,
+      "loss": 0.7811,
+      "step": 4372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9942330946344768,
+      "learning_rate": 8.450065435664686e-06,
+      "loss": 0.7125,
+      "step": 4373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9793160263892866,
+      "learning_rate": 8.449315155340369e-06,
+      "loss": 0.9645,
+      "step": 4374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1429750706148507,
+      "learning_rate": 8.4485647267913e-06,
+      "loss": 0.886,
+      "step": 4375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8531629024640173,
+      "learning_rate": 8.44781415004972e-06,
+      "loss": 1.0042,
+      "step": 4376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2994785484308977,
+      "learning_rate": 8.447063425147891e-06,
+      "loss": 0.8307,
+      "step": 4377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9410804937430557,
+      "learning_rate": 8.446312552118068e-06,
+      "loss": 0.6989,
+      "step": 4378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9451403049390426,
+      "learning_rate": 8.44556153099252e-06,
+      "loss": 0.7625,
+      "step": 4379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6932015634396975,
+      "learning_rate": 8.44481036180352e-06,
+      "loss": 0.843,
+      "step": 4380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0088550116596084,
+      "learning_rate": 8.444059044583344e-06,
+      "loss": 0.9495,
+      "step": 4381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1024547941270146,
+      "learning_rate": 8.443307579364282e-06,
+      "loss": 0.8071,
+      "step": 4382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.448347184041665,
+      "learning_rate": 8.442555966178626e-06,
+      "loss": 0.8013,
+      "step": 4383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9583709268177245,
+      "learning_rate": 8.441804205058672e-06,
+      "loss": 0.8668,
+      "step": 4384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.952369506699993,
+      "learning_rate": 8.441052296036724e-06,
+      "loss": 0.9605,
+      "step": 4385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8726051984222143,
+      "learning_rate": 8.440300239145098e-06,
+      "loss": 0.9837,
+      "step": 4386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.701529180991134,
+      "learning_rate": 8.439548034416108e-06,
+      "loss": 0.9192,
+      "step": 4387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5496098183848463,
+      "learning_rate": 8.43879568188208e-06,
+      "loss": 1.0779,
+      "step": 4388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6432412959353937,
+      "learning_rate": 8.43804318157534e-06,
+      "loss": 1.1613,
+      "step": 4389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.075879900628084,
+      "learning_rate": 8.437290533528231e-06,
+      "loss": 0.9445,
+      "step": 4390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.973568891982425,
+      "learning_rate": 8.436537737773089e-06,
+      "loss": 0.6679,
+      "step": 4391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9769578678444961,
+      "learning_rate": 8.43578479434227e-06,
+      "loss": 0.8633,
+      "step": 4392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0486071249693925,
+      "learning_rate": 8.435031703268123e-06,
+      "loss": 0.7367,
+      "step": 4393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.202725460053181,
+      "learning_rate": 8.434278464583018e-06,
+      "loss": 0.8266,
+      "step": 4394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.527636221520004,
+      "learning_rate": 8.433525078319316e-06,
+      "loss": 0.8757,
+      "step": 4395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3057149876251093,
+      "learning_rate": 8.432771544509395e-06,
+      "loss": 0.9061,
+      "step": 4396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.053600690434964,
+      "learning_rate": 8.432017863185635e-06,
+      "loss": 0.9019,
+      "step": 4397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4067057367306153,
+      "learning_rate": 8.431264034380424e-06,
+      "loss": 0.8486,
+      "step": 4398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0842704892961965,
+      "learning_rate": 8.430510058126156e-06,
+      "loss": 0.7755,
+      "step": 4399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9100651038131968,
+      "learning_rate": 8.42975593445523e-06,
+      "loss": 0.6106,
+      "step": 4400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3563491418934297,
+      "learning_rate": 8.429001663400054e-06,
+      "loss": 0.8201,
+      "step": 4401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.351963023821807,
+      "learning_rate": 8.428247244993038e-06,
+      "loss": 0.8091,
+      "step": 4402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.896079342360544,
+      "learning_rate": 8.427492679266605e-06,
+      "loss": 0.7309,
+      "step": 4403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2651614060213285,
+      "learning_rate": 8.426737966253176e-06,
+      "loss": 0.8036,
+      "step": 4404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7781032880943983,
+      "learning_rate": 8.425983105985188e-06,
+      "loss": 0.8034,
+      "step": 4405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7812642649210868,
+      "learning_rate": 8.425228098495073e-06,
+      "loss": 0.8664,
+      "step": 4406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9376853518110355,
+      "learning_rate": 8.424472943815278e-06,
+      "loss": 0.9105,
+      "step": 4407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.029863836173955,
+      "learning_rate": 8.423717641978255e-06,
+      "loss": 0.7846,
+      "step": 4408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7293575941665322,
+      "learning_rate": 8.422962193016459e-06,
+      "loss": 0.9031,
+      "step": 4409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9346585419891043,
+      "learning_rate": 8.422206596962357e-06,
+      "loss": 0.8561,
+      "step": 4410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.834739784301807,
+      "learning_rate": 8.421450853848414e-06,
+      "loss": 0.8086,
+      "step": 4411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.795213310338317,
+      "learning_rate": 8.420694963707108e-06,
+      "loss": 0.8135,
+      "step": 4412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.099421183107295,
+      "learning_rate": 8.419938926570921e-06,
+      "loss": 0.7824,
+      "step": 4413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9995408974972264,
+      "learning_rate": 8.41918274247234e-06,
+      "loss": 0.7755,
+      "step": 4414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7091512621121117,
+      "learning_rate": 8.418426411443864e-06,
+      "loss": 0.9316,
+      "step": 4415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3318243200979154,
+      "learning_rate": 8.41766993351799e-06,
+      "loss": 1.0156,
+      "step": 4416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.25404820219589,
+      "learning_rate": 8.416913308727229e-06,
+      "loss": 0.9151,
+      "step": 4417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7836695522765216,
+      "learning_rate": 8.416156537104092e-06,
+      "loss": 0.8596,
+      "step": 4418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9104058509338357,
+      "learning_rate": 8.415399618681101e-06,
+      "loss": 0.6798,
+      "step": 4419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3765779529612563,
+      "learning_rate": 8.414642553490783e-06,
+      "loss": 0.9048,
+      "step": 4420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.232315411152863,
+      "learning_rate": 8.413885341565668e-06,
+      "loss": 0.7112,
+      "step": 4421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.677171579170749,
+      "learning_rate": 8.413127982938299e-06,
+      "loss": 1.124,
+      "step": 4422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1610458286989127,
+      "learning_rate": 8.412370477641215e-06,
+      "loss": 0.7704,
+      "step": 4423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9122995953990258,
+      "learning_rate": 8.411612825706976e-06,
+      "loss": 0.653,
+      "step": 4424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8875300181667194,
+      "learning_rate": 8.410855027168134e-06,
+      "loss": 0.9202,
+      "step": 4425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7753179497196854,
+      "learning_rate": 8.410097082057256e-06,
+      "loss": 0.783,
+      "step": 4426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.919962758705532,
+      "learning_rate": 8.40933899040691e-06,
+      "loss": 0.8522,
+      "step": 4427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9765486072052456,
+      "learning_rate": 8.408580752249676e-06,
+      "loss": 0.8977,
+      "step": 4428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.337850569670209,
+      "learning_rate": 8.407822367618135e-06,
+      "loss": 0.7437,
+      "step": 4429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7902039260576303,
+      "learning_rate": 8.407063836544877e-06,
+      "loss": 0.8016,
+      "step": 4430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6699887138316194,
+      "learning_rate": 8.4063051590625e-06,
+      "loss": 0.8995,
+      "step": 4431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1216287967194836,
+      "learning_rate": 8.405546335203602e-06,
+      "loss": 0.8842,
+      "step": 4432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0498601844346207,
+      "learning_rate": 8.404787365000796e-06,
+      "loss": 0.865,
+      "step": 4433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7600923226871739,
+      "learning_rate": 8.40402824848669e-06,
+      "loss": 0.6976,
+      "step": 4434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9785968214489935,
+      "learning_rate": 8.403268985693912e-06,
+      "loss": 0.8494,
+      "step": 4435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8545107331235737,
+      "learning_rate": 8.402509576655085e-06,
+      "loss": 0.98,
+      "step": 4436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9285042431553847,
+      "learning_rate": 8.401750021402843e-06,
+      "loss": 0.7712,
+      "step": 4437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7998668333893377,
+      "learning_rate": 8.400990319969829e-06,
+      "loss": 0.8164,
+      "step": 4438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0720517339883755,
+      "learning_rate": 8.400230472388684e-06,
+      "loss": 0.8512,
+      "step": 4439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7899874774214577,
+      "learning_rate": 8.399470478692064e-06,
+      "loss": 0.9341,
+      "step": 4440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0435206050873278,
+      "learning_rate": 8.398710338912626e-06,
+      "loss": 0.6526,
+      "step": 4441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0349137759747598,
+      "learning_rate": 8.397950053083036e-06,
+      "loss": 0.6677,
+      "step": 4442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0440819325892268,
+      "learning_rate": 8.397189621235964e-06,
+      "loss": 0.7109,
+      "step": 4443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.433644251127517,
+      "learning_rate": 8.396429043404088e-06,
+      "loss": 0.7216,
+      "step": 4444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1444936201774212,
+      "learning_rate": 8.395668319620092e-06,
+      "loss": 0.7438,
+      "step": 4445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2202613111475036,
+      "learning_rate": 8.394907449916667e-06,
+      "loss": 0.7915,
+      "step": 4446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3981567841795197,
+      "learning_rate": 8.394146434326507e-06,
+      "loss": 0.9266,
+      "step": 4447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.520025211662831,
+      "learning_rate": 8.393385272882315e-06,
+      "loss": 0.7757,
+      "step": 4448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.054611691581527,
+      "learning_rate": 8.3926239656168e-06,
+      "loss": 0.9518,
+      "step": 4449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5169343054197513,
+      "learning_rate": 8.391862512562679e-06,
+      "loss": 0.9542,
+      "step": 4450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.389897402721558,
+      "learning_rate": 8.39110091375267e-06,
+      "loss": 0.9083,
+      "step": 4451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9791196142135488,
+      "learning_rate": 8.390339169219504e-06,
+      "loss": 0.8186,
+      "step": 4452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4340223101502976,
+      "learning_rate": 8.389577278995913e-06,
+      "loss": 0.8105,
+      "step": 4453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6249715636251483,
+      "learning_rate": 8.388815243114637e-06,
+      "loss": 0.8608,
+      "step": 4454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0535702107061506,
+      "learning_rate": 8.388053061608421e-06,
+      "loss": 0.6613,
+      "step": 4455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1236851332926379,
+      "learning_rate": 8.387290734510022e-06,
+      "loss": 0.6301,
+      "step": 4456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2339626186107857,
+      "learning_rate": 8.386528261852196e-06,
+      "loss": 0.7284,
+      "step": 4457
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8689720678327846,
+      "learning_rate": 8.385765643667707e-06,
+      "loss": 0.8299,
+      "step": 4458
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.091887485390256,
+      "learning_rate": 8.385002879989328e-06,
+      "loss": 0.6942,
+      "step": 4459
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9792812754771607,
+      "learning_rate": 8.384239970849837e-06,
+      "loss": 0.7845,
+      "step": 4460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.979595646623584,
+      "learning_rate": 8.383476916282015e-06,
+      "loss": 0.9919,
+      "step": 4461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.083196666920942,
+      "learning_rate": 8.382713716318658e-06,
+      "loss": 0.7057,
+      "step": 4462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0445060215350623,
+      "learning_rate": 8.381950370992557e-06,
+      "loss": 0.9045,
+      "step": 4463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1074017724728145,
+      "learning_rate": 8.381186880336518e-06,
+      "loss": 0.7652,
+      "step": 4464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5035010325928275,
+      "learning_rate": 8.380423244383347e-06,
+      "loss": 0.7525,
+      "step": 4465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.094556415948721,
+      "learning_rate": 8.37965946316586e-06,
+      "loss": 0.8509,
+      "step": 4466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4298296520730212,
+      "learning_rate": 8.378895536716882e-06,
+      "loss": 0.7949,
+      "step": 4467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.056491623314358,
+      "learning_rate": 8.378131465069235e-06,
+      "loss": 0.8966,
+      "step": 4468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9884144146810485,
+      "learning_rate": 8.377367248255757e-06,
+      "loss": 0.8741,
+      "step": 4469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.065628166936878,
+      "learning_rate": 8.376602886309285e-06,
+      "loss": 0.9271,
+      "step": 4470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9113311398100998,
+      "learning_rate": 8.375838379262667e-06,
+      "loss": 0.7389,
+      "step": 4471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.143881599702912,
+      "learning_rate": 8.375073727148756e-06,
+      "loss": 0.7623,
+      "step": 4472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1603938757733525,
+      "learning_rate": 8.37430893000041e-06,
+      "loss": 0.6959,
+      "step": 4473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.121599698678351,
+      "learning_rate": 8.373543987850494e-06,
+      "loss": 0.9671,
+      "step": 4474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8116112301859655,
+      "learning_rate": 8.372778900731882e-06,
+      "loss": 0.7451,
+      "step": 4475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5531995167810064,
+      "learning_rate": 8.372013668677446e-06,
+      "loss": 0.7029,
+      "step": 4476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7793672020202032,
+      "learning_rate": 8.371248291720073e-06,
+      "loss": 0.8972,
+      "step": 4477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8861892874245982,
+      "learning_rate": 8.370482769892654e-06,
+      "loss": 0.8162,
+      "step": 4478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0477293077644787,
+      "learning_rate": 8.369717103228084e-06,
+      "loss": 0.9033,
+      "step": 4479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3392853290153632,
+      "learning_rate": 8.368951291759264e-06,
+      "loss": 0.7722,
+      "step": 4480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0095002526181425,
+      "learning_rate": 8.368185335519106e-06,
+      "loss": 0.8308,
+      "step": 4481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.188495180534236,
+      "learning_rate": 8.367419234540522e-06,
+      "loss": 0.6845,
+      "step": 4482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.604970595532673,
+      "learning_rate": 8.366652988856432e-06,
+      "loss": 0.7677,
+      "step": 4483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4395043644211425,
+      "learning_rate": 8.365886598499766e-06,
+      "loss": 0.823,
+      "step": 4484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.414187438145303,
+      "learning_rate": 8.365120063503458e-06,
+      "loss": 0.7103,
+      "step": 4485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.164118374593245,
+      "learning_rate": 8.364353383900445e-06,
+      "loss": 0.8384,
+      "step": 4486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0789638295512827,
+      "learning_rate": 8.363586559723675e-06,
+      "loss": 0.9041,
+      "step": 4487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3749871504574607,
+      "learning_rate": 8.3628195910061e-06,
+      "loss": 0.9121,
+      "step": 4488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.330882956102571,
+      "learning_rate": 8.362052477780677e-06,
+      "loss": 1.0438,
+      "step": 4489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8536887627919543,
+      "learning_rate": 8.36128522008037e-06,
+      "loss": 0.6753,
+      "step": 4490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4101590189310422,
+      "learning_rate": 8.360517817938154e-06,
+      "loss": 0.7856,
+      "step": 4491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.202359141415349,
+      "learning_rate": 8.359750271386999e-06,
+      "loss": 0.9014,
+      "step": 4492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.401711470153252,
+      "learning_rate": 8.358982580459896e-06,
+      "loss": 0.8576,
+      "step": 4493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.739716770297364,
+      "learning_rate": 8.35821474518983e-06,
+      "loss": 0.8278,
+      "step": 4494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5092072176268703,
+      "learning_rate": 8.357446765609796e-06,
+      "loss": 0.9004,
+      "step": 4495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.296764330274167,
+      "learning_rate": 8.356678641752797e-06,
+      "loss": 0.8806,
+      "step": 4496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0469202398872244,
+      "learning_rate": 8.355910373651844e-06,
+      "loss": 1.0182,
+      "step": 4497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.08820043545513,
+      "learning_rate": 8.355141961339945e-06,
+      "loss": 0.8874,
+      "step": 4498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7486737730550268,
+      "learning_rate": 8.354373404850124e-06,
+      "loss": 0.7261,
+      "step": 4499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9918649158794057,
+      "learning_rate": 8.353604704215408e-06,
+      "loss": 0.8141,
+      "step": 4500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0651012581822452,
+      "learning_rate": 8.352835859468829e-06,
+      "loss": 0.63,
+      "step": 4501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1902659150344292,
+      "learning_rate": 8.352066870643424e-06,
+      "loss": 0.9122,
+      "step": 4502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.135873518425155,
+      "learning_rate": 8.351297737772244e-06,
+      "loss": 0.9385,
+      "step": 4503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2052319444468145,
+      "learning_rate": 8.350528460888334e-06,
+      "loss": 0.9919,
+      "step": 4504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.682731235766537,
+      "learning_rate": 8.349759040024753e-06,
+      "loss": 0.8861,
+      "step": 4505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4600759756685386,
+      "learning_rate": 8.348989475214568e-06,
+      "loss": 0.9567,
+      "step": 4506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9428071813711103,
+      "learning_rate": 8.348219766490845e-06,
+      "loss": 0.9608,
+      "step": 4507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4066143581831603,
+      "learning_rate": 8.347449913886662e-06,
+      "loss": 0.832,
+      "step": 4508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9399628773803232,
+      "learning_rate": 8.346679917435104e-06,
+      "loss": 0.8196,
+      "step": 4509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.271238688504737,
+      "learning_rate": 8.345909777169252e-06,
+      "loss": 1.008,
+      "step": 4510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.493826747248264,
+      "learning_rate": 8.345139493122208e-06,
+      "loss": 0.9481,
+      "step": 4511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.161459676243732,
+      "learning_rate": 8.34436906532707e-06,
+      "loss": 0.6864,
+      "step": 4512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2331271908264556,
+      "learning_rate": 8.343598493816944e-06,
+      "loss": 0.6934,
+      "step": 4513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1749866964002793,
+      "learning_rate": 8.342827778624943e-06,
+      "loss": 0.6477,
+      "step": 4514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.327732847107845,
+      "learning_rate": 8.34205691978419e-06,
+      "loss": 1.0601,
+      "step": 4515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.126292934512357,
+      "learning_rate": 8.341285917327807e-06,
+      "loss": 0.8559,
+      "step": 4516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3942152231512517,
+      "learning_rate": 8.340514771288926e-06,
+      "loss": 0.9913,
+      "step": 4517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.82852460395202,
+      "learning_rate": 8.339743481700685e-06,
+      "loss": 0.7318,
+      "step": 4518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8111096889843967,
+      "learning_rate": 8.33897204859623e-06,
+      "loss": 0.9445,
+      "step": 4519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8274536346532473,
+      "learning_rate": 8.338200472008708e-06,
+      "loss": 0.8698,
+      "step": 4520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4695980574824654,
+      "learning_rate": 8.337428751971279e-06,
+      "loss": 0.7485,
+      "step": 4521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.525446089021819,
+      "learning_rate": 8.336656888517103e-06,
+      "loss": 0.8268,
+      "step": 4522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1327933973247695,
+      "learning_rate": 8.33588488167935e-06,
+      "loss": 0.8628,
+      "step": 4523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.910995039120968,
+      "learning_rate": 8.335112731491192e-06,
+      "loss": 0.9164,
+      "step": 4524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2760170963002584,
+      "learning_rate": 8.334340437985814e-06,
+      "loss": 0.8066,
+      "step": 4525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.906101783208721,
+      "learning_rate": 8.333568001196402e-06,
+      "loss": 0.7816,
+      "step": 4526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2940420969894935,
+      "learning_rate": 8.332795421156147e-06,
+      "loss": 0.8867,
+      "step": 4527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3162465003822637,
+      "learning_rate": 8.332022697898253e-06,
+      "loss": 0.9841,
+      "step": 4528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1115585993370938,
+      "learning_rate": 8.331249831455921e-06,
+      "loss": 0.9929,
+      "step": 4529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.318318014010111,
+      "learning_rate": 8.330476821862366e-06,
+      "loss": 0.9087,
+      "step": 4530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.243315109810106,
+      "learning_rate": 8.329703669150801e-06,
+      "loss": 0.9153,
+      "step": 4531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6885829656089169,
+      "learning_rate": 8.328930373354457e-06,
+      "loss": 0.7662,
+      "step": 4532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3947953478408865,
+      "learning_rate": 8.328156934506559e-06,
+      "loss": 0.8751,
+      "step": 4533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.400320223219075,
+      "learning_rate": 8.327383352640347e-06,
+      "loss": 0.9443,
+      "step": 4534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3004594368407725,
+      "learning_rate": 8.32660962778906e-06,
+      "loss": 0.7115,
+      "step": 4535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.037642470714612,
+      "learning_rate": 8.325835759985951e-06,
+      "loss": 0.6949,
+      "step": 4536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.714055728328404,
+      "learning_rate": 8.32506174926427e-06,
+      "loss": 0.9845,
+      "step": 4537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9805546499295237,
+      "learning_rate": 8.324287595657284e-06,
+      "loss": 0.8659,
+      "step": 4538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.518013515925005,
+      "learning_rate": 8.323513299198252e-06,
+      "loss": 0.7926,
+      "step": 4539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.571740183457717,
+      "learning_rate": 8.322738859920453e-06,
+      "loss": 0.7901,
+      "step": 4540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9998044219780737,
+      "learning_rate": 8.321964277857167e-06,
+      "loss": 0.9318,
+      "step": 4541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.002284686834264,
+      "learning_rate": 8.321189553041675e-06,
+      "loss": 0.8827,
+      "step": 4542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.09283659631095,
+      "learning_rate": 8.320414685507272e-06,
+      "loss": 0.8307,
+      "step": 4543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1853508211590933,
+      "learning_rate": 8.319639675287255e-06,
+      "loss": 0.846,
+      "step": 4544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.445545496308092,
+      "learning_rate": 8.318864522414928e-06,
+      "loss": 0.8742,
+      "step": 4545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3346524950917367,
+      "learning_rate": 8.318089226923602e-06,
+      "loss": 0.9292,
+      "step": 4546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3647515435128486,
+      "learning_rate": 8.317313788846591e-06,
+      "loss": 0.8915,
+      "step": 4547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.978678561039664,
+      "learning_rate": 8.31653820821722e-06,
+      "loss": 0.6186,
+      "step": 4548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.94532730968498,
+      "learning_rate": 8.315762485068815e-06,
+      "loss": 1.072,
+      "step": 4549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.219960896235725,
+      "learning_rate": 8.31498661943471e-06,
+      "loss": 0.7898,
+      "step": 4550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1123453844282358,
+      "learning_rate": 8.314210611348249e-06,
+      "loss": 0.7099,
+      "step": 4551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9783400605601704,
+      "learning_rate": 8.313434460842775e-06,
+      "loss": 0.9289,
+      "step": 4552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1570617360385707,
+      "learning_rate": 8.312658167951644e-06,
+      "loss": 0.8253,
+      "step": 4553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.096051863645469,
+      "learning_rate": 8.311881732708213e-06,
+      "loss": 0.807,
+      "step": 4554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3486222776822734,
+      "learning_rate": 8.311105155145849e-06,
+      "loss": 0.8981,
+      "step": 4555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.174195893737039,
+      "learning_rate": 8.310328435297923e-06,
+      "loss": 0.8419,
+      "step": 4556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.030857076553585,
+      "learning_rate": 8.309551573197809e-06,
+      "loss": 0.8392,
+      "step": 4557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.531069260334235,
+      "learning_rate": 8.308774568878896e-06,
+      "loss": 0.6731,
+      "step": 4558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0237079132852185,
+      "learning_rate": 8.307997422374569e-06,
+      "loss": 0.8034,
+      "step": 4559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9561531509421002,
+      "learning_rate": 8.307220133718225e-06,
+      "loss": 0.8828,
+      "step": 4560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.315436968403611,
+      "learning_rate": 8.306442702943268e-06,
+      "loss": 0.7616,
+      "step": 4561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.307220651914893,
+      "learning_rate": 8.305665130083102e-06,
+      "loss": 0.8394,
+      "step": 4562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.316606324476112,
+      "learning_rate": 8.304887415171146e-06,
+      "loss": 0.6467,
+      "step": 4563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9266614012597738,
+      "learning_rate": 8.304109558240817e-06,
+      "loss": 0.8396,
+      "step": 4564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2056932180049427,
+      "learning_rate": 8.30333155932554e-06,
+      "loss": 0.9107,
+      "step": 4565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2276922999810782,
+      "learning_rate": 8.302553418458749e-06,
+      "loss": 0.7731,
+      "step": 4566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2734414712367426,
+      "learning_rate": 8.301775135673884e-06,
+      "loss": 0.6908,
+      "step": 4567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2377617629289546,
+      "learning_rate": 8.300996711004387e-06,
+      "loss": 0.8264,
+      "step": 4568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3055524283172604,
+      "learning_rate": 8.300218144483709e-06,
+      "loss": 0.9496,
+      "step": 4569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.079000949343692,
+      "learning_rate": 8.299439436145309e-06,
+      "loss": 0.7994,
+      "step": 4570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.375103087182569,
+      "learning_rate": 8.298660586022646e-06,
+      "loss": 0.8512,
+      "step": 4571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8340304206835136,
+      "learning_rate": 8.297881594149193e-06,
+      "loss": 0.8041,
+      "step": 4572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.975904913249608,
+      "learning_rate": 8.297102460558421e-06,
+      "loss": 0.6983,
+      "step": 4573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1012892167103514,
+      "learning_rate": 8.296323185283816e-06,
+      "loss": 0.7383,
+      "step": 4574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.161958615360072,
+      "learning_rate": 8.29554376835886e-06,
+      "loss": 0.8223,
+      "step": 4575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.382808683778344,
+      "learning_rate": 8.29476420981705e-06,
+      "loss": 1.0881,
+      "step": 4576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2346527206459788,
+      "learning_rate": 8.293984509691885e-06,
+      "loss": 0.7416,
+      "step": 4577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.239517530791074,
+      "learning_rate": 8.293204668016867e-06,
+      "loss": 0.872,
+      "step": 4578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0627448094181946,
+      "learning_rate": 8.292424684825514e-06,
+      "loss": 0.8762,
+      "step": 4579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1396159852373033,
+      "learning_rate": 8.291644560151335e-06,
+      "loss": 0.9079,
+      "step": 4580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8470551616716613,
+      "learning_rate": 8.29086429402786e-06,
+      "loss": 0.864,
+      "step": 4581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.343073101367461,
+      "learning_rate": 8.290083886488618e-06,
+      "loss": 0.7713,
+      "step": 4582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6188525383971668,
+      "learning_rate": 8.289303337567145e-06,
+      "loss": 0.7385,
+      "step": 4583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.333903941675175,
+      "learning_rate": 8.28852264729698e-06,
+      "loss": 0.8221,
+      "step": 4584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0879189429535034,
+      "learning_rate": 8.287741815711674e-06,
+      "loss": 0.6841,
+      "step": 4585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.258480348306817,
+      "learning_rate": 8.286960842844779e-06,
+      "loss": 0.8432,
+      "step": 4586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7606798662821315,
+      "learning_rate": 8.286179728729856e-06,
+      "loss": 0.7309,
+      "step": 4587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1214535874622302,
+      "learning_rate": 8.285398473400471e-06,
+      "loss": 0.9974,
+      "step": 4588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8690776414082493,
+      "learning_rate": 8.284617076890199e-06,
+      "loss": 1.0121,
+      "step": 4589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.268498298139327,
+      "learning_rate": 8.283835539232614e-06,
+      "loss": 0.7585,
+      "step": 4590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.160471493461634,
+      "learning_rate": 8.283053860461303e-06,
+      "loss": 0.8419,
+      "step": 4591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6727456544630606,
+      "learning_rate": 8.282272040609855e-06,
+      "loss": 0.6858,
+      "step": 4592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.932139982976684,
+      "learning_rate": 8.281490079711866e-06,
+      "loss": 0.9586,
+      "step": 4593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2563442321092766,
+      "learning_rate": 8.280707977800944e-06,
+      "loss": 0.9058,
+      "step": 4594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9303174784507506,
+      "learning_rate": 8.279925734910691e-06,
+      "loss": 0.8579,
+      "step": 4595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4993414946506016,
+      "learning_rate": 8.279143351074726e-06,
+      "loss": 0.6692,
+      "step": 4596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.063626436518251,
+      "learning_rate": 8.278360826326669e-06,
+      "loss": 0.5827,
+      "step": 4597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.676242197429218,
+      "learning_rate": 8.277578160700143e-06,
+      "loss": 0.8348,
+      "step": 4598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.259354454984838,
+      "learning_rate": 8.276795354228785e-06,
+      "loss": 0.8799,
+      "step": 4599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.059952257267768,
+      "learning_rate": 8.276012406946232e-06,
+      "loss": 0.7144,
+      "step": 4600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4732611487344722,
+      "learning_rate": 8.275229318886132e-06,
+      "loss": 0.7466,
+      "step": 4601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.99966785620237,
+      "learning_rate": 8.274446090082133e-06,
+      "loss": 0.8814,
+      "step": 4602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.250239969379915,
+      "learning_rate": 8.273662720567892e-06,
+      "loss": 0.9918,
+      "step": 4603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8526144787310475,
+      "learning_rate": 8.272879210377074e-06,
+      "loss": 0.8588,
+      "step": 4604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2941646902660717,
+      "learning_rate": 8.272095559543346e-06,
+      "loss": 0.8318,
+      "step": 4605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.538028817731692,
+      "learning_rate": 8.271311768100386e-06,
+      "loss": 0.7982,
+      "step": 4606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.254046415184496,
+      "learning_rate": 8.270527836081872e-06,
+      "loss": 0.932,
+      "step": 4607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9198094225826612,
+      "learning_rate": 8.269743763521495e-06,
+      "loss": 0.8329,
+      "step": 4608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2024886437168976,
+      "learning_rate": 8.268959550452946e-06,
+      "loss": 0.7669,
+      "step": 4609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.832268285445359,
+      "learning_rate": 8.268175196909924e-06,
+      "loss": 0.8822,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7776361490339756,
+      "learning_rate": 8.267390702926135e-06,
+      "loss": 0.8073,
+      "step": 4611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1463797274143963,
+      "learning_rate": 8.26660606853529e-06,
+      "loss": 0.8742,
+      "step": 4612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.270340958032394,
+      "learning_rate": 8.26582129377111e-06,
+      "loss": 0.6863,
+      "step": 4613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.681982807597737,
+      "learning_rate": 8.265036378667312e-06,
+      "loss": 0.8869,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0888530066807434,
+      "learning_rate": 8.264251323257633e-06,
+      "loss": 0.8725,
+      "step": 4615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0101591262728817,
+      "learning_rate": 8.263466127575801e-06,
+      "loss": 0.9751,
+      "step": 4616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3299443929664276,
+      "learning_rate": 8.262680791655565e-06,
+      "loss": 0.7186,
+      "step": 4617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0028624004342137,
+      "learning_rate": 8.261895315530667e-06,
+      "loss": 0.8758,
+      "step": 4618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.334774662170465,
+      "learning_rate": 8.261109699234862e-06,
+      "loss": 0.6135,
+      "step": 4619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.625996574374654,
+      "learning_rate": 8.26032394280191e-06,
+      "loss": 0.7683,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.130015979357719,
+      "learning_rate": 8.259538046265578e-06,
+      "loss": 0.9583,
+      "step": 4621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9987185509720486,
+      "learning_rate": 8.258752009659638e-06,
+      "loss": 0.8214,
+      "step": 4622
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1192161353315337,
+      "learning_rate": 8.257965833017864e-06,
+      "loss": 0.7998,
+      "step": 4623
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7745389617635907,
+      "learning_rate": 8.257179516374045e-06,
+      "loss": 0.8798,
+      "step": 4624
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.944554695065326,
+      "learning_rate": 8.256393059761966e-06,
+      "loss": 0.9971,
+      "step": 4625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9444418950576137,
+      "learning_rate": 8.255606463215426e-06,
+      "loss": 0.9544,
+      "step": 4626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.647763134934171,
+      "learning_rate": 8.254819726768224e-06,
+      "loss": 0.8044,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.130133761061346,
+      "learning_rate": 8.254032850454173e-06,
+      "loss": 0.9533,
+      "step": 4628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.034368795786201,
+      "learning_rate": 8.253245834307079e-06,
+      "loss": 0.7543,
+      "step": 4629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1214537863666938,
+      "learning_rate": 8.252458678360769e-06,
+      "loss": 0.7292,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8728433654108625,
+      "learning_rate": 8.251671382649065e-06,
+      "loss": 0.9062,
+      "step": 4631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0712451794804048,
+      "learning_rate": 8.250883947205799e-06,
+      "loss": 0.7192,
+      "step": 4632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1908901481797138,
+      "learning_rate": 8.25009637206481e-06,
+      "loss": 0.6589,
+      "step": 4633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1466566514094927,
+      "learning_rate": 8.249308657259943e-06,
+      "loss": 0.8293,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7861877713869454,
+      "learning_rate": 8.248520802825046e-06,
+      "loss": 0.733,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3077513385369377,
+      "learning_rate": 8.247732808793975e-06,
+      "loss": 0.837,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0879853383436107,
+      "learning_rate": 8.246944675200592e-06,
+      "loss": 0.9266,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.596854578010218,
+      "learning_rate": 8.246156402078765e-06,
+      "loss": 0.7931,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.345147519457627,
+      "learning_rate": 8.245367989462368e-06,
+      "loss": 0.8098,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.835292836991519,
+      "learning_rate": 8.244579437385281e-06,
+      "loss": 0.9576,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.188540979035097,
+      "learning_rate": 8.243790745881389e-06,
+      "loss": 0.9093,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.387450297203892,
+      "learning_rate": 8.243001914984583e-06,
+      "loss": 0.7906,
+      "step": 4642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2290106954123736,
+      "learning_rate": 8.242212944728762e-06,
+      "loss": 0.6597,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.018136376424666,
+      "learning_rate": 8.241423835147833e-06,
+      "loss": 0.9117,
+      "step": 4644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.035362662164121,
+      "learning_rate": 8.240634586275701e-06,
+      "loss": 0.8782,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8774608738148428,
+      "learning_rate": 8.239845198146284e-06,
+      "loss": 0.7061,
+      "step": 4646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.658507196474391,
+      "learning_rate": 8.239055670793503e-06,
+      "loss": 0.8383,
+      "step": 4647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.572241259916796,
+      "learning_rate": 8.238266004251284e-06,
+      "loss": 0.8927,
+      "step": 4648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.283073566789163,
+      "learning_rate": 8.237476198553567e-06,
+      "loss": 0.8704,
+      "step": 4649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3772953417503504,
+      "learning_rate": 8.236686253734285e-06,
+      "loss": 0.9289,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.960137573334007,
+      "learning_rate": 8.235896169827386e-06,
+      "loss": 0.6628,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.423674449150862,
+      "learning_rate": 8.235105946866823e-06,
+      "loss": 0.7543,
+      "step": 4652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1944940442044325,
+      "learning_rate": 8.234315584886551e-06,
+      "loss": 0.7995,
+      "step": 4653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8154899393194976,
+      "learning_rate": 8.233525083920536e-06,
+      "loss": 0.8876,
+      "step": 4654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9344844223544144,
+      "learning_rate": 8.232734444002748e-06,
+      "loss": 0.8045,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4533836962223616,
+      "learning_rate": 8.23194366516716e-06,
+      "loss": 0.8291,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5258695122378905,
+      "learning_rate": 8.231152747447753e-06,
+      "loss": 0.9136,
+      "step": 4657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.411765391306894,
+      "learning_rate": 8.230361690878519e-06,
+      "loss": 0.9321,
+      "step": 4658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9863562153949028,
+      "learning_rate": 8.229570495493447e-06,
+      "loss": 0.9732,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.116227900173955,
+      "learning_rate": 8.22877916132654e-06,
+      "loss": 0.7269,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1144988498707673,
+      "learning_rate": 8.2279876884118e-06,
+      "loss": 0.8841,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7540827499164187,
+      "learning_rate": 8.22719607678324e-06,
+      "loss": 0.7216,
+      "step": 4662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.222880959760755,
+      "learning_rate": 8.226404326474878e-06,
+      "loss": 0.8055,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0556524750465663,
+      "learning_rate": 8.225612437520736e-06,
+      "loss": 0.9423,
+      "step": 4664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0259534325996027,
+      "learning_rate": 8.224820409954844e-06,
+      "loss": 0.8101,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9785129435876794,
+      "learning_rate": 8.224028243811237e-06,
+      "loss": 1.0059,
+      "step": 4666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4576547719581594,
+      "learning_rate": 8.223235939123955e-06,
+      "loss": 0.894,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5778876780390174,
+      "learning_rate": 8.222443495927049e-06,
+      "loss": 0.9849,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2893561088144923,
+      "learning_rate": 8.221650914254566e-06,
+      "loss": 0.6734,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.178472490394523,
+      "learning_rate": 8.220858194140568e-06,
+      "loss": 0.9316,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.766990695529529,
+      "learning_rate": 8.220065335619123e-06,
+      "loss": 0.8979,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8006511733036512,
+      "learning_rate": 8.219272338724297e-06,
+      "loss": 0.9203,
+      "step": 4672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2750366021235775,
+      "learning_rate": 8.21847920349017e-06,
+      "loss": 0.6403,
+      "step": 4673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0687915418039515,
+      "learning_rate": 8.217685929950823e-06,
+      "loss": 0.8095,
+      "step": 4674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.207648324857878,
+      "learning_rate": 8.216892518140346e-06,
+      "loss": 0.7203,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.186430460734418,
+      "learning_rate": 8.216098968092833e-06,
+      "loss": 0.9248,
+      "step": 4676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.182112612669978,
+      "learning_rate": 8.215305279842385e-06,
+      "loss": 0.8789,
+      "step": 4677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5692917443645706,
+      "learning_rate": 8.214511453423107e-06,
+      "loss": 0.8159,
+      "step": 4678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8149752735783387,
+      "learning_rate": 8.213717488869113e-06,
+      "loss": 0.8711,
+      "step": 4679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0849208490383755,
+      "learning_rate": 8.212923386214522e-06,
+      "loss": 0.6076,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1778753853328783,
+      "learning_rate": 8.212129145493457e-06,
+      "loss": 0.853,
+      "step": 4681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.085000683254722,
+      "learning_rate": 8.211334766740048e-06,
+      "loss": 0.8705,
+      "step": 4682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1995231036466545,
+      "learning_rate": 8.210540249988435e-06,
+      "loss": 0.9813,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7643787175533276,
+      "learning_rate": 8.209745595272755e-06,
+      "loss": 0.9396,
+      "step": 4684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2194985852899545,
+      "learning_rate": 8.208950802627157e-06,
+      "loss": 0.8565,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5160278768456843,
+      "learning_rate": 8.208155872085799e-06,
+      "loss": 0.9293,
+      "step": 4686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0711583088177794,
+      "learning_rate": 8.207360803682837e-06,
+      "loss": 0.6358,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2363166445212848,
+      "learning_rate": 8.206565597452438e-06,
+      "loss": 1.1058,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.260108666952518,
+      "learning_rate": 8.205770253428775e-06,
+      "loss": 0.8431,
+      "step": 4689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1282054846464145,
+      "learning_rate": 8.204974771646023e-06,
+      "loss": 0.9188,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.173288931534606,
+      "learning_rate": 8.204179152138367e-06,
+      "loss": 0.9771,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.469732138220622,
+      "learning_rate": 8.203383394939998e-06,
+      "loss": 0.9437,
+      "step": 4692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5267845154234534,
+      "learning_rate": 8.20258750008511e-06,
+      "loss": 0.8135,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.163255677572875,
+      "learning_rate": 8.201791467607905e-06,
+      "loss": 0.7415,
+      "step": 4694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3655948201857777,
+      "learning_rate": 8.200995297542589e-06,
+      "loss": 1.0428,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0699562915235994,
+      "learning_rate": 8.200198989923376e-06,
+      "loss": 0.7277,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.722779822106159,
+      "learning_rate": 8.199402544784485e-06,
+      "loss": 0.8706,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8725904722684852,
+      "learning_rate": 8.198605962160143e-06,
+      "loss": 0.7252,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.5975779671006425,
+      "learning_rate": 8.197809242084575e-06,
+      "loss": 0.6852,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.105650377199651,
+      "learning_rate": 8.197012384592027e-06,
+      "loss": 0.7305,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8187698552639073,
+      "learning_rate": 8.196215389716732e-06,
+      "loss": 0.8185,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.773766868012755,
+      "learning_rate": 8.195418257492948e-06,
+      "loss": 0.8182,
+      "step": 4702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6579348184947866,
+      "learning_rate": 8.194620987954922e-06,
+      "loss": 0.8304,
+      "step": 4703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2222016755817986,
+      "learning_rate": 8.193823581136919e-06,
+      "loss": 0.8588,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.146148455087148,
+      "learning_rate": 8.193026037073201e-06,
+      "loss": 0.8478,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3588039963884078,
+      "learning_rate": 8.192228355798045e-06,
+      "loss": 0.9128,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8103280281531509,
+      "learning_rate": 8.191430537345728e-06,
+      "loss": 0.896,
+      "step": 4707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1297806774370938,
+      "learning_rate": 8.19063258175053e-06,
+      "loss": 0.8722,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1450418455513156,
+      "learning_rate": 8.189834489046746e-06,
+      "loss": 0.9314,
+      "step": 4709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3529080503383697,
+      "learning_rate": 8.18903625926867e-06,
+      "loss": 0.8212,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.496053047393003,
+      "learning_rate": 8.188237892450603e-06,
+      "loss": 0.8662,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.602384865195676,
+      "learning_rate": 8.187439388626855e-06,
+      "loss": 0.9434,
+      "step": 4712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.999103904478709,
+      "learning_rate": 8.186640747831735e-06,
+      "loss": 0.8652,
+      "step": 4713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.687704040440643,
+      "learning_rate": 8.185841970099566e-06,
+      "loss": 0.8791,
+      "step": 4714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2461857805610905,
+      "learning_rate": 8.185043055464673e-06,
+      "loss": 0.7684,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2801884656845277,
+      "learning_rate": 8.184244003961386e-06,
+      "loss": 0.9645,
+      "step": 4716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.611584709148049,
+      "learning_rate": 8.183444815624041e-06,
+      "loss": 0.9437,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.405833209686475,
+      "learning_rate": 8.182645490486986e-06,
+      "loss": 0.8392,
+      "step": 4718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5645173811036526,
+      "learning_rate": 8.181846028584563e-06,
+      "loss": 0.6957,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1480801524811977,
+      "learning_rate": 8.181046429951131e-06,
+      "loss": 0.8402,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.162213216858755,
+      "learning_rate": 8.180246694621048e-06,
+      "loss": 0.9671,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.156508177561667,
+      "learning_rate": 8.179446822628684e-06,
+      "loss": 0.8203,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0916151199269788,
+      "learning_rate": 8.178646814008407e-06,
+      "loss": 0.6704,
+      "step": 4723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.237830379481574,
+      "learning_rate": 8.177846668794598e-06,
+      "loss": 0.6864,
+      "step": 4724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.13615511806954,
+      "learning_rate": 8.177046387021641e-06,
+      "loss": 0.7453,
+      "step": 4725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0622429560086144,
+      "learning_rate": 8.176245968723924e-06,
+      "loss": 0.917,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.634778618763897,
+      "learning_rate": 8.175445413935844e-06,
+      "loss": 0.8903,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.248466460749978,
+      "learning_rate": 8.174644722691802e-06,
+      "loss": 0.6766,
+      "step": 4728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0952842954208506,
+      "learning_rate": 8.173843895026207e-06,
+      "loss": 0.798,
+      "step": 4729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.160619046903679,
+      "learning_rate": 8.17304293097347e-06,
+      "loss": 0.6639,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3836693334623855,
+      "learning_rate": 8.172241830568012e-06,
+      "loss": 0.8977,
+      "step": 4731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.511828605185999,
+      "learning_rate": 8.17144059384426e-06,
+      "loss": 0.8372,
+      "step": 4732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3275419393677654,
+      "learning_rate": 8.170639220836639e-06,
+      "loss": 0.7828,
+      "step": 4733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.16006094684137,
+      "learning_rate": 8.169837711579591e-06,
+      "loss": 0.8437,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1974055961582994,
+      "learning_rate": 8.169036066107556e-06,
+      "loss": 0.8025,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1665090325573484,
+      "learning_rate": 8.168234284454984e-06,
+      "loss": 0.6385,
+      "step": 4736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.842873846332037,
+      "learning_rate": 8.16743236665633e-06,
+      "loss": 0.9584,
+      "step": 4737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2841000860701097,
+      "learning_rate": 8.166630312746051e-06,
+      "loss": 0.7419,
+      "step": 4738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7401252918563195,
+      "learning_rate": 8.165828122758615e-06,
+      "loss": 0.9662,
+      "step": 4739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8204067874932766,
+      "learning_rate": 8.165025796728495e-06,
+      "loss": 0.8708,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.280939833909823,
+      "learning_rate": 8.164223334690168e-06,
+      "loss": 0.8052,
+      "step": 4741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7430734116988602,
+      "learning_rate": 8.163420736678115e-06,
+      "loss": 0.7577,
+      "step": 4742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.310115496337784,
+      "learning_rate": 8.162618002726829e-06,
+      "loss": 0.663,
+      "step": 4743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3875448942505413,
+      "learning_rate": 8.161815132870806e-06,
+      "loss": 0.9142,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1453321070632656,
+      "learning_rate": 8.161012127144543e-06,
+      "loss": 0.6352,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0897590403663835,
+      "learning_rate": 8.160208985582547e-06,
+      "loss": 0.7113,
+      "step": 4746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.143971738119541,
+      "learning_rate": 8.159405708219335e-06,
+      "loss": 0.8461,
+      "step": 4747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3316385417456744,
+      "learning_rate": 8.158602295089423e-06,
+      "loss": 0.9094,
+      "step": 4748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0493470221445333,
+      "learning_rate": 8.157798746227337e-06,
+      "loss": 0.9283,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.430433174651269,
+      "learning_rate": 8.156995061667605e-06,
+      "loss": 0.8933,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3114963440318075,
+      "learning_rate": 8.156191241444763e-06,
+      "loss": 0.9962,
+      "step": 4751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.837602416107323,
+      "learning_rate": 8.155387285593356e-06,
+      "loss": 0.7771,
+      "step": 4752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.822601452039977,
+      "learning_rate": 8.154583194147929e-06,
+      "loss": 1.0071,
+      "step": 4753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.243041865543386,
+      "learning_rate": 8.153778967143035e-06,
+      "loss": 1.0404,
+      "step": 4754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.291688455767539,
+      "learning_rate": 8.152974604613238e-06,
+      "loss": 0.8497,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0047985473520518,
+      "learning_rate": 8.152170106593099e-06,
+      "loss": 1.0596,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0823814707283033,
+      "learning_rate": 8.15136547311719e-06,
+      "loss": 0.8198,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9912293212532098,
+      "learning_rate": 8.150560704220087e-06,
+      "loss": 0.6182,
+      "step": 4758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3952477457595895,
+      "learning_rate": 8.149755799936377e-06,
+      "loss": 0.7883,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1537412470640174,
+      "learning_rate": 8.148950760300642e-06,
+      "loss": 0.6955,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6024693954427174,
+      "learning_rate": 8.148145585347482e-06,
+      "loss": 0.8566,
+      "step": 4761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.159627332996029,
+      "learning_rate": 8.147340275111492e-06,
+      "loss": 0.8212,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4503567970047024,
+      "learning_rate": 8.146534829627285e-06,
+      "loss": 0.9573,
+      "step": 4763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.027511260704952,
+      "learning_rate": 8.145729248929466e-06,
+      "loss": 0.8352,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.509769658392434,
+      "learning_rate": 8.144923533052655e-06,
+      "loss": 0.8674,
+      "step": 4765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1801347510924518,
+      "learning_rate": 8.144117682031478e-06,
+      "loss": 0.7267,
+      "step": 4766
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0861065084201573,
+      "learning_rate": 8.14331169590056e-06,
+      "loss": 0.6219,
+      "step": 4767
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.753961241638263,
+      "learning_rate": 8.142505574694539e-06,
+      "loss": 0.8,
+      "step": 4768
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9368852959751184,
+      "learning_rate": 8.141699318448053e-06,
+      "loss": 0.8533,
+      "step": 4769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.169161523298369,
+      "learning_rate": 8.14089292719575e-06,
+      "loss": 0.9064,
+      "step": 4770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.090821206075882,
+      "learning_rate": 8.140086400972285e-06,
+      "loss": 1.0533,
+      "step": 4771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2442506994024694,
+      "learning_rate": 8.139279739812314e-06,
+      "loss": 1.1113,
+      "step": 4772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7626843471537406,
+      "learning_rate": 8.1384729437505e-06,
+      "loss": 0.9178,
+      "step": 4773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.190956588824387,
+      "learning_rate": 8.137666012821514e-06,
+      "loss": 0.7651,
+      "step": 4774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9790939301665071,
+      "learning_rate": 8.136858947060032e-06,
+      "loss": 0.8833,
+      "step": 4775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.128109597952058,
+      "learning_rate": 8.136051746500735e-06,
+      "loss": 0.761,
+      "step": 4776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.375412694312191,
+      "learning_rate": 8.135244411178309e-06,
+      "loss": 0.9955,
+      "step": 4777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.953027142928454,
+      "learning_rate": 8.134436941127448e-06,
+      "loss": 0.7301,
+      "step": 4778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9891241446987107,
+      "learning_rate": 8.13362933638285e-06,
+      "loss": 0.837,
+      "step": 4779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3835858981157534,
+      "learning_rate": 8.132821596979223e-06,
+      "loss": 0.9211,
+      "step": 4780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.690416122811506,
+      "learning_rate": 8.132013722951275e-06,
+      "loss": 0.6592,
+      "step": 4781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9237504585808383,
+      "learning_rate": 8.13120571433372e-06,
+      "loss": 0.8826,
+      "step": 4782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2002655579605825,
+      "learning_rate": 8.130397571161283e-06,
+      "loss": 0.913,
+      "step": 4783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3241606877511807,
+      "learning_rate": 8.129589293468689e-06,
+      "loss": 0.6168,
+      "step": 4784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4923143925364437,
+      "learning_rate": 8.128780881290674e-06,
+      "loss": 0.7353,
+      "step": 4785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.509139748723546,
+      "learning_rate": 8.127972334661978e-06,
+      "loss": 0.6971,
+      "step": 4786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5200035541943677,
+      "learning_rate": 8.127163653617342e-06,
+      "loss": 0.8399,
+      "step": 4787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4307493962287636,
+      "learning_rate": 8.12635483819152e-06,
+      "loss": 0.7985,
+      "step": 4788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3719232740507925,
+      "learning_rate": 8.125545888419269e-06,
+      "loss": 0.7877,
+      "step": 4789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2288604289873228,
+      "learning_rate": 8.124736804335348e-06,
+      "loss": 0.7077,
+      "step": 4790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8755002579664066,
+      "learning_rate": 8.123927585974529e-06,
+      "loss": 0.8517,
+      "step": 4791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6999553076095824,
+      "learning_rate": 8.123118233371585e-06,
+      "loss": 0.8079,
+      "step": 4792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.192107025832407,
+      "learning_rate": 8.122308746561294e-06,
+      "loss": 0.7539,
+      "step": 4793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0431914018552852,
+      "learning_rate": 8.12149912557844e-06,
+      "loss": 0.8376,
+      "step": 4794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.757341227077783,
+      "learning_rate": 8.12068937045782e-06,
+      "loss": 0.9492,
+      "step": 4795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.240289802433996,
+      "learning_rate": 8.119879481234228e-06,
+      "loss": 1.0265,
+      "step": 4796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.205505375227648,
+      "learning_rate": 8.119069457942464e-06,
+      "loss": 0.8345,
+      "step": 4797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2111478321156324,
+      "learning_rate": 8.118259300617339e-06,
+      "loss": 0.9162,
+      "step": 4798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.332963940506511,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.8411,
+      "step": 4799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.076752384988596,
+      "learning_rate": 8.11663858400627e-06,
+      "loss": 0.8589,
+      "step": 4800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1819773717292674,
+      "learning_rate": 8.115828024789972e-06,
+      "loss": 0.7403,
+      "step": 4801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5903007359089405,
+      "learning_rate": 8.115017331679602e-06,
+      "loss": 1.0136,
+      "step": 4802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.460744752481784,
+      "learning_rate": 8.114206504710002e-06,
+      "loss": 0.9867,
+      "step": 4803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3900727516361016,
+      "learning_rate": 8.113395543916012e-06,
+      "loss": 0.6098,
+      "step": 4804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5454566352298853,
+      "learning_rate": 8.112584449332481e-06,
+      "loss": 0.7621,
+      "step": 4805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6675859757391909,
+      "learning_rate": 8.111773220994264e-06,
+      "loss": 0.8728,
+      "step": 4806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.378796070992985,
+      "learning_rate": 8.110961858936222e-06,
+      "loss": 1.0293,
+      "step": 4807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7957048595273943,
+      "learning_rate": 8.11015036319322e-06,
+      "loss": 0.7329,
+      "step": 4808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8194913477784311,
+      "learning_rate": 8.109338733800132e-06,
+      "loss": 0.9401,
+      "step": 4809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.411442944441635,
+      "learning_rate": 8.108526970791832e-06,
+      "loss": 0.9341,
+      "step": 4810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5012927694332263,
+      "learning_rate": 8.107715074203205e-06,
+      "loss": 0.7522,
+      "step": 4811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1227792179334033,
+      "learning_rate": 8.10690304406914e-06,
+      "loss": 0.8387,
+      "step": 4812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.404732454806915,
+      "learning_rate": 8.106090880424534e-06,
+      "loss": 0.8665,
+      "step": 4813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.993839668043478,
+      "learning_rate": 8.10527858330428e-06,
+      "loss": 0.7318,
+      "step": 4814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.869125189023947,
+      "learning_rate": 8.104466152743295e-06,
+      "loss": 0.8252,
+      "step": 4815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.471212608739906,
+      "learning_rate": 8.103653588776483e-06,
+      "loss": 0.9784,
+      "step": 4816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3387607080836252,
+      "learning_rate": 8.102840891438764e-06,
+      "loss": 0.7827,
+      "step": 4817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8418838000486213,
+      "learning_rate": 8.10202806076506e-06,
+      "loss": 0.7869,
+      "step": 4818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.078187505640914,
+      "learning_rate": 8.101215096790305e-06,
+      "loss": 0.8734,
+      "step": 4819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.473562623853066,
+      "learning_rate": 8.10040199954943e-06,
+      "loss": 0.7808,
+      "step": 4820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.160358122001241,
+      "learning_rate": 8.099588769077374e-06,
+      "loss": 0.6519,
+      "step": 4821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9750704991136059,
+      "learning_rate": 8.098775405409087e-06,
+      "loss": 0.839,
+      "step": 4822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7385652564935943,
+      "learning_rate": 8.09796190857952e-06,
+      "loss": 0.8958,
+      "step": 4823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8456506723941657,
+      "learning_rate": 8.097148278623628e-06,
+      "loss": 0.7588,
+      "step": 4824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2786981423248567,
+      "learning_rate": 8.096334515576379e-06,
+      "loss": 0.7867,
+      "step": 4825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3865238715080865,
+      "learning_rate": 8.095520619472739e-06,
+      "loss": 1.0115,
+      "step": 4826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1784686239833033,
+      "learning_rate": 8.094706590347683e-06,
+      "loss": 0.9294,
+      "step": 4827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.156376251416278,
+      "learning_rate": 8.093892428236192e-06,
+      "loss": 0.7731,
+      "step": 4828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6384315310081643,
+      "learning_rate": 8.093078133173256e-06,
+      "loss": 0.978,
+      "step": 4829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.714816309140252,
+      "learning_rate": 8.09226370519386e-06,
+      "loss": 0.7602,
+      "step": 4830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.323020576799651,
+      "learning_rate": 8.091449144333009e-06,
+      "loss": 0.844,
+      "step": 4831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1827228868013178,
+      "learning_rate": 8.090634450625704e-06,
+      "loss": 0.6817,
+      "step": 4832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4179288215905714,
+      "learning_rate": 8.08981962410695e-06,
+      "loss": 0.7548,
+      "step": 4833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.228039453698041,
+      "learning_rate": 8.089004664811767e-06,
+      "loss": 0.9605,
+      "step": 4834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0363157724476615,
+      "learning_rate": 8.088189572775173e-06,
+      "loss": 0.8386,
+      "step": 4835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3524158366936994,
+      "learning_rate": 8.087374348032198e-06,
+      "loss": 0.7617,
+      "step": 4836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9294052799909376,
+      "learning_rate": 8.08655899061787e-06,
+      "loss": 0.6684,
+      "step": 4837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2676666845498707,
+      "learning_rate": 8.085743500567228e-06,
+      "loss": 0.9606,
+      "step": 4838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7684266051802895,
+      "learning_rate": 8.084927877915314e-06,
+      "loss": 0.8917,
+      "step": 4839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1963079549278306,
+      "learning_rate": 8.084112122697178e-06,
+      "loss": 0.6904,
+      "step": 4840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5372077938208912,
+      "learning_rate": 8.083296234947878e-06,
+      "loss": 0.7894,
+      "step": 4841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9848453407136912,
+      "learning_rate": 8.082480214702472e-06,
+      "loss": 0.6358,
+      "step": 4842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.015428327384403,
+      "learning_rate": 8.081664061996024e-06,
+      "loss": 1.0289,
+      "step": 4843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.046231569826014,
+      "learning_rate": 8.080847776863609e-06,
+      "loss": 0.8617,
+      "step": 4844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7611039769644883,
+      "learning_rate": 8.080031359340303e-06,
+      "loss": 0.816,
+      "step": 4845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9353089653487239,
+      "learning_rate": 8.079214809461192e-06,
+      "loss": 0.7905,
+      "step": 4846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1715908449402024,
+      "learning_rate": 8.07839812726136e-06,
+      "loss": 0.9486,
+      "step": 4847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1053007660820415,
+      "learning_rate": 8.077581312775908e-06,
+      "loss": 0.7991,
+      "step": 4848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2622618284706464,
+      "learning_rate": 8.07676436603993e-06,
+      "loss": 1.0742,
+      "step": 4849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0628334614003787,
+      "learning_rate": 8.075947287088535e-06,
+      "loss": 0.997,
+      "step": 4850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1344115078337447,
+      "learning_rate": 8.075130075956836e-06,
+      "loss": 0.7179,
+      "step": 4851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3835028856136287,
+      "learning_rate": 8.074312732679948e-06,
+      "loss": 0.8358,
+      "step": 4852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.997399136365772,
+      "learning_rate": 8.073495257292996e-06,
+      "loss": 0.7691,
+      "step": 4853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2221119197366361,
+      "learning_rate": 8.072677649831107e-06,
+      "loss": 0.7211,
+      "step": 4854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8187946360014564,
+      "learning_rate": 8.071859910329416e-06,
+      "loss": 0.7419,
+      "step": 4855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.985483139850896,
+      "learning_rate": 8.071042038823063e-06,
+      "loss": 0.7542,
+      "step": 4856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2451293845112703,
+      "learning_rate": 8.070224035347193e-06,
+      "loss": 0.6819,
+      "step": 4857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.072582771244923,
+      "learning_rate": 8.069405899936961e-06,
+      "loss": 0.8321,
+      "step": 4858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0523248797983116,
+      "learning_rate": 8.068587632627521e-06,
+      "loss": 0.8376,
+      "step": 4859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2901995386130483,
+      "learning_rate": 8.067769233454037e-06,
+      "loss": 0.9556,
+      "step": 4860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.251588119376198,
+      "learning_rate": 8.066950702451673e-06,
+      "loss": 0.896,
+      "step": 4861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0972082585205953,
+      "learning_rate": 8.066132039655609e-06,
+      "loss": 0.6808,
+      "step": 4862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9747000669203334,
+      "learning_rate": 8.065313245101024e-06,
+      "loss": 1.023,
+      "step": 4863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1792212243203388,
+      "learning_rate": 8.064494318823102e-06,
+      "loss": 0.6226,
+      "step": 4864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8042006970772078,
+      "learning_rate": 8.06367526085703e-06,
+      "loss": 0.7478,
+      "step": 4865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3265864365941904,
+      "learning_rate": 8.062856071238014e-06,
+      "loss": 1.0051,
+      "step": 4866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.094085884266054,
+      "learning_rate": 8.062036750001249e-06,
+      "loss": 0.9908,
+      "step": 4867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9915167151326207,
+      "learning_rate": 8.061217297181946e-06,
+      "loss": 0.6497,
+      "step": 4868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1376006531453853,
+      "learning_rate": 8.060397712815318e-06,
+      "loss": 1.0608,
+      "step": 4869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.465842319245621,
+      "learning_rate": 8.059577996936583e-06,
+      "loss": 0.8713,
+      "step": 4870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.102877447360486,
+      "learning_rate": 8.05875814958097e-06,
+      "loss": 0.7732,
+      "step": 4871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5853051578274946,
+      "learning_rate": 8.057938170783704e-06,
+      "loss": 0.8519,
+      "step": 4872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9815283851379437,
+      "learning_rate": 8.057118060580027e-06,
+      "loss": 0.6265,
+      "step": 4873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9066664871955445,
+      "learning_rate": 8.056297819005177e-06,
+      "loss": 0.7299,
+      "step": 4874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7366834315622044,
+      "learning_rate": 8.055477446094404e-06,
+      "loss": 0.7258,
+      "step": 4875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2058060551429741,
+      "learning_rate": 8.054656941882959e-06,
+      "loss": 0.5631,
+      "step": 4876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5440477653546423,
+      "learning_rate": 8.053836306406102e-06,
+      "loss": 0.8454,
+      "step": 4877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.581833567109994,
+      "learning_rate": 8.053015539699099e-06,
+      "loss": 0.8059,
+      "step": 4878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.050342874945105,
+      "learning_rate": 8.052194641797217e-06,
+      "loss": 0.7237,
+      "step": 4879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0221539300480496,
+      "learning_rate": 8.051373612735736e-06,
+      "loss": 0.7189,
+      "step": 4880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9794379547500194,
+      "learning_rate": 8.050552452549933e-06,
+      "loss": 0.7981,
+      "step": 4881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1730497888402627,
+      "learning_rate": 8.049731161275099e-06,
+      "loss": 0.919,
+      "step": 4882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.201649294346357,
+      "learning_rate": 8.048909738946523e-06,
+      "loss": 0.791,
+      "step": 4883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0931707629484575,
+      "learning_rate": 8.048088185599507e-06,
+      "loss": 0.9376,
+      "step": 4884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1743975994998173,
+      "learning_rate": 8.047266501269352e-06,
+      "loss": 0.7738,
+      "step": 4885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.088692499918127,
+      "learning_rate": 8.046444685991369e-06,
+      "loss": 0.7136,
+      "step": 4886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.275805382246357,
+      "learning_rate": 8.045622739800873e-06,
+      "loss": 0.9408,
+      "step": 4887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2130300283640696,
+      "learning_rate": 8.044800662733185e-06,
+      "loss": 1.0978,
+      "step": 4888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5495173523293975,
+      "learning_rate": 8.043978454823632e-06,
+      "loss": 0.9699,
+      "step": 4889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0611104760814536,
+      "learning_rate": 8.043156116107545e-06,
+      "loss": 0.9008,
+      "step": 4890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.400807998314811,
+      "learning_rate": 8.042333646620262e-06,
+      "loss": 0.6442,
+      "step": 4891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.726655615232083,
+      "learning_rate": 8.041511046397127e-06,
+      "loss": 0.9009,
+      "step": 4892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.230985517879595,
+      "learning_rate": 8.040688315473489e-06,
+      "loss": 0.8798,
+      "step": 4893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2336525405286132,
+      "learning_rate": 8.0398654538847e-06,
+      "loss": 0.8058,
+      "step": 4894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0644253980226215,
+      "learning_rate": 8.039042461666125e-06,
+      "loss": 0.7778,
+      "step": 4895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.490327774985131,
+      "learning_rate": 8.038219338853126e-06,
+      "loss": 0.8738,
+      "step": 4896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5135255641860175,
+      "learning_rate": 8.037396085481075e-06,
+      "loss": 0.6222,
+      "step": 4897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8199540513429044,
+      "learning_rate": 8.03657270158535e-06,
+      "loss": 0.8681,
+      "step": 4898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0102628611451614,
+      "learning_rate": 8.035749187201333e-06,
+      "loss": 0.8494,
+      "step": 4899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8670752983769756,
+      "learning_rate": 8.034925542364412e-06,
+      "loss": 0.6891,
+      "step": 4900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7829245821498048,
+      "learning_rate": 8.034101767109982e-06,
+      "loss": 0.9411,
+      "step": 4901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6670081465078654,
+      "learning_rate": 8.033277861473441e-06,
+      "loss": 0.8168,
+      "step": 4902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.415294221715343,
+      "learning_rate": 8.032453825490197e-06,
+      "loss": 0.8576,
+      "step": 4903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.852289564774182,
+      "learning_rate": 8.031629659195657e-06,
+      "loss": 0.8427,
+      "step": 4904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9286195639861183,
+      "learning_rate": 8.030805362625238e-06,
+      "loss": 0.8749,
+      "step": 4905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.586572244508462,
+      "learning_rate": 8.029980935814363e-06,
+      "loss": 1.0343,
+      "step": 4906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.205873926887469,
+      "learning_rate": 8.029156378798459e-06,
+      "loss": 0.9879,
+      "step": 4907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0257880369750265,
+      "learning_rate": 8.028331691612961e-06,
+      "loss": 0.8545,
+      "step": 4908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.085783007447975,
+      "learning_rate": 8.027506874293304e-06,
+      "loss": 0.6674,
+      "step": 4909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9985151985364151,
+      "learning_rate": 8.026681926874932e-06,
+      "loss": 0.9289,
+      "step": 4910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8004810587874938,
+      "learning_rate": 8.025856849393301e-06,
+      "loss": 0.7972,
+      "step": 4911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1922087555856655,
+      "learning_rate": 8.02503164188386e-06,
+      "loss": 0.8919,
+      "step": 4912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8519639591453763,
+      "learning_rate": 8.024206304382072e-06,
+      "loss": 0.8556,
+      "step": 4913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7188704327146853,
+      "learning_rate": 8.023380836923404e-06,
+      "loss": 0.7893,
+      "step": 4914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7303183331752041,
+      "learning_rate": 8.022555239543328e-06,
+      "loss": 0.9473,
+      "step": 4915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.122050434767405,
+      "learning_rate": 8.021729512277323e-06,
+      "loss": 0.6335,
+      "step": 4916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0575555910875467,
+      "learning_rate": 8.02090365516087e-06,
+      "loss": 0.8305,
+      "step": 4917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8340083026821399,
+      "learning_rate": 8.02007766822946e-06,
+      "loss": 0.8537,
+      "step": 4918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.226240695996376,
+      "learning_rate": 8.019251551518585e-06,
+      "loss": 0.9807,
+      "step": 4919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.663220677697593,
+      "learning_rate": 8.018425305063749e-06,
+      "loss": 0.8376,
+      "step": 4920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0229129203528784,
+      "learning_rate": 8.017598928900452e-06,
+      "loss": 0.669,
+      "step": 4921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.935382544385908,
+      "learning_rate": 8.01677242306421e-06,
+      "loss": 0.8292,
+      "step": 4922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.254538777745868,
+      "learning_rate": 8.015945787590537e-06,
+      "loss": 0.7295,
+      "step": 4923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.655148577265021,
+      "learning_rate": 8.015119022514958e-06,
+      "loss": 0.8335,
+      "step": 4924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0973970617891478,
+      "learning_rate": 8.014292127873002e-06,
+      "loss": 0.7468,
+      "step": 4925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0269154105420495,
+      "learning_rate": 8.013465103700197e-06,
+      "loss": 0.8008,
+      "step": 4926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0936890246638518,
+      "learning_rate": 8.012637950032086e-06,
+      "loss": 0.8426,
+      "step": 4927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.00050376439011,
+      "learning_rate": 8.011810666904212e-06,
+      "loss": 0.87,
+      "step": 4928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1706163105693665,
+      "learning_rate": 8.010983254352127e-06,
+      "loss": 0.8743,
+      "step": 4929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9573590147945477,
+      "learning_rate": 8.010155712411386e-06,
+      "loss": 0.9415,
+      "step": 4930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3988540977124826,
+      "learning_rate": 8.009328041117547e-06,
+      "loss": 0.8988,
+      "step": 4931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5370062812307808,
+      "learning_rate": 8.008500240506182e-06,
+      "loss": 0.7544,
+      "step": 4932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9851131112564435,
+      "learning_rate": 8.007672310612863e-06,
+      "loss": 0.9242,
+      "step": 4933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9310922623261944,
+      "learning_rate": 8.006844251473165e-06,
+      "loss": 1.012,
+      "step": 4934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.794400786462738,
+      "learning_rate": 8.006016063122672e-06,
+      "loss": 0.8697,
+      "step": 4935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5205575594961553,
+      "learning_rate": 8.005187745596976e-06,
+      "loss": 0.8695,
+      "step": 4936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.712311020114758,
+      "learning_rate": 8.004359298931668e-06,
+      "loss": 0.8194,
+      "step": 4937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.5756826283248,
+      "learning_rate": 8.003530723162352e-06,
+      "loss": 0.8584,
+      "step": 4938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.053238185000056,
+      "learning_rate": 8.002702018324629e-06,
+      "loss": 0.8742,
+      "step": 4939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1751419781482735,
+      "learning_rate": 8.001873184454114e-06,
+      "loss": 0.655,
+      "step": 4940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.081486007732901,
+      "learning_rate": 8.001044221586423e-06,
+      "loss": 0.7464,
+      "step": 4941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1544601322710584,
+      "learning_rate": 8.000215129757178e-06,
+      "loss": 0.8335,
+      "step": 4942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2473469823142893,
+      "learning_rate": 7.999385909002007e-06,
+      "loss": 0.8774,
+      "step": 4943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8685205740875392,
+      "learning_rate": 7.998556559356543e-06,
+      "loss": 0.7477,
+      "step": 4944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.796584638645999,
+      "learning_rate": 7.997727080856425e-06,
+      "loss": 0.8703,
+      "step": 4945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3345780627774704,
+      "learning_rate": 7.9968974735373e-06,
+      "loss": 0.8483,
+      "step": 4946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.604017570865661,
+      "learning_rate": 7.996067737434815e-06,
+      "loss": 0.9123,
+      "step": 4947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0756460535617696,
+      "learning_rate": 7.995237872584627e-06,
+      "loss": 0.9915,
+      "step": 4948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.20957145690792,
+      "learning_rate": 7.994407879022397e-06,
+      "loss": 0.8744,
+      "step": 4949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2005176055398161,
+      "learning_rate": 7.99357775678379e-06,
+      "loss": 0.7445,
+      "step": 4950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8968311124795372,
+      "learning_rate": 7.99274750590448e-06,
+      "loss": 1.0112,
+      "step": 4951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9564628823836587,
+      "learning_rate": 7.991917126420147e-06,
+      "loss": 0.993,
+      "step": 4952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2356779168405505,
+      "learning_rate": 7.99108661836647e-06,
+      "loss": 0.885,
+      "step": 4953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.941678670226812,
+      "learning_rate": 7.990255981779139e-06,
+      "loss": 0.7826,
+      "step": 4954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.517931070114798,
+      "learning_rate": 7.989425216693849e-06,
+      "loss": 0.8707,
+      "step": 4955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.082435241739974,
+      "learning_rate": 7.9885943231463e-06,
+      "loss": 0.9213,
+      "step": 4956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2926147836863286,
+      "learning_rate": 7.987763301172197e-06,
+      "loss": 0.8967,
+      "step": 4957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7076690945583786,
+      "learning_rate": 7.986932150807253e-06,
+      "loss": 0.8813,
+      "step": 4958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3320820871986108,
+      "learning_rate": 7.986100872087177e-06,
+      "loss": 1.0476,
+      "step": 4959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.59147836764515,
+      "learning_rate": 7.9852694650477e-06,
+      "loss": 0.6923,
+      "step": 4960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1262256357527063,
+      "learning_rate": 7.984437929724546e-06,
+      "loss": 0.8844,
+      "step": 4961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.995850582127941,
+      "learning_rate": 7.983606266153445e-06,
+      "loss": 0.8336,
+      "step": 4962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0781977203662445,
+      "learning_rate": 7.98277447437014e-06,
+      "loss": 0.6231,
+      "step": 4963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.789360017292589,
+      "learning_rate": 7.981942554410371e-06,
+      "loss": 0.7881,
+      "step": 4964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1398684618334927,
+      "learning_rate": 7.98111050630989e-06,
+      "loss": 0.8694,
+      "step": 4965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9838835821532763,
+      "learning_rate": 7.980278330104452e-06,
+      "loss": 0.6909,
+      "step": 4966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8062089101355727,
+      "learning_rate": 7.979446025829815e-06,
+      "loss": 0.9459,
+      "step": 4967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9351442431626333,
+      "learning_rate": 7.978613593521748e-06,
+      "loss": 0.8219,
+      "step": 4968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9826097344736036,
+      "learning_rate": 7.97778103321602e-06,
+      "loss": 0.7611,
+      "step": 4969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1714593284581325,
+      "learning_rate": 7.976948344948412e-06,
+      "loss": 0.7481,
+      "step": 4970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0363218727567913,
+      "learning_rate": 7.976115528754702e-06,
+      "loss": 0.807,
+      "step": 4971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.800881578420221,
+      "learning_rate": 7.975282584670679e-06,
+      "loss": 0.7913,
+      "step": 4972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.005297953627428,
+      "learning_rate": 7.974449512732137e-06,
+      "loss": 1.0466,
+      "step": 4973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9478035093752113,
+      "learning_rate": 7.973616312974876e-06,
+      "loss": 0.8754,
+      "step": 4974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.040403827746014,
+      "learning_rate": 7.972782985434697e-06,
+      "loss": 0.6626,
+      "step": 4975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.074207844184691,
+      "learning_rate": 7.971949530147417e-06,
+      "loss": 0.9182,
+      "step": 4976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4134469947393278,
+      "learning_rate": 7.971115947148842e-06,
+      "loss": 0.9887,
+      "step": 4977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1777536195883793,
+      "learning_rate": 7.9702822364748e-06,
+      "loss": 0.8124,
+      "step": 4978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5301266911425255,
+      "learning_rate": 7.969448398161115e-06,
+      "loss": 1.0368,
+      "step": 4979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6071565212257486,
+      "learning_rate": 7.968614432243617e-06,
+      "loss": 0.7138,
+      "step": 4980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6281409761867212,
+      "learning_rate": 7.96778033875815e-06,
+      "loss": 0.8197,
+      "step": 4981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7381702081180084,
+      "learning_rate": 7.966946117740548e-06,
+      "loss": 0.837,
+      "step": 4982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3174409023777014,
+      "learning_rate": 7.966111769226666e-06,
+      "loss": 0.8555,
+      "step": 4983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.272960673169631,
+      "learning_rate": 7.965277293252354e-06,
+      "loss": 0.8298,
+      "step": 4984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7965374661693656,
+      "learning_rate": 7.964442689853473e-06,
+      "loss": 0.9719,
+      "step": 4985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8034161339343828,
+      "learning_rate": 7.963607959065887e-06,
+      "loss": 0.8355,
+      "step": 4986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3919347520103735,
+      "learning_rate": 7.962773100925467e-06,
+      "loss": 0.9297,
+      "step": 4987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2639972916751443,
+      "learning_rate": 7.961938115468088e-06,
+      "loss": 1.0426,
+      "step": 4988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5098666615137173,
+      "learning_rate": 7.961103002729634e-06,
+      "loss": 0.8859,
+      "step": 4989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1876379921466778,
+      "learning_rate": 7.960267762745985e-06,
+      "loss": 0.9399,
+      "step": 4990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1597808175475324,
+      "learning_rate": 7.95943239555304e-06,
+      "loss": 0.7827,
+      "step": 4991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.575598717518532,
+      "learning_rate": 7.958596901186695e-06,
+      "loss": 0.8975,
+      "step": 4992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9410290154083583,
+      "learning_rate": 7.95776127968285e-06,
+      "loss": 0.8979,
+      "step": 4993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.358502332628991,
+      "learning_rate": 7.956925531077417e-06,
+      "loss": 0.9783,
+      "step": 4994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.51154821813832,
+      "learning_rate": 7.956089655406307e-06,
+      "loss": 0.8355,
+      "step": 4995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3385522410193873,
+      "learning_rate": 7.955253652705442e-06,
+      "loss": 0.7412,
+      "step": 4996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.505926897755842,
+      "learning_rate": 7.954417523010747e-06,
+      "loss": 0.7068,
+      "step": 4997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.071547012669307,
+      "learning_rate": 7.953581266358148e-06,
+      "loss": 0.8544,
+      "step": 4998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.010416366714736,
+      "learning_rate": 7.952744882783587e-06,
+      "loss": 0.6773,
+      "step": 4999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0782649270427092,
+      "learning_rate": 7.951908372323001e-06,
+      "loss": 0.5928,
+      "step": 5000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1834468051808824,
+      "learning_rate": 7.951071735012337e-06,
+      "loss": 0.8331,
+      "step": 5001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6150548590855416,
+      "learning_rate": 7.95023497088755e-06,
+      "loss": 0.681,
+      "step": 5002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9853677006202608,
+      "learning_rate": 7.949398079984594e-06,
+      "loss": 0.7595,
+      "step": 5003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.402157030286248,
+      "learning_rate": 7.948561062339435e-06,
+      "loss": 0.8234,
+      "step": 5004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1568003808943872,
+      "learning_rate": 7.94772391798804e-06,
+      "loss": 0.6521,
+      "step": 5005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.990783492907402,
+      "learning_rate": 7.946886646966384e-06,
+      "loss": 0.8781,
+      "step": 5006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8527415616625105,
+      "learning_rate": 7.946049249310445e-06,
+      "loss": 0.9275,
+      "step": 5007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9806736075826445,
+      "learning_rate": 7.945211725056208e-06,
+      "loss": 0.758,
+      "step": 5008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3884546970721163,
+      "learning_rate": 7.944374074239665e-06,
+      "loss": 0.8205,
+      "step": 5009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6326804191809217,
+      "learning_rate": 7.94353629689681e-06,
+      "loss": 0.8235,
+      "step": 5010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0629042791646848,
+      "learning_rate": 7.942698393063646e-06,
+      "loss": 0.8498,
+      "step": 5011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.774943612944439,
+      "learning_rate": 7.941860362776176e-06,
+      "loss": 0.7335,
+      "step": 5012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.174123882956257,
+      "learning_rate": 7.941022206070415e-06,
+      "loss": 0.7188,
+      "step": 5013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.095064032713728,
+      "learning_rate": 7.940183922982381e-06,
+      "loss": 0.9108,
+      "step": 5014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2783703197920677,
+      "learning_rate": 7.939345513548095e-06,
+      "loss": 0.9139,
+      "step": 5015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.062197053239144,
+      "learning_rate": 7.938506977803586e-06,
+      "loss": 0.8314,
+      "step": 5016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.201781026522572,
+      "learning_rate": 7.937668315784888e-06,
+      "loss": 0.8423,
+      "step": 5017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0189241823284774,
+      "learning_rate": 7.93682952752804e-06,
+      "loss": 0.6129,
+      "step": 5018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1621530397479214,
+      "learning_rate": 7.935990613069087e-06,
+      "loss": 0.8125,
+      "step": 5019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0293956685854724,
+      "learning_rate": 7.935151572444076e-06,
+      "loss": 1.0328,
+      "step": 5020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1519821617307926,
+      "learning_rate": 7.934312405689067e-06,
+      "loss": 0.7927,
+      "step": 5021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6299420813679353,
+      "learning_rate": 7.933473112840117e-06,
+      "loss": 0.9187,
+      "step": 5022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3379208860613163,
+      "learning_rate": 7.932633693933298e-06,
+      "loss": 0.7491,
+      "step": 5023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9384727856496736,
+      "learning_rate": 7.931794149004675e-06,
+      "loss": 0.6466,
+      "step": 5024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9186164775051155,
+      "learning_rate": 7.930954478090329e-06,
+      "loss": 1.0491,
+      "step": 5025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.519814727168082,
+      "learning_rate": 7.930114681226341e-06,
+      "loss": 0.7607,
+      "step": 5026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0890279953890107,
+      "learning_rate": 7.9292747584488e-06,
+      "loss": 0.6369,
+      "step": 5027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1979144045853354,
+      "learning_rate": 7.928434709793798e-06,
+      "loss": 0.9223,
+      "step": 5028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0595641161996103,
+      "learning_rate": 7.927594535297433e-06,
+      "loss": 0.9674,
+      "step": 5029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2037327629790053,
+      "learning_rate": 7.926754234995814e-06,
+      "loss": 0.8653,
+      "step": 5030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6025236856075056,
+      "learning_rate": 7.925913808925045e-06,
+      "loss": 0.864,
+      "step": 5031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.393117100554993,
+      "learning_rate": 7.925073257121245e-06,
+      "loss": 0.742,
+      "step": 5032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8268968263578926,
+      "learning_rate": 7.924232579620533e-06,
+      "loss": 0.8145,
+      "step": 5033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8758194923976845,
+      "learning_rate": 7.923391776459031e-06,
+      "loss": 0.8834,
+      "step": 5034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1437148201926641,
+      "learning_rate": 7.922550847672876e-06,
+      "loss": 0.7275,
+      "step": 5035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3017247642579493,
+      "learning_rate": 7.921709793298202e-06,
+      "loss": 0.7189,
+      "step": 5036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.953042913673986,
+      "learning_rate": 7.920868613371151e-06,
+      "loss": 0.7565,
+      "step": 5037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.062773683458912,
+      "learning_rate": 7.920027307927872e-06,
+      "loss": 0.7319,
+      "step": 5038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1328031547267376,
+      "learning_rate": 7.919185877004515e-06,
+      "loss": 0.7279,
+      "step": 5039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1725135788365635,
+      "learning_rate": 7.91834432063724e-06,
+      "loss": 0.702,
+      "step": 5040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0866140004701346,
+      "learning_rate": 7.917502638862209e-06,
+      "loss": 0.9656,
+      "step": 5041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.703508601473017,
+      "learning_rate": 7.916660831715592e-06,
+      "loss": 0.8437,
+      "step": 5042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1473331797449937,
+      "learning_rate": 7.915818899233564e-06,
+      "loss": 0.999,
+      "step": 5043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.111545001893912,
+      "learning_rate": 7.914976841452304e-06,
+      "loss": 0.9081,
+      "step": 5044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9502231815712923,
+      "learning_rate": 7.914134658407997e-06,
+      "loss": 0.9888,
+      "step": 5045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.295191509316323,
+      "learning_rate": 7.913292350136834e-06,
+      "loss": 0.7705,
+      "step": 5046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.182957432066007,
+      "learning_rate": 7.912449916675008e-06,
+      "loss": 0.6985,
+      "step": 5047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0142604890664666,
+      "learning_rate": 7.911607358058726e-06,
+      "loss": 1.0019,
+      "step": 5048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.809034070211368,
+      "learning_rate": 7.91076467432419e-06,
+      "loss": 0.7416,
+      "step": 5049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8378398970611094,
+      "learning_rate": 7.909921865507614e-06,
+      "loss": 0.8393,
+      "step": 5050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2299383951818794,
+      "learning_rate": 7.909078931645215e-06,
+      "loss": 0.6377,
+      "step": 5051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1199505691788247,
+      "learning_rate": 7.908235872773214e-06,
+      "loss": 0.884,
+      "step": 5052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0705726344548006,
+      "learning_rate": 7.907392688927842e-06,
+      "loss": 0.6807,
+      "step": 5053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3673322622756903,
+      "learning_rate": 7.90654938014533e-06,
+      "loss": 0.9125,
+      "step": 5054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1835396390881106,
+      "learning_rate": 7.905705946461918e-06,
+      "loss": 0.9817,
+      "step": 5055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.944830498595022,
+      "learning_rate": 7.90486238791385e-06,
+      "loss": 0.8494,
+      "step": 5056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6268119435138804,
+      "learning_rate": 7.904018704537376e-06,
+      "loss": 0.7039,
+      "step": 5057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.129995812573889,
+      "learning_rate": 7.90317489636875e-06,
+      "loss": 0.8648,
+      "step": 5058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2934558591477203,
+      "learning_rate": 7.902330963444234e-06,
+      "loss": 0.7895,
+      "step": 5059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.969982296978998,
+      "learning_rate": 7.901486905800092e-06,
+      "loss": 0.8354,
+      "step": 5060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.316309841280496,
+      "learning_rate": 7.900642723472596e-06,
+      "loss": 0.8005,
+      "step": 5061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.7918001183271945,
+      "learning_rate": 7.899798416498023e-06,
+      "loss": 0.7294,
+      "step": 5062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.309969283575044,
+      "learning_rate": 7.898953984912652e-06,
+      "loss": 0.6745,
+      "step": 5063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0576456520271162,
+      "learning_rate": 7.898109428752773e-06,
+      "loss": 0.6983,
+      "step": 5064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2720000515832366,
+      "learning_rate": 7.897264748054679e-06,
+      "loss": 1.0614,
+      "step": 5065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2320675568336776,
+      "learning_rate": 7.896419942854662e-06,
+      "loss": 0.781,
+      "step": 5066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.14227696195202,
+      "learning_rate": 7.895575013189032e-06,
+      "loss": 0.8408,
+      "step": 5067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9661584838835064,
+      "learning_rate": 7.894729959094097e-06,
+      "loss": 0.7365,
+      "step": 5068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.174480308414642,
+      "learning_rate": 7.893884780606164e-06,
+      "loss": 0.6479,
+      "step": 5069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.32899840770779,
+      "learning_rate": 7.893039477761562e-06,
+      "loss": 0.7888,
+      "step": 5070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.963248401988271,
+      "learning_rate": 7.892194050596606e-06,
+      "loss": 1.0361,
+      "step": 5071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.010562903162415,
+      "learning_rate": 7.891348499147634e-06,
+      "loss": 0.7778,
+      "step": 5072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1447131569936038,
+      "learning_rate": 7.890502823450975e-06,
+      "loss": 0.9454,
+      "step": 5073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5259302720318404,
+      "learning_rate": 7.889657023542973e-06,
+      "loss": 0.7518,
+      "step": 5074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8864565755139138,
+      "learning_rate": 7.888811099459974e-06,
+      "loss": 0.8019,
+      "step": 5075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0661353907658182,
+      "learning_rate": 7.887965051238326e-06,
+      "loss": 0.7794,
+      "step": 5076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7924306486732218,
+      "learning_rate": 7.88711887891439e-06,
+      "loss": 0.8256,
+      "step": 5077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4455089908866239,
+      "learning_rate": 7.886272582524525e-06,
+      "loss": 0.7824,
+      "step": 5078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3042314553102257,
+      "learning_rate": 7.885426162105101e-06,
+      "loss": 0.9376,
+      "step": 5079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9920571721869171,
+      "learning_rate": 7.884579617692487e-06,
+      "loss": 0.9531,
+      "step": 5080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2101311714595226,
+      "learning_rate": 7.883732949323061e-06,
+      "loss": 0.6158,
+      "step": 5081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1238030069852765,
+      "learning_rate": 7.882886157033209e-06,
+      "loss": 0.8166,
+      "step": 5082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1043625924978206,
+      "learning_rate": 7.882039240859318e-06,
+      "loss": 0.6792,
+      "step": 5083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3679599306716845,
+      "learning_rate": 7.881192200837785e-06,
+      "loss": 0.8895,
+      "step": 5084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8873421050511707,
+      "learning_rate": 7.880345037005002e-06,
+      "loss": 0.9798,
+      "step": 5085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.468807210753197,
+      "learning_rate": 7.879497749397382e-06,
+      "loss": 0.9325,
+      "step": 5086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3452054284996176,
+      "learning_rate": 7.878650338051328e-06,
+      "loss": 0.7823,
+      "step": 5087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9536984185962512,
+      "learning_rate": 7.877802803003258e-06,
+      "loss": 0.8247,
+      "step": 5088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2499912060367335,
+      "learning_rate": 7.876955144289594e-06,
+      "loss": 0.9948,
+      "step": 5089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7907125501545307,
+      "learning_rate": 7.876107361946758e-06,
+      "loss": 0.8837,
+      "step": 5090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3277346916260626,
+      "learning_rate": 7.875259456011185e-06,
+      "loss": 0.8598,
+      "step": 5091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9716847465196592,
+      "learning_rate": 7.874411426519308e-06,
+      "loss": 0.8084,
+      "step": 5092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0807654258317518,
+      "learning_rate": 7.873563273507571e-06,
+      "loss": 0.6628,
+      "step": 5093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.46071525515466,
+      "learning_rate": 7.872714997012421e-06,
+      "loss": 0.9608,
+      "step": 5094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4967527115617574,
+      "learning_rate": 7.87186659707031e-06,
+      "loss": 0.7452,
+      "step": 5095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.642592426604549,
+      "learning_rate": 7.871018073717693e-06,
+      "loss": 0.9075,
+      "step": 5096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.40587187790782,
+      "learning_rate": 7.870169426991038e-06,
+      "loss": 0.8824,
+      "step": 5097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.354773333445508,
+      "learning_rate": 7.86932065692681e-06,
+      "loss": 0.9739,
+      "step": 5098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.048620854638149,
+      "learning_rate": 7.868471763561482e-06,
+      "loss": 0.7911,
+      "step": 5099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.103079600418708,
+      "learning_rate": 7.867622746931533e-06,
+      "loss": 0.6866,
+      "step": 5100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1100570073767004,
+      "learning_rate": 7.86677360707345e-06,
+      "loss": 0.7963,
+      "step": 5101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0503285880848443,
+      "learning_rate": 7.865924344023722e-06,
+      "loss": 0.8359,
+      "step": 5102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.072808072027394,
+      "learning_rate": 7.865074957818839e-06,
+      "loss": 0.7258,
+      "step": 5103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0279428227686955,
+      "learning_rate": 7.864225448495304e-06,
+      "loss": 0.581,
+      "step": 5104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.107738092070089,
+      "learning_rate": 7.863375816089624e-06,
+      "loss": 0.6957,
+      "step": 5105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1312285342928328,
+      "learning_rate": 7.862526060638309e-06,
+      "loss": 0.8095,
+      "step": 5106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.00798974948055,
+      "learning_rate": 7.861676182177873e-06,
+      "loss": 0.8364,
+      "step": 5107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8557630282266986,
+      "learning_rate": 7.860826180744837e-06,
+      "loss": 0.8308,
+      "step": 5108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.196728094256672,
+      "learning_rate": 7.85997605637573e-06,
+      "loss": 0.9311,
+      "step": 5109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.453621596888353,
+      "learning_rate": 7.859125809107082e-06,
+      "loss": 0.9936,
+      "step": 5110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.62836006401649,
+      "learning_rate": 7.85827543897543e-06,
+      "loss": 1.0426,
+      "step": 5111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.290046674204825,
+      "learning_rate": 7.857424946017317e-06,
+      "loss": 0.7375,
+      "step": 5112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.245042459615523,
+      "learning_rate": 7.85657433026929e-06,
+      "loss": 0.7105,
+      "step": 5113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.351846692936656,
+      "learning_rate": 7.855723591767903e-06,
+      "loss": 0.8327,
+      "step": 5114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2930332638990003,
+      "learning_rate": 7.854872730549712e-06,
+      "loss": 0.7288,
+      "step": 5115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9938708515270491,
+      "learning_rate": 7.854021746651283e-06,
+      "loss": 0.7485,
+      "step": 5116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.43877847453149,
+      "learning_rate": 7.853170640109182e-06,
+      "loss": 0.8701,
+      "step": 5117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0750311369128684,
+      "learning_rate": 7.852319410959984e-06,
+      "loss": 0.6942,
+      "step": 5118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.321215312739264,
+      "learning_rate": 7.85146805924027e-06,
+      "loss": 0.7396,
+      "step": 5119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8566818300763066,
+      "learning_rate": 7.850616584986621e-06,
+      "loss": 1.0407,
+      "step": 5120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1132324146951875,
+      "learning_rate": 7.84976498823563e-06,
+      "loss": 0.6398,
+      "step": 5121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.176704602908476,
+      "learning_rate": 7.84891326902389e-06,
+      "loss": 0.7486,
+      "step": 5122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3628027220317915,
+      "learning_rate": 7.848061427388002e-06,
+      "loss": 0.9287,
+      "step": 5123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9129942491444663,
+      "learning_rate": 7.847209463364574e-06,
+      "loss": 0.7384,
+      "step": 5124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9791587192311715,
+      "learning_rate": 7.846357376990213e-06,
+      "loss": 0.7858,
+      "step": 5125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.114017585454991,
+      "learning_rate": 7.845505168301533e-06,
+      "loss": 0.7141,
+      "step": 5126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.585334740773491,
+      "learning_rate": 7.844652837335162e-06,
+      "loss": 0.7105,
+      "step": 5127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.351543974026045,
+      "learning_rate": 7.843800384127721e-06,
+      "loss": 0.7542,
+      "step": 5128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3557278863330295,
+      "learning_rate": 7.842947808715848e-06,
+      "loss": 0.6789,
+      "step": 5129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7408728800401256,
+      "learning_rate": 7.842095111136173e-06,
+      "loss": 1.0276,
+      "step": 5130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0925893574218204,
+      "learning_rate": 7.841242291425342e-06,
+      "loss": 0.8685,
+      "step": 5131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9466860223937956,
+      "learning_rate": 7.840389349620002e-06,
+      "loss": 0.9428,
+      "step": 5132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0038753259303055,
+      "learning_rate": 7.839536285756804e-06,
+      "loss": 0.5805,
+      "step": 5133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.359900454063397,
+      "learning_rate": 7.83868309987241e-06,
+      "loss": 0.8162,
+      "step": 5134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1158583976875693,
+      "learning_rate": 7.837829792003479e-06,
+      "loss": 0.6999,
+      "step": 5135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.571414781971994,
+      "learning_rate": 7.836976362186684e-06,
+      "loss": 0.9202,
+      "step": 5136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4446700577884477,
+      "learning_rate": 7.836122810458696e-06,
+      "loss": 0.682,
+      "step": 5137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.51304164752676,
+      "learning_rate": 7.835269136856194e-06,
+      "loss": 0.9414,
+      "step": 5138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0813053138056692,
+      "learning_rate": 7.834415341415862e-06,
+      "loss": 0.6245,
+      "step": 5139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4691465964675516,
+      "learning_rate": 7.833561424174392e-06,
+      "loss": 0.8499,
+      "step": 5140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.016926533655261,
+      "learning_rate": 7.832707385168474e-06,
+      "loss": 0.7213,
+      "step": 5141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4424720015412418,
+      "learning_rate": 7.831853224434815e-06,
+      "loss": 0.7335,
+      "step": 5142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.262355194254056,
+      "learning_rate": 7.830998942010113e-06,
+      "loss": 0.7033,
+      "step": 5143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.587131292132497,
+      "learning_rate": 7.830144537931082e-06,
+      "loss": 0.7959,
+      "step": 5144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.105356460831295,
+      "learning_rate": 7.829290012234438e-06,
+      "loss": 0.8014,
+      "step": 5145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1130514022708935,
+      "learning_rate": 7.8284353649569e-06,
+      "loss": 0.8984,
+      "step": 5146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.302612232406404,
+      "learning_rate": 7.827580596135196e-06,
+      "loss": 0.636,
+      "step": 5147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5262118864136167,
+      "learning_rate": 7.826725705806056e-06,
+      "loss": 0.7702,
+      "step": 5148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2576501267855174,
+      "learning_rate": 7.825870694006217e-06,
+      "loss": 0.7023,
+      "step": 5149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5178594139467416,
+      "learning_rate": 7.82501556077242e-06,
+      "loss": 0.7295,
+      "step": 5150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3506411056200287,
+      "learning_rate": 7.824160306141413e-06,
+      "loss": 0.7285,
+      "step": 5151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2316649670190243,
+      "learning_rate": 7.823304930149949e-06,
+      "loss": 0.7398,
+      "step": 5152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.037015540077958,
+      "learning_rate": 7.822449432834784e-06,
+      "loss": 0.9592,
+      "step": 5153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.765080354405162,
+      "learning_rate": 7.82159381423268e-06,
+      "loss": 0.7755,
+      "step": 5154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8803280483548486,
+      "learning_rate": 7.820738074380404e-06,
+      "loss": 0.9044,
+      "step": 5155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2315427486536787,
+      "learning_rate": 7.819882213314736e-06,
+      "loss": 0.8924,
+      "step": 5156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.513119964907687,
+      "learning_rate": 7.819026231072445e-06,
+      "loss": 1.0176,
+      "step": 5157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.856221964445883,
+      "learning_rate": 7.81817012769032e-06,
+      "loss": 0.7118,
+      "step": 5158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.831711003727446,
+      "learning_rate": 7.817313903205148e-06,
+      "loss": 0.7922,
+      "step": 5159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.906830484678884,
+      "learning_rate": 7.816457557653722e-06,
+      "loss": 0.8039,
+      "step": 5160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.360209633520047,
+      "learning_rate": 7.815601091072845e-06,
+      "loss": 0.8017,
+      "step": 5161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.628542631565887,
+      "learning_rate": 7.814744503499317e-06,
+      "loss": 0.9261,
+      "step": 5162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6538852221331883,
+      "learning_rate": 7.813887794969948e-06,
+      "loss": 0.8647,
+      "step": 5163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1851086220412292,
+      "learning_rate": 7.813030965521554e-06,
+      "loss": 0.7491,
+      "step": 5164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8620055749265654,
+      "learning_rate": 7.812174015190955e-06,
+      "loss": 0.6525,
+      "step": 5165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2654089858329938,
+      "learning_rate": 7.811316944014974e-06,
+      "loss": 0.7357,
+      "step": 5166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8723311524469999,
+      "learning_rate": 7.810459752030444e-06,
+      "loss": 0.7092,
+      "step": 5167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.150805409009453,
+      "learning_rate": 7.809602439274198e-06,
+      "loss": 0.8686,
+      "step": 5168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.363525165464555,
+      "learning_rate": 7.80874500578308e-06,
+      "loss": 0.8508,
+      "step": 5169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0190104915099374,
+      "learning_rate": 7.807887451593932e-06,
+      "loss": 1.0197,
+      "step": 5170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5830552005955574,
+      "learning_rate": 7.807029776743608e-06,
+      "loss": 0.9757,
+      "step": 5171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.251585268396228,
+      "learning_rate": 7.806171981268962e-06,
+      "loss": 0.74,
+      "step": 5172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8528499359596515,
+      "learning_rate": 7.805314065206857e-06,
+      "loss": 0.8548,
+      "step": 5173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.071245247843414,
+      "learning_rate": 7.804456028594158e-06,
+      "loss": 0.8378,
+      "step": 5174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0475669590241306,
+      "learning_rate": 7.803597871467738e-06,
+      "loss": 0.7651,
+      "step": 5175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1228940437267576,
+      "learning_rate": 7.802739593864475e-06,
+      "loss": 0.7781,
+      "step": 5176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1444442520237537,
+      "learning_rate": 7.801881195821246e-06,
+      "loss": 0.677,
+      "step": 5177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.209149837258506,
+      "learning_rate": 7.801022677374945e-06,
+      "loss": 1.1006,
+      "step": 5178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1490894103536875,
+      "learning_rate": 7.80016403856246e-06,
+      "loss": 0.8685,
+      "step": 5179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9717083296971447,
+      "learning_rate": 7.799305279420691e-06,
+      "loss": 0.8819,
+      "step": 5180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0919459025966853,
+      "learning_rate": 7.79844639998654e-06,
+      "loss": 0.8716,
+      "step": 5181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0360096421945943,
+      "learning_rate": 7.797587400296913e-06,
+      "loss": 1.1224,
+      "step": 5182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.603031020266383,
+      "learning_rate": 7.796728280388725e-06,
+      "loss": 0.9239,
+      "step": 5183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0577820762537404,
+      "learning_rate": 7.795869040298895e-06,
+      "loss": 0.6367,
+      "step": 5184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.336683118647983,
+      "learning_rate": 7.795009680064344e-06,
+      "loss": 0.9808,
+      "step": 5185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.098466578492551,
+      "learning_rate": 7.794150199722004e-06,
+      "loss": 1.0042,
+      "step": 5186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9582115091891081,
+      "learning_rate": 7.793290599308807e-06,
+      "loss": 0.9161,
+      "step": 5187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.33912020820875,
+      "learning_rate": 7.792430878861691e-06,
+      "loss": 0.77,
+      "step": 5188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.234895992309908,
+      "learning_rate": 7.791571038417602e-06,
+      "loss": 0.9802,
+      "step": 5189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9079746391591816,
+      "learning_rate": 7.79071107801349e-06,
+      "loss": 0.9679,
+      "step": 5190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3516066220490228,
+      "learning_rate": 7.789850997686303e-06,
+      "loss": 0.6863,
+      "step": 5191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3840830860182773,
+      "learning_rate": 7.78899079747301e-06,
+      "loss": 0.6543,
+      "step": 5192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.768364866348871,
+      "learning_rate": 7.788130477410567e-06,
+      "loss": 0.9073,
+      "step": 5193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7392689530790848,
+      "learning_rate": 7.78727003753595e-06,
+      "loss": 0.9551,
+      "step": 5194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.388719797559871,
+      "learning_rate": 7.786409477886133e-06,
+      "loss": 0.6039,
+      "step": 5195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2122614556523854,
+      "learning_rate": 7.785548798498093e-06,
+      "loss": 0.7251,
+      "step": 5196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.169550423141791,
+      "learning_rate": 7.784687999408817e-06,
+      "loss": 0.968,
+      "step": 5197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1212491580861472,
+      "learning_rate": 7.7838270806553e-06,
+      "loss": 0.6428,
+      "step": 5198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6453226877625693,
+      "learning_rate": 7.782966042274529e-06,
+      "loss": 0.8547,
+      "step": 5199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8804253064324448,
+      "learning_rate": 7.782104884303512e-06,
+      "loss": 0.838,
+      "step": 5200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.147308556959783,
+      "learning_rate": 7.78124360677925e-06,
+      "loss": 0.8462,
+      "step": 5201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.292934227463149,
+      "learning_rate": 7.780382209738756e-06,
+      "loss": 0.791,
+      "step": 5202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.912064861180169,
+      "learning_rate": 7.779520693219047e-06,
+      "loss": 0.9343,
+      "step": 5203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.125050720160777,
+      "learning_rate": 7.778659057257144e-06,
+      "loss": 0.9559,
+      "step": 5204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1904288583979574,
+      "learning_rate": 7.777797301890073e-06,
+      "loss": 0.7686,
+      "step": 5205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8208665483980533,
+      "learning_rate": 7.776935427154864e-06,
+      "loss": 0.7258,
+      "step": 5206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3651159916437137,
+      "learning_rate": 7.776073433088555e-06,
+      "loss": 1.0185,
+      "step": 5207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.164063247617283,
+      "learning_rate": 7.775211319728191e-06,
+      "loss": 0.7913,
+      "step": 5208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0511026781031845,
+      "learning_rate": 7.774349087110813e-06,
+      "loss": 0.8427,
+      "step": 5209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.856128181988273,
+      "learning_rate": 7.773486735273476e-06,
+      "loss": 0.8274,
+      "step": 5210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9468185631993853,
+      "learning_rate": 7.772624264253237e-06,
+      "loss": 0.774,
+      "step": 5211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7345738074824606,
+      "learning_rate": 7.77176167408716e-06,
+      "loss": 0.9432,
+      "step": 5212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2618391912499023,
+      "learning_rate": 7.77089896481231e-06,
+      "loss": 0.7447,
+      "step": 5213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.969587399264163,
+      "learning_rate": 7.77003613646576e-06,
+      "loss": 0.9339,
+      "step": 5214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.529744260480344,
+      "learning_rate": 7.769173189084589e-06,
+      "loss": 0.8214,
+      "step": 5215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9724215018454523,
+      "learning_rate": 7.768310122705877e-06,
+      "loss": 0.8307,
+      "step": 5216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8017408447791086,
+      "learning_rate": 7.767446937366716e-06,
+      "loss": 0.7612,
+      "step": 5217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5434888605755632,
+      "learning_rate": 7.766583633104195e-06,
+      "loss": 0.843,
+      "step": 5218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.13367429817688,
+      "learning_rate": 7.765720209955414e-06,
+      "loss": 0.711,
+      "step": 5219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2117633691883563,
+      "learning_rate": 7.764856667957477e-06,
+      "loss": 0.7842,
+      "step": 5220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.152897388137202,
+      "learning_rate": 7.76399300714749e-06,
+      "loss": 0.8699,
+      "step": 5221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7714480950769964,
+      "learning_rate": 7.763129227562568e-06,
+      "loss": 1.0476,
+      "step": 5222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1418190537446637,
+      "learning_rate": 7.762265329239829e-06,
+      "loss": 0.784,
+      "step": 5223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0023793765948494,
+      "learning_rate": 7.761401312216398e-06,
+      "loss": 1.0101,
+      "step": 5224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5660397264845187,
+      "learning_rate": 7.760537176529402e-06,
+      "loss": 0.9137,
+      "step": 5225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8621697177300744,
+      "learning_rate": 7.759672922215976e-06,
+      "loss": 0.7964,
+      "step": 5226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3060603644168947,
+      "learning_rate": 7.758808549313258e-06,
+      "loss": 0.8087,
+      "step": 5227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0324322136958792,
+      "learning_rate": 7.757944057858392e-06,
+      "loss": 0.7968,
+      "step": 5228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9209752026969904,
+      "learning_rate": 7.757079447888529e-06,
+      "loss": 0.9178,
+      "step": 5229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.596846341305376,
+      "learning_rate": 7.75621471944082e-06,
+      "loss": 0.9369,
+      "step": 5230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1094103926256644,
+      "learning_rate": 7.755349872552426e-06,
+      "loss": 0.833,
+      "step": 5231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.238786911335419,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 0.836,
+      "step": 5232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8044342408417569,
+      "learning_rate": 7.753619823602249e-06,
+      "loss": 0.7664,
+      "step": 5233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.137813554082169,
+      "learning_rate": 7.752754621614807e-06,
+      "loss": 0.7456,
+      "step": 5234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.123423231062343,
+      "learning_rate": 7.751889301335369e-06,
+      "loss": 0.8018,
+      "step": 5235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.277391526516686,
+      "learning_rate": 7.75102386280112e-06,
+      "loss": 0.8474,
+      "step": 5236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1655455383286824,
+      "learning_rate": 7.750158306049247e-06,
+      "loss": 1.011,
+      "step": 5237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.32160734782969,
+      "learning_rate": 7.749292631116946e-06,
+      "loss": 0.8038,
+      "step": 5238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.924725180108933,
+      "learning_rate": 7.748426838041421e-06,
+      "loss": 0.7667,
+      "step": 5239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7131829738148077,
+      "learning_rate": 7.747560926859872e-06,
+      "loss": 0.7243,
+      "step": 5240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1928773018860237,
+      "learning_rate": 7.74669489760951e-06,
+      "loss": 0.6853,
+      "step": 5241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8987230394414198,
+      "learning_rate": 7.745828750327551e-06,
+      "loss": 0.8652,
+      "step": 5242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3321131085892812,
+      "learning_rate": 7.744962485051217e-06,
+      "loss": 0.8031,
+      "step": 5243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2524681246920966,
+      "learning_rate": 7.744096101817731e-06,
+      "loss": 0.9989,
+      "step": 5244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8412721614797767,
+      "learning_rate": 7.743229600664322e-06,
+      "loss": 0.9458,
+      "step": 5245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4131058898361113,
+      "learning_rate": 7.742362981628229e-06,
+      "loss": 0.7368,
+      "step": 5246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9052244123890334,
+      "learning_rate": 7.741496244746692e-06,
+      "loss": 0.9998,
+      "step": 5247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.512203510837276,
+      "learning_rate": 7.740629390056955e-06,
+      "loss": 0.8886,
+      "step": 5248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0073707970658563,
+      "learning_rate": 7.73976241759627e-06,
+      "loss": 0.6502,
+      "step": 5249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.542441392888268,
+      "learning_rate": 7.738895327401891e-06,
+      "loss": 0.7582,
+      "step": 5250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1045497382734637,
+      "learning_rate": 7.738028119511081e-06,
+      "loss": 0.8519,
+      "step": 5251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4034580021612544,
+      "learning_rate": 7.737160793961103e-06,
+      "loss": 0.7772,
+      "step": 5252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.343908562441648,
+      "learning_rate": 7.736293350789232e-06,
+      "loss": 0.8986,
+      "step": 5253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.18444997691811,
+      "learning_rate": 7.73542579003274e-06,
+      "loss": 0.8992,
+      "step": 5254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.231078823864015,
+      "learning_rate": 7.734558111728911e-06,
+      "loss": 0.6701,
+      "step": 5255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2217364986087262,
+      "learning_rate": 7.733690315915028e-06,
+      "loss": 0.8129,
+      "step": 5256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9558914843783077,
+      "learning_rate": 7.732822402628385e-06,
+      "loss": 0.8247,
+      "step": 5257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3166093330126896,
+      "learning_rate": 7.731954371906279e-06,
+      "loss": 0.9332,
+      "step": 5258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2896600897754626,
+      "learning_rate": 7.731086223786006e-06,
+      "loss": 0.936,
+      "step": 5259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0451331093866285,
+      "learning_rate": 7.730217958304876e-06,
+      "loss": 0.6684,
+      "step": 5260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.347014760859459,
+      "learning_rate": 7.729349575500202e-06,
+      "loss": 0.9095,
+      "step": 5261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9298751130037102,
+      "learning_rate": 7.728481075409297e-06,
+      "loss": 0.9995,
+      "step": 5262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1679050559351976,
+      "learning_rate": 7.727612458069482e-06,
+      "loss": 0.682,
+      "step": 5263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9196814805573792,
+      "learning_rate": 7.726743723518087e-06,
+      "loss": 0.7442,
+      "step": 5264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1093580791840973,
+      "learning_rate": 7.72587487179244e-06,
+      "loss": 0.5415,
+      "step": 5265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9999446673768029,
+      "learning_rate": 7.72500590292988e-06,
+      "loss": 0.761,
+      "step": 5266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.068325706155967,
+      "learning_rate": 7.724136816967746e-06,
+      "loss": 0.7202,
+      "step": 5267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2200122163275364,
+      "learning_rate": 7.723267613943388e-06,
+      "loss": 0.9223,
+      "step": 5268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0644503814130482,
+      "learning_rate": 7.722398293894153e-06,
+      "loss": 0.6943,
+      "step": 5269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.676915722550001,
+      "learning_rate": 7.721528856857403e-06,
+      "loss": 0.8767,
+      "step": 5270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.063945879172518,
+      "learning_rate": 7.720659302870496e-06,
+      "loss": 0.8774,
+      "step": 5271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1802259797785724,
+      "learning_rate": 7.719789631970798e-06,
+      "loss": 0.7341,
+      "step": 5272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7814331912149766,
+      "learning_rate": 7.718919844195685e-06,
+      "loss": 0.8731,
+      "step": 5273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2815224529533205,
+      "learning_rate": 7.718049939582529e-06,
+      "loss": 0.8212,
+      "step": 5274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1277157813843859,
+      "learning_rate": 7.717179918168713e-06,
+      "loss": 0.6238,
+      "step": 5275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1511976644229982,
+      "learning_rate": 7.716309779991625e-06,
+      "loss": 0.7637,
+      "step": 5276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.452624557735539,
+      "learning_rate": 7.715439525088657e-06,
+      "loss": 0.9087,
+      "step": 5277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1369278608034756,
+      "learning_rate": 7.714569153497204e-06,
+      "loss": 0.7838,
+      "step": 5278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1389222555121186,
+      "learning_rate": 7.713698665254669e-06,
+      "loss": 0.9097,
+      "step": 5279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1583030206977374,
+      "learning_rate": 7.71282806039846e-06,
+      "loss": 0.8053,
+      "step": 5280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1871304814080443,
+      "learning_rate": 7.711957338965986e-06,
+      "loss": 0.7737,
+      "step": 5281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5481165287286083,
+      "learning_rate": 7.711086500994667e-06,
+      "loss": 0.9152,
+      "step": 5282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5027283669308744,
+      "learning_rate": 7.710215546521921e-06,
+      "loss": 0.8075,
+      "step": 5283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2327051696725015,
+      "learning_rate": 7.70934447558518e-06,
+      "loss": 0.9825,
+      "step": 5284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7368187186416733,
+      "learning_rate": 7.708473288221868e-06,
+      "loss": 0.7597,
+      "step": 5285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6869079968108296,
+      "learning_rate": 7.70760198446943e-06,
+      "loss": 0.8424,
+      "step": 5286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.851042657024052,
+      "learning_rate": 7.706730564365304e-06,
+      "loss": 0.7732,
+      "step": 5287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2348428990702205,
+      "learning_rate": 7.705859027946939e-06,
+      "loss": 1.0007,
+      "step": 5288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1820300917730795,
+      "learning_rate": 7.704987375251782e-06,
+      "loss": 0.8214,
+      "step": 5289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8435197098841047,
+      "learning_rate": 7.704115606317296e-06,
+      "loss": 0.9305,
+      "step": 5290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1674313604992843,
+      "learning_rate": 7.703243721180938e-06,
+      "loss": 0.8151,
+      "step": 5291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4216283811323012,
+      "learning_rate": 7.702371719880178e-06,
+      "loss": 0.7536,
+      "step": 5292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2854033649280816,
+      "learning_rate": 7.701499602452487e-06,
+      "loss": 0.7509,
+      "step": 5293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1106462828019965,
+      "learning_rate": 7.70062736893534e-06,
+      "loss": 0.6843,
+      "step": 5294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2480455450065147,
+      "learning_rate": 7.699755019366221e-06,
+      "loss": 0.6715,
+      "step": 5295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.696576862311743,
+      "learning_rate": 7.698882553782617e-06,
+      "loss": 0.8834,
+      "step": 5296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9671869213708615,
+      "learning_rate": 7.698009972222017e-06,
+      "loss": 0.7434,
+      "step": 5297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4675182647608067,
+      "learning_rate": 7.697137274721922e-06,
+      "loss": 0.7976,
+      "step": 5298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2000364830436783,
+      "learning_rate": 7.696264461319831e-06,
+      "loss": 0.9281,
+      "step": 5299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2473525554171614,
+      "learning_rate": 7.69539153205325e-06,
+      "loss": 1.0209,
+      "step": 5300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.072669087949397,
+      "learning_rate": 7.694518486959695e-06,
+      "loss": 0.8938,
+      "step": 5301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.25991301864628,
+      "learning_rate": 7.693645326076677e-06,
+      "loss": 0.8164,
+      "step": 5302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.184257637862257,
+      "learning_rate": 7.692772049441718e-06,
+      "loss": 0.8141,
+      "step": 5303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.089601975049657,
+      "learning_rate": 7.69189865709235e-06,
+      "loss": 0.5891,
+      "step": 5304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1366357182833364,
+      "learning_rate": 7.691025149066103e-06,
+      "loss": 0.6732,
+      "step": 5305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3549952393407834,
+      "learning_rate": 7.69015152540051e-06,
+      "loss": 0.7836,
+      "step": 5306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.00713834290777,
+      "learning_rate": 7.689277786133113e-06,
+      "loss": 1.0053,
+      "step": 5307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.641339223947017,
+      "learning_rate": 7.688403931301462e-06,
+      "loss": 0.9788,
+      "step": 5308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.331762587206155,
+      "learning_rate": 7.687529960943107e-06,
+      "loss": 0.8331,
+      "step": 5309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9104340862773477,
+      "learning_rate": 7.686655875095603e-06,
+      "loss": 0.8212,
+      "step": 5310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5823315686465156,
+      "learning_rate": 7.685781673796515e-06,
+      "loss": 0.8242,
+      "step": 5311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0438382777371116,
+      "learning_rate": 7.684907357083404e-06,
+      "loss": 0.951,
+      "step": 5312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0725403593799254,
+      "learning_rate": 7.684032924993845e-06,
+      "loss": 0.8993,
+      "step": 5313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.022956789630063,
+      "learning_rate": 7.683158377565415e-06,
+      "loss": 0.811,
+      "step": 5314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7790714421784557,
+      "learning_rate": 7.682283714835691e-06,
+      "loss": 0.8776,
+      "step": 5315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3835696839049034,
+      "learning_rate": 7.681408936842266e-06,
+      "loss": 0.7269,
+      "step": 5316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1038739844076282,
+      "learning_rate": 7.680534043622725e-06,
+      "loss": 0.7009,
+      "step": 5317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.21433856803231,
+      "learning_rate": 7.679659035214666e-06,
+      "loss": 0.7141,
+      "step": 5318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.744020159140638,
+      "learning_rate": 7.678783911655691e-06,
+      "loss": 0.8745,
+      "step": 5319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5787320219272547,
+      "learning_rate": 7.677908672983404e-06,
+      "loss": 0.9681,
+      "step": 5320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.950231848335508,
+      "learning_rate": 7.677033319235418e-06,
+      "loss": 0.7529,
+      "step": 5321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.252659378596618,
+      "learning_rate": 7.676157850449348e-06,
+      "loss": 0.8834,
+      "step": 5322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.495367004845683,
+      "learning_rate": 7.675282266662816e-06,
+      "loss": 0.8436,
+      "step": 5323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.890946690719468,
+      "learning_rate": 7.674406567913447e-06,
+      "loss": 0.8064,
+      "step": 5324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1281440015378426,
+      "learning_rate": 7.67353075423887e-06,
+      "loss": 0.6007,
+      "step": 5325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9458904239143966,
+      "learning_rate": 7.672654825676724e-06,
+      "loss": 0.8377,
+      "step": 5326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0540945368261943,
+      "learning_rate": 7.671778782264647e-06,
+      "loss": 0.7539,
+      "step": 5327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1731027855788554,
+      "learning_rate": 7.670902624040286e-06,
+      "loss": 0.8135,
+      "step": 5328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.033148456913297,
+      "learning_rate": 7.67002635104129e-06,
+      "loss": 0.6587,
+      "step": 5329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.167907793487384,
+      "learning_rate": 7.669149963305314e-06,
+      "loss": 0.9464,
+      "step": 5330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.526667722655177,
+      "learning_rate": 7.668273460870022e-06,
+      "loss": 0.659,
+      "step": 5331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.162222853640468,
+      "learning_rate": 7.667396843773074e-06,
+      "loss": 0.6683,
+      "step": 5332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2341464411883227,
+      "learning_rate": 7.666520112052145e-06,
+      "loss": 0.7401,
+      "step": 5333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.045235475124907,
+      "learning_rate": 7.66564326574491e-06,
+      "loss": 0.9751,
+      "step": 5334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.83129592515627,
+      "learning_rate": 7.664766304889044e-06,
+      "loss": 0.8018,
+      "step": 5335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0610084655971264,
+      "learning_rate": 7.663889229522235e-06,
+      "loss": 0.7926,
+      "step": 5336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.683974357746062,
+      "learning_rate": 7.663012039682174e-06,
+      "loss": 0.812,
+      "step": 5337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.050479282116955,
+      "learning_rate": 7.662134735406555e-06,
+      "loss": 0.8215,
+      "step": 5338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1606859539341126,
+      "learning_rate": 7.661257316733078e-06,
+      "loss": 0.7786,
+      "step": 5339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9405560510000566,
+      "learning_rate": 7.660379783699446e-06,
+      "loss": 0.7946,
+      "step": 5340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.724903901883774,
+      "learning_rate": 7.65950213634337e-06,
+      "loss": 0.9007,
+      "step": 5341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2857133047532074,
+      "learning_rate": 7.658624374702565e-06,
+      "loss": 0.8408,
+      "step": 5342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2242274329014522,
+      "learning_rate": 7.657746498814749e-06,
+      "loss": 0.6587,
+      "step": 5343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0641512533304214,
+      "learning_rate": 7.656868508717648e-06,
+      "loss": 0.7687,
+      "step": 5344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2071291561451574,
+      "learning_rate": 7.655990404448988e-06,
+      "loss": 0.9933,
+      "step": 5345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9912731953453622,
+      "learning_rate": 7.655112186046508e-06,
+      "loss": 0.8295,
+      "step": 5346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4533049675865315,
+      "learning_rate": 7.65423385354794e-06,
+      "loss": 0.7372,
+      "step": 5347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3100462850880894,
+      "learning_rate": 7.653355406991034e-06,
+      "loss": 0.8373,
+      "step": 5348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3934618007267177,
+      "learning_rate": 7.652476846413537e-06,
+      "loss": 0.8307,
+      "step": 5349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2590443666444613,
+      "learning_rate": 7.651598171853203e-06,
+      "loss": 0.8731,
+      "step": 5350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.142602038635085,
+      "learning_rate": 7.65071938334779e-06,
+      "loss": 0.9125,
+      "step": 5351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4303011190438,
+      "learning_rate": 7.64984048093506e-06,
+      "loss": 0.6489,
+      "step": 5352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.081440310099479,
+      "learning_rate": 7.648961464652784e-06,
+      "loss": 0.7421,
+      "step": 5353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1372147827793513,
+      "learning_rate": 7.648082334538735e-06,
+      "loss": 0.617,
+      "step": 5354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5494447875190085,
+      "learning_rate": 7.64720309063069e-06,
+      "loss": 0.6303,
+      "step": 5355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.39074196096392,
+      "learning_rate": 7.646323732966433e-06,
+      "loss": 0.7812,
+      "step": 5356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0337757965477157,
+      "learning_rate": 7.645444261583752e-06,
+      "loss": 0.8397,
+      "step": 5357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.491494053427519,
+      "learning_rate": 7.644564676520439e-06,
+      "loss": 0.768,
+      "step": 5358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0399009660825285,
+      "learning_rate": 7.64368497781429e-06,
+      "loss": 0.9623,
+      "step": 5359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3793859622748887,
+      "learning_rate": 7.642805165503112e-06,
+      "loss": 0.8543,
+      "step": 5360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0164990318503633,
+      "learning_rate": 7.641925239624708e-06,
+      "loss": 0.6633,
+      "step": 5361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1811033273809632,
+      "learning_rate": 7.641045200216896e-06,
+      "loss": 0.7774,
+      "step": 5362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1842654045073342,
+      "learning_rate": 7.640165047317488e-06,
+      "loss": 0.6417,
+      "step": 5363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7558514066428543,
+      "learning_rate": 7.639284780964307e-06,
+      "loss": 0.884,
+      "step": 5364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.44624267581796,
+      "learning_rate": 7.638404401195182e-06,
+      "loss": 0.8493,
+      "step": 5365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0649675247056316,
+      "learning_rate": 7.637523908047943e-06,
+      "loss": 0.6893,
+      "step": 5366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.92002785002901,
+      "learning_rate": 7.63664330156043e-06,
+      "loss": 0.8858,
+      "step": 5367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.595059185796166,
+      "learning_rate": 7.635762581770478e-06,
+      "loss": 0.7785,
+      "step": 5368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7170538550564847,
+      "learning_rate": 7.634881748715941e-06,
+      "loss": 0.7612,
+      "step": 5369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2191144977734694,
+      "learning_rate": 7.634000802434667e-06,
+      "loss": 0.7672,
+      "step": 5370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.992091148333544,
+      "learning_rate": 7.633119742964511e-06,
+      "loss": 0.903,
+      "step": 5371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.080178348497653,
+      "learning_rate": 7.632238570343335e-06,
+      "loss": 0.6674,
+      "step": 5372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.143727459422405,
+      "learning_rate": 7.631357284609006e-06,
+      "loss": 0.7094,
+      "step": 5373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5546391363358003,
+      "learning_rate": 7.630475885799395e-06,
+      "loss": 0.7184,
+      "step": 5374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9496003121159546,
+      "learning_rate": 7.629594373952375e-06,
+      "loss": 0.977,
+      "step": 5375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.161040062223872,
+      "learning_rate": 7.628712749105831e-06,
+      "loss": 0.909,
+      "step": 5376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.118613795940489,
+      "learning_rate": 7.627831011297644e-06,
+      "loss": 0.7706,
+      "step": 5377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.297062663043031,
+      "learning_rate": 7.626949160565706e-06,
+      "loss": 0.8334,
+      "step": 5378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.983568610836711,
+      "learning_rate": 7.626067196947913e-06,
+      "loss": 0.7833,
+      "step": 5379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.031972982695026,
+      "learning_rate": 7.625185120482162e-06,
+      "loss": 0.6567,
+      "step": 5380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5185619555923795,
+      "learning_rate": 7.624302931206362e-06,
+      "loss": 0.9528,
+      "step": 5381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6842631124982888,
+      "learning_rate": 7.623420629158418e-06,
+      "loss": 0.7792,
+      "step": 5382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.450557913502267,
+      "learning_rate": 7.622538214376248e-06,
+      "loss": 0.7445,
+      "step": 5383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3528899268894183,
+      "learning_rate": 7.621655686897771e-06,
+      "loss": 0.8537,
+      "step": 5384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9527794862171042,
+      "learning_rate": 7.62077304676091e-06,
+      "loss": 0.9684,
+      "step": 5385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2144281743126728,
+      "learning_rate": 7.619890294003593e-06,
+      "loss": 0.7381,
+      "step": 5386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9578718895363133,
+      "learning_rate": 7.6190074286637575e-06,
+      "loss": 0.5935,
+      "step": 5387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.448017515986289,
+      "learning_rate": 7.618124450779338e-06,
+      "loss": 0.7518,
+      "step": 5388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5705017014892384,
+      "learning_rate": 7.617241360388282e-06,
+      "loss": 0.8289,
+      "step": 5389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1597763251856006,
+      "learning_rate": 7.616358157528535e-06,
+      "loss": 0.786,
+      "step": 5390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.169811282656929,
+      "learning_rate": 7.615474842238051e-06,
+      "loss": 0.6836,
+      "step": 5391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.904059147825222,
+      "learning_rate": 7.614591414554787e-06,
+      "loss": 0.8729,
+      "step": 5392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.100101550734466,
+      "learning_rate": 7.613707874516708e-06,
+      "loss": 0.625,
+      "step": 5393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2802920907501818,
+      "learning_rate": 7.612824222161781e-06,
+      "loss": 0.7379,
+      "step": 5394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.46659413019773,
+      "learning_rate": 7.611940457527976e-06,
+      "loss": 1.0221,
+      "step": 5395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.929223599271201,
+      "learning_rate": 7.6110565806532745e-06,
+      "loss": 0.8496,
+      "step": 5396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9438074695676308,
+      "learning_rate": 7.610172591575656e-06,
+      "loss": 0.769,
+      "step": 5397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.880823980650164,
+      "learning_rate": 7.609288490333109e-06,
+      "loss": 0.5966,
+      "step": 5398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.540876123958633,
+      "learning_rate": 7.608404276963623e-06,
+      "loss": 0.831,
+      "step": 5399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3772693629383315,
+      "learning_rate": 7.6075199515051955e-06,
+      "loss": 0.6375,
+      "step": 5400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6927097022149025,
+      "learning_rate": 7.60663551399583e-06,
+      "loss": 0.8466,
+      "step": 5401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7137741717965835,
+      "learning_rate": 7.605750964473531e-06,
+      "loss": 0.893,
+      "step": 5402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6645340313184245,
+      "learning_rate": 7.604866302976309e-06,
+      "loss": 0.4597,
+      "step": 5403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1144533745156116,
+      "learning_rate": 7.60398152954218e-06,
+      "loss": 0.8416,
+      "step": 5404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.04945165145335,
+      "learning_rate": 7.6030966442091665e-06,
+      "loss": 0.8661,
+      "step": 5405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4263984011842163,
+      "learning_rate": 7.602211647015292e-06,
+      "loss": 0.9123,
+      "step": 5406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0619221945310233,
+      "learning_rate": 7.601326537998589e-06,
+      "loss": 0.7447,
+      "step": 5407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0977738735664477,
+      "learning_rate": 7.60044131719709e-06,
+      "loss": 0.6326,
+      "step": 5408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5484964605592957,
+      "learning_rate": 7.599555984648836e-06,
+      "loss": 0.8797,
+      "step": 5409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.913216674976656,
+      "learning_rate": 7.598670540391872e-06,
+      "loss": 0.8587,
+      "step": 5410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0046260015543615,
+      "learning_rate": 7.597784984464248e-06,
+      "loss": 0.8257,
+      "step": 5411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.725583150553362,
+      "learning_rate": 7.596899316904018e-06,
+      "loss": 0.8056,
+      "step": 5412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8205753627421557,
+      "learning_rate": 7.596013537749238e-06,
+      "loss": 0.7748,
+      "step": 5413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.952738570186866,
+      "learning_rate": 7.595127647037976e-06,
+      "loss": 0.734,
+      "step": 5414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1973401606348912,
+      "learning_rate": 7.5942416448083e-06,
+      "loss": 0.826,
+      "step": 5415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.780051925733984,
+      "learning_rate": 7.593355531098284e-06,
+      "loss": 0.6998,
+      "step": 5416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1306319557418,
+      "learning_rate": 7.592469305946003e-06,
+      "loss": 0.7397,
+      "step": 5417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0814188255212844,
+      "learning_rate": 7.5915829693895435e-06,
+      "loss": 0.7709,
+      "step": 5418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.209687438139963,
+      "learning_rate": 7.590696521466992e-06,
+      "loss": 0.9114,
+      "step": 5419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1001201989506817,
+      "learning_rate": 7.589809962216439e-06,
+      "loss": 1.0397,
+      "step": 5420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0919749954494393,
+      "learning_rate": 7.588923291675987e-06,
+      "loss": 0.8514,
+      "step": 5421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8169844012906,
+      "learning_rate": 7.588036509883734e-06,
+      "loss": 0.798,
+      "step": 5422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.434911477356655,
+      "learning_rate": 7.587149616877789e-06,
+      "loss": 0.7786,
+      "step": 5423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3255042447400953,
+      "learning_rate": 7.586262612696263e-06,
+      "loss": 0.774,
+      "step": 5424
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.244831477187899,
+      "learning_rate": 7.585375497377271e-06,
+      "loss": 0.6733,
+      "step": 5425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.777534476935265,
+      "learning_rate": 7.584488270958937e-06,
+      "loss": 0.8897,
+      "step": 5426
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0319645014190857,
+      "learning_rate": 7.583600933479387e-06,
+      "loss": 0.7004,
+      "step": 5427
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8720735982229557,
+      "learning_rate": 7.582713484976749e-06,
+      "loss": 0.6901,
+      "step": 5428
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6439321688443114,
+      "learning_rate": 7.5818259254891614e-06,
+      "loss": 0.8521,
+      "step": 5429
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.002579469501395,
+      "learning_rate": 7.580938255054765e-06,
+      "loss": 0.946,
+      "step": 5430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2616908392936663,
+      "learning_rate": 7.580050473711702e-06,
+      "loss": 0.7748,
+      "step": 5431
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.21818915023055,
+      "learning_rate": 7.579162581498125e-06,
+      "loss": 0.7556,
+      "step": 5432
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7846138405418825,
+      "learning_rate": 7.5782745784521875e-06,
+      "loss": 0.9569,
+      "step": 5433
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.599503575076504,
+      "learning_rate": 7.577386464612049e-06,
+      "loss": 0.8126,
+      "step": 5434
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.257235638775955,
+      "learning_rate": 7.576498240015875e-06,
+      "loss": 0.6639,
+      "step": 5435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.118854015686393,
+      "learning_rate": 7.5756099047018315e-06,
+      "loss": 0.6748,
+      "step": 5436
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8846950081425555,
+      "learning_rate": 7.574721458708096e-06,
+      "loss": 0.8348,
+      "step": 5437
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.324104842237872,
+      "learning_rate": 7.573832902072844e-06,
+      "loss": 0.8784,
+      "step": 5438
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2564115354303658,
+      "learning_rate": 7.572944234834261e-06,
+      "loss": 0.9515,
+      "step": 5439
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.564620013129591,
+      "learning_rate": 7.572055457030533e-06,
+      "loss": 0.8154,
+      "step": 5440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1728490925415214,
+      "learning_rate": 7.5711665686998545e-06,
+      "loss": 1.0971,
+      "step": 5441
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5904765723083067,
+      "learning_rate": 7.570277569880422e-06,
+      "loss": 0.8078,
+      "step": 5442
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7741534299575739,
+      "learning_rate": 7.569388460610437e-06,
+      "loss": 0.9855,
+      "step": 5443
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6120098837275108,
+      "learning_rate": 7.568499240928109e-06,
+      "loss": 0.7112,
+      "step": 5444
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9553512083421234,
+      "learning_rate": 7.5676099108716485e-06,
+      "loss": 0.8486,
+      "step": 5445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8987765505903453,
+      "learning_rate": 7.5667204704792706e-06,
+      "loss": 0.7525,
+      "step": 5446
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.234644695856194,
+      "learning_rate": 7.565830919789198e-06,
+      "loss": 1.0896,
+      "step": 5447
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.196062717089123,
+      "learning_rate": 7.564941258839655e-06,
+      "loss": 0.7001,
+      "step": 5448
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8991739190008878,
+      "learning_rate": 7.5640514876688765e-06,
+      "loss": 1.0026,
+      "step": 5449
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.044689527654753,
+      "learning_rate": 7.563161606315095e-06,
+      "loss": 0.9187,
+      "step": 5450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0978720871688905,
+      "learning_rate": 7.5622716148165475e-06,
+      "loss": 0.789,
+      "step": 5451
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.008986022230485,
+      "learning_rate": 7.561381513211486e-06,
+      "loss": 0.6269,
+      "step": 5452
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8277211951720589,
+      "learning_rate": 7.5604913015381535e-06,
+      "loss": 1.1124,
+      "step": 5453
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7437237227545457,
+      "learning_rate": 7.559600979834809e-06,
+      "loss": 0.7967,
+      "step": 5454
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3078990042325875,
+      "learning_rate": 7.5587105481397095e-06,
+      "loss": 0.888,
+      "step": 5455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2995276326756826,
+      "learning_rate": 7.557820006491118e-06,
+      "loss": 0.7804,
+      "step": 5456
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2041318510608496,
+      "learning_rate": 7.5569293549273046e-06,
+      "loss": 0.9624,
+      "step": 5457
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3636833800313446,
+      "learning_rate": 7.556038593486543e-06,
+      "loss": 0.7716,
+      "step": 5458
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9058486390430232,
+      "learning_rate": 7.555147722207111e-06,
+      "loss": 0.6463,
+      "step": 5459
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9090442133933136,
+      "learning_rate": 7.554256741127291e-06,
+      "loss": 0.8272,
+      "step": 5460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3663166701081537,
+      "learning_rate": 7.553365650285368e-06,
+      "loss": 0.6336,
+      "step": 5461
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.163786273059354,
+      "learning_rate": 7.5524744497196376e-06,
+      "loss": 0.8369,
+      "step": 5462
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.136516864710223,
+      "learning_rate": 7.5515831394683954e-06,
+      "loss": 0.8332,
+      "step": 5463
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1701650379208246,
+      "learning_rate": 7.550691719569944e-06,
+      "loss": 0.7763,
+      "step": 5464
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8809307706217424,
+      "learning_rate": 7.549800190062586e-06,
+      "loss": 0.7532,
+      "step": 5465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.096885489843313,
+      "learning_rate": 7.548908550984637e-06,
+      "loss": 0.7429,
+      "step": 5466
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2898683516653149,
+      "learning_rate": 7.548016802374412e-06,
+      "loss": 0.705,
+      "step": 5467
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.62092823095436,
+      "learning_rate": 7.547124944270229e-06,
+      "loss": 1.0235,
+      "step": 5468
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7625470628926454,
+      "learning_rate": 7.546232976710413e-06,
+      "loss": 0.7708,
+      "step": 5469
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.998127502004544,
+      "learning_rate": 7.545340899733297e-06,
+      "loss": 0.9302,
+      "step": 5470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1292650137939497,
+      "learning_rate": 7.544448713377213e-06,
+      "loss": 1.0268,
+      "step": 5471
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9874725053073712,
+      "learning_rate": 7.543556417680502e-06,
+      "loss": 0.9241,
+      "step": 5472
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1684784693267194,
+      "learning_rate": 7.542664012681507e-06,
+      "loss": 0.683,
+      "step": 5473
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1504171050733587,
+      "learning_rate": 7.541771498418575e-06,
+      "loss": 0.6905,
+      "step": 5474
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0460955654270996,
+      "learning_rate": 7.540878874930062e-06,
+      "loss": 0.8699,
+      "step": 5475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8179066082200752,
+      "learning_rate": 7.539986142254324e-06,
+      "loss": 0.7969,
+      "step": 5476
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9350259332990738,
+      "learning_rate": 7.539093300429727e-06,
+      "loss": 0.7872,
+      "step": 5477
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4008734575932666,
+      "learning_rate": 7.538200349494633e-06,
+      "loss": 0.6993,
+      "step": 5478
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8173394828118576,
+      "learning_rate": 7.537307289487419e-06,
+      "loss": 0.7747,
+      "step": 5479
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5196154643549447,
+      "learning_rate": 7.536414120446458e-06,
+      "loss": 0.7948,
+      "step": 5480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1817259085409106,
+      "learning_rate": 7.535520842410136e-06,
+      "loss": 0.8297,
+      "step": 5481
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5690759225365967,
+      "learning_rate": 7.534627455416835e-06,
+      "loss": 0.7257,
+      "step": 5482
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.060274135080897,
+      "learning_rate": 7.533733959504947e-06,
+      "loss": 0.7872,
+      "step": 5483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8966660303974208,
+      "learning_rate": 7.532840354712868e-06,
+      "loss": 1.0048,
+      "step": 5484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.268492719660553,
+      "learning_rate": 7.531946641078998e-06,
+      "loss": 0.716,
+      "step": 5485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7232627535639145,
+      "learning_rate": 7.531052818641743e-06,
+      "loss": 0.7539,
+      "step": 5486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.073877390130371,
+      "learning_rate": 7.530158887439509e-06,
+      "loss": 0.9757,
+      "step": 5487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0141402265454835,
+      "learning_rate": 7.529264847510715e-06,
+      "loss": 0.5725,
+      "step": 5488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.174175304365285,
+      "learning_rate": 7.5283706988937765e-06,
+      "loss": 0.8104,
+      "step": 5489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2017689777434883,
+      "learning_rate": 7.5274764416271185e-06,
+      "loss": 0.7381,
+      "step": 5490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7024495561524287,
+      "learning_rate": 7.526582075749169e-06,
+      "loss": 0.8852,
+      "step": 5491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8428691496198304,
+      "learning_rate": 7.525687601298361e-06,
+      "loss": 0.6778,
+      "step": 5492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.253536190247792,
+      "learning_rate": 7.5247930183131305e-06,
+      "loss": 0.9519,
+      "step": 5493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.879130251124265,
+      "learning_rate": 7.523898326831921e-06,
+      "loss": 0.8402,
+      "step": 5494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8959836191505925,
+      "learning_rate": 7.52300352689318e-06,
+      "loss": 0.9552,
+      "step": 5495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2266712783467142,
+      "learning_rate": 7.522108618535359e-06,
+      "loss": 0.6803,
+      "step": 5496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9766175461716632,
+      "learning_rate": 7.521213601796913e-06,
+      "loss": 0.7313,
+      "step": 5497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1148159475567643,
+      "learning_rate": 7.520318476716303e-06,
+      "loss": 0.7247,
+      "step": 5498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2236768827559907,
+      "learning_rate": 7.5194232433319955e-06,
+      "loss": 0.7426,
+      "step": 5499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.426439943365704,
+      "learning_rate": 7.51852790168246e-06,
+      "loss": 0.7804,
+      "step": 5500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1031277687321053,
+      "learning_rate": 7.517632451806172e-06,
+      "loss": 0.9104,
+      "step": 5501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.823563216516996,
+      "learning_rate": 7.516736893741611e-06,
+      "loss": 0.8808,
+      "step": 5502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.041129927630248,
+      "learning_rate": 7.515841227527258e-06,
+      "loss": 0.946,
+      "step": 5503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5845866971427944,
+      "learning_rate": 7.514945453201608e-06,
+      "loss": 0.8876,
+      "step": 5504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.005390956132794,
+      "learning_rate": 7.514049570803148e-06,
+      "loss": 0.8004,
+      "step": 5505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7467645858283687,
+      "learning_rate": 7.5131535803703805e-06,
+      "loss": 0.7215,
+      "step": 5506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.068479696315261,
+      "learning_rate": 7.512257481941805e-06,
+      "loss": 0.7476,
+      "step": 5507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2551866405438457,
+      "learning_rate": 7.511361275555931e-06,
+      "loss": 0.6882,
+      "step": 5508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8805308370594047,
+      "learning_rate": 7.510464961251271e-06,
+      "loss": 0.7579,
+      "step": 5509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8602098433486853,
+      "learning_rate": 7.50956853906634e-06,
+      "loss": 0.8396,
+      "step": 5510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8893112635395006,
+      "learning_rate": 7.508672009039659e-06,
+      "loss": 0.5897,
+      "step": 5511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0141380401002733,
+      "learning_rate": 7.507775371209754e-06,
+      "loss": 0.6425,
+      "step": 5512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6845982989328077,
+      "learning_rate": 7.5068786256151585e-06,
+      "loss": 0.9733,
+      "step": 5513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9985353101686218,
+      "learning_rate": 7.505981772294404e-06,
+      "loss": 0.9259,
+      "step": 5514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.677632637529228,
+      "learning_rate": 7.505084811286032e-06,
+      "loss": 0.9762,
+      "step": 5515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.827956322548608,
+      "learning_rate": 7.5041877426285856e-06,
+      "loss": 0.8608,
+      "step": 5516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1940822819360402,
+      "learning_rate": 7.503290566360615e-06,
+      "loss": 0.6667,
+      "step": 5517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0785937475913248,
+      "learning_rate": 7.502393282520675e-06,
+      "loss": 0.5753,
+      "step": 5518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1474935378536966,
+      "learning_rate": 7.501495891147322e-06,
+      "loss": 0.755,
+      "step": 5519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2492168191589594,
+      "learning_rate": 7.500598392279119e-06,
+      "loss": 0.7342,
+      "step": 5520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1508312791543873,
+      "learning_rate": 7.499700785954633e-06,
+      "loss": 0.7973,
+      "step": 5521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.196472358627178,
+      "learning_rate": 7.498803072212439e-06,
+      "loss": 0.6731,
+      "step": 5522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1707762323720883,
+      "learning_rate": 7.49790525109111e-06,
+      "loss": 0.7071,
+      "step": 5523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.005505283905552,
+      "learning_rate": 7.497007322629231e-06,
+      "loss": 0.9442,
+      "step": 5524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0314514024110895,
+      "learning_rate": 7.496109286865384e-06,
+      "loss": 0.6946,
+      "step": 5525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0330608367177136,
+      "learning_rate": 7.495211143838164e-06,
+      "loss": 0.7157,
+      "step": 5526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1195739376068543,
+      "learning_rate": 7.494312893586163e-06,
+      "loss": 1.1269,
+      "step": 5527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1905237825182855,
+      "learning_rate": 7.493414536147983e-06,
+      "loss": 0.7906,
+      "step": 5528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5095461598761517,
+      "learning_rate": 7.492516071562226e-06,
+      "loss": 0.9921,
+      "step": 5529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7283398127062641,
+      "learning_rate": 7.491617499867502e-06,
+      "loss": 0.6085,
+      "step": 5530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.03630655063665,
+      "learning_rate": 7.4907188211024254e-06,
+      "loss": 1.0029,
+      "step": 5531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2869454018708302,
+      "learning_rate": 7.489820035305613e-06,
+      "loss": 0.6103,
+      "step": 5532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.53167810325461,
+      "learning_rate": 7.488921142515691e-06,
+      "loss": 0.8385,
+      "step": 5533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1592297904821103,
+      "learning_rate": 7.488022142771282e-06,
+      "loss": 0.6583,
+      "step": 5534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2782556249272983,
+      "learning_rate": 7.4871230361110214e-06,
+      "loss": 0.6299,
+      "step": 5535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8666505423823025,
+      "learning_rate": 7.486223822573545e-06,
+      "loss": 0.8106,
+      "step": 5536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4465855066203572,
+      "learning_rate": 7.485324502197494e-06,
+      "loss": 0.6985,
+      "step": 5537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.227888873567727,
+      "learning_rate": 7.484425075021513e-06,
+      "loss": 0.7937,
+      "step": 5538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9768322881216048,
+      "learning_rate": 7.483525541084253e-06,
+      "loss": 0.6592,
+      "step": 5539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0236110475967095,
+      "learning_rate": 7.482625900424372e-06,
+      "loss": 0.9932,
+      "step": 5540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.109927674218301,
+      "learning_rate": 7.481726153080524e-06,
+      "loss": 0.9455,
+      "step": 5541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8285218094516875,
+      "learning_rate": 7.480826299091378e-06,
+      "loss": 0.7386,
+      "step": 5542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0417384394807785,
+      "learning_rate": 7.4799263384956e-06,
+      "loss": 0.9243,
+      "step": 5543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2707552552823516,
+      "learning_rate": 7.479026271331864e-06,
+      "loss": 0.6868,
+      "step": 5544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.927538345893023,
+      "learning_rate": 7.478126097638848e-06,
+      "loss": 0.6615,
+      "step": 5545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.980784237823109,
+      "learning_rate": 7.477225817455236e-06,
+      "loss": 0.8047,
+      "step": 5546
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.05967278009586,
+      "learning_rate": 7.476325430819713e-06,
+      "loss": 0.6917,
+      "step": 5547
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0982220310268123,
+      "learning_rate": 7.47542493777097e-06,
+      "loss": 0.708,
+      "step": 5548
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7636672794294777,
+      "learning_rate": 7.4745243383477055e-06,
+      "loss": 0.7425,
+      "step": 5549
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4698586711514463,
+      "learning_rate": 7.473623632588618e-06,
+      "loss": 0.8239,
+      "step": 5550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0719742261230365,
+      "learning_rate": 7.472722820532414e-06,
+      "loss": 0.6968,
+      "step": 5551
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2356748423242863,
+      "learning_rate": 7.471821902217804e-06,
+      "loss": 0.7679,
+      "step": 5552
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0872908301322886,
+      "learning_rate": 7.470920877683503e-06,
+      "loss": 0.6894,
+      "step": 5553
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.190461635263598,
+      "learning_rate": 7.470019746968226e-06,
+      "loss": 0.7725,
+      "step": 5554
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.535734414576722,
+      "learning_rate": 7.469118510110702e-06,
+      "loss": 0.7845,
+      "step": 5555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9446541839410003,
+      "learning_rate": 7.468217167149656e-06,
+      "loss": 0.9912,
+      "step": 5556
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4729311245260552,
+      "learning_rate": 7.467315718123821e-06,
+      "loss": 0.9341,
+      "step": 5557
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8068427390893513,
+      "learning_rate": 7.466414163071934e-06,
+      "loss": 0.7482,
+      "step": 5558
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9300237409258325,
+      "learning_rate": 7.4655125020327376e-06,
+      "loss": 0.9137,
+      "step": 5559
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4231493908224913,
+      "learning_rate": 7.464610735044979e-06,
+      "loss": 0.7567,
+      "step": 5560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3386430803091276,
+      "learning_rate": 7.463708862147408e-06,
+      "loss": 0.8045,
+      "step": 5561
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2933558142568768,
+      "learning_rate": 7.462806883378779e-06,
+      "loss": 0.6479,
+      "step": 5562
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.78784124593996,
+      "learning_rate": 7.4619047987778536e-06,
+      "loss": 0.897,
+      "step": 5563
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.175694074517188,
+      "learning_rate": 7.461002608383396e-06,
+      "loss": 0.8812,
+      "step": 5564
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4492748349257298,
+      "learning_rate": 7.460100312234176e-06,
+      "loss": 0.7117,
+      "step": 5565
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9569235980159545,
+      "learning_rate": 7.4591979103689675e-06,
+      "loss": 0.8982,
+      "step": 5566
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.626375119118812,
+      "learning_rate": 7.458295402826546e-06,
+      "loss": 0.7382,
+      "step": 5567
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8641216222220338,
+      "learning_rate": 7.4573927896456966e-06,
+      "loss": 0.6711,
+      "step": 5568
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.917751824607234,
+      "learning_rate": 7.456490070865206e-06,
+      "loss": 0.8597,
+      "step": 5569
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.918827375552496,
+      "learning_rate": 7.455587246523868e-06,
+      "loss": 0.8504,
+      "step": 5570
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9466055678641674,
+      "learning_rate": 7.454684316660475e-06,
+      "loss": 0.5989,
+      "step": 5571
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4784022580321174,
+      "learning_rate": 7.453781281313831e-06,
+      "loss": 0.8722,
+      "step": 5572
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.015604341848856,
+      "learning_rate": 7.45287814052274e-06,
+      "loss": 0.7977,
+      "step": 5573
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9865060524673614,
+      "learning_rate": 7.4519748943260126e-06,
+      "loss": 0.7629,
+      "step": 5574
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3198946734594443,
+      "learning_rate": 7.4510715427624635e-06,
+      "loss": 0.745,
+      "step": 5575
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0464124004616844,
+      "learning_rate": 7.450168085870911e-06,
+      "loss": 0.9852,
+      "step": 5576
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1825179908091323,
+      "learning_rate": 7.4492645236901795e-06,
+      "loss": 0.8342,
+      "step": 5577
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9894943653478345,
+      "learning_rate": 7.4483608562590955e-06,
+      "loss": 0.8344,
+      "step": 5578
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.227223353345404,
+      "learning_rate": 7.447457083616494e-06,
+      "loss": 0.9014,
+      "step": 5579
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.260156037041933,
+      "learning_rate": 7.4465532058012105e-06,
+      "loss": 0.7552,
+      "step": 5580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0731942641743544,
+      "learning_rate": 7.445649222852086e-06,
+      "loss": 0.9251,
+      "step": 5581
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.545738839845344,
+      "learning_rate": 7.444745134807969e-06,
+      "loss": 0.9868,
+      "step": 5582
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9162248254602763,
+      "learning_rate": 7.443840941707709e-06,
+      "loss": 0.8589,
+      "step": 5583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.250520931974145,
+      "learning_rate": 7.44293664359016e-06,
+      "loss": 0.6654,
+      "step": 5584
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9720358813737435,
+      "learning_rate": 7.442032240494184e-06,
+      "loss": 0.8251,
+      "step": 5585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.222311562238274,
+      "learning_rate": 7.441127732458642e-06,
+      "loss": 0.8573,
+      "step": 5586
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6436989648313767,
+      "learning_rate": 7.440223119522405e-06,
+      "loss": 0.8012,
+      "step": 5587
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.90230451979359,
+      "learning_rate": 7.439318401724347e-06,
+      "loss": 0.8842,
+      "step": 5588
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.266378026470148,
+      "learning_rate": 7.438413579103344e-06,
+      "loss": 1.0646,
+      "step": 5589
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8415722903444964,
+      "learning_rate": 7.4375086516982775e-06,
+      "loss": 0.777,
+      "step": 5590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1359413543870371,
+      "learning_rate": 7.436603619548038e-06,
+      "loss": 0.7236,
+      "step": 5591
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1304242979129984,
+      "learning_rate": 7.4356984826915135e-06,
+      "loss": 0.945,
+      "step": 5592
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2054401660608973,
+      "learning_rate": 7.434793241167601e-06,
+      "loss": 0.9693,
+      "step": 5593
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0070656553633155,
+      "learning_rate": 7.433887895015199e-06,
+      "loss": 0.5982,
+      "step": 5594
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.269782008455272,
+      "learning_rate": 7.432982444273215e-06,
+      "loss": 0.7529,
+      "step": 5595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0879681150060034,
+      "learning_rate": 7.432076888980556e-06,
+      "loss": 0.8065,
+      "step": 5596
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0680545612238945,
+      "learning_rate": 7.431171229176136e-06,
+      "loss": 0.6783,
+      "step": 5597
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9530490357948262,
+      "learning_rate": 7.4302654648988755e-06,
+      "loss": 0.7706,
+      "step": 5598
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6064105524436516,
+      "learning_rate": 7.429359596187694e-06,
+      "loss": 0.6597,
+      "step": 5599
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9596197454304671,
+      "learning_rate": 7.428453623081522e-06,
+      "loss": 0.9266,
+      "step": 5600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1788456636497116,
+      "learning_rate": 7.427547545619288e-06,
+      "loss": 0.7974,
+      "step": 5601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.184406724870993,
+      "learning_rate": 7.4266413638399284e-06,
+      "loss": 1.0072,
+      "step": 5602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.927509941799466,
+      "learning_rate": 7.425735077782387e-06,
+      "loss": 1.0061,
+      "step": 5603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.126517761317861,
+      "learning_rate": 7.424828687485606e-06,
+      "loss": 0.897,
+      "step": 5604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2209180169337346,
+      "learning_rate": 7.423922192988536e-06,
+      "loss": 0.9509,
+      "step": 5605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1204620648272297,
+      "learning_rate": 7.423015594330131e-06,
+      "loss": 0.9955,
+      "step": 5606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1051815383697177,
+      "learning_rate": 7.422108891549349e-06,
+      "loss": 0.7248,
+      "step": 5607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2416673935167781,
+      "learning_rate": 7.421202084685155e-06,
+      "loss": 0.7294,
+      "step": 5608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0614864908001245,
+      "learning_rate": 7.420295173776515e-06,
+      "loss": 0.8278,
+      "step": 5609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.314824021213368,
+      "learning_rate": 7.4193881588624e-06,
+      "loss": 0.6937,
+      "step": 5610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.144658510705721,
+      "learning_rate": 7.418481039981787e-06,
+      "loss": 0.7441,
+      "step": 5611
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1645495358628115,
+      "learning_rate": 7.41757381717366e-06,
+      "loss": 0.9056,
+      "step": 5612
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1751803006254122,
+      "learning_rate": 7.416666490476999e-06,
+      "loss": 0.957,
+      "step": 5613
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.060209506307927,
+      "learning_rate": 7.415759059930799e-06,
+      "loss": 0.8409,
+      "step": 5614
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3862455761055346,
+      "learning_rate": 7.41485152557405e-06,
+      "loss": 0.9273,
+      "step": 5615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.233158987966399,
+      "learning_rate": 7.4139438874457536e-06,
+      "loss": 0.6776,
+      "step": 5616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2456515344185344,
+      "learning_rate": 7.413036145584913e-06,
+      "loss": 0.9202,
+      "step": 5617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.788834788300273,
+      "learning_rate": 7.412128300030534e-06,
+      "loss": 0.9135,
+      "step": 5618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.663653456682062,
+      "learning_rate": 7.411220350821631e-06,
+      "loss": 0.8193,
+      "step": 5619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9367615986446869,
+      "learning_rate": 7.410312297997219e-06,
+      "loss": 0.7001,
+      "step": 5620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0001831833969137,
+      "learning_rate": 7.409404141596319e-06,
+      "loss": 0.6541,
+      "step": 5621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3423855797156485,
+      "learning_rate": 7.4084958816579575e-06,
+      "loss": 0.9606,
+      "step": 5622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.210105976620731,
+      "learning_rate": 7.407587518221164e-06,
+      "loss": 0.877,
+      "step": 5623
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.987665722012593,
+      "learning_rate": 7.406679051324972e-06,
+      "loss": 0.8589,
+      "step": 5624
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.692112542796545,
+      "learning_rate": 7.4057704810084215e-06,
+      "loss": 0.823,
+      "step": 5625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.766490739789179,
+      "learning_rate": 7.404861807310556e-06,
+      "loss": 0.8494,
+      "step": 5626
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0154108576761276,
+      "learning_rate": 7.4039530302704236e-06,
+      "loss": 0.7867,
+      "step": 5627
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.274669180440114,
+      "learning_rate": 7.403044149927074e-06,
+      "loss": 0.7913,
+      "step": 5628
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.169636028156221,
+      "learning_rate": 7.402135166319567e-06,
+      "loss": 0.6204,
+      "step": 5629
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1726994523894667,
+      "learning_rate": 7.40122607948696e-06,
+      "loss": 0.6392,
+      "step": 5630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.046863544642998,
+      "learning_rate": 7.400316889468322e-06,
+      "loss": 0.7894,
+      "step": 5631
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.187971668722158,
+      "learning_rate": 7.399407596302721e-06,
+      "loss": 0.8362,
+      "step": 5632
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5804411435070005,
+      "learning_rate": 7.398498200029232e-06,
+      "loss": 0.9355,
+      "step": 5633
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.180867613944648,
+      "learning_rate": 7.397588700686933e-06,
+      "loss": 0.8116,
+      "step": 5634
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5454885358947017,
+      "learning_rate": 7.396679098314908e-06,
+      "loss": 0.8418,
+      "step": 5635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.033098696602037,
+      "learning_rate": 7.395769392952246e-06,
+      "loss": 0.8311,
+      "step": 5636
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.157595476971495,
+      "learning_rate": 7.394859584638035e-06,
+      "loss": 0.7413,
+      "step": 5637
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1899116520713457,
+      "learning_rate": 7.393949673411375e-06,
+      "loss": 0.9096,
+      "step": 5638
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.310089542621758,
+      "learning_rate": 7.393039659311366e-06,
+      "loss": 0.7658,
+      "step": 5639
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.601121089600706,
+      "learning_rate": 7.392129542377114e-06,
+      "loss": 0.9219,
+      "step": 5640
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9806849289018562,
+      "learning_rate": 7.391219322647728e-06,
+      "loss": 0.7722,
+      "step": 5641
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9629407057739288,
+      "learning_rate": 7.390309000162321e-06,
+      "loss": 0.8718,
+      "step": 5642
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.990489816096123,
+      "learning_rate": 7.389398574960014e-06,
+      "loss": 0.7561,
+      "step": 5643
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9197396886109916,
+      "learning_rate": 7.388488047079927e-06,
+      "loss": 0.7346,
+      "step": 5644
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9840702898631608,
+      "learning_rate": 7.387577416561191e-06,
+      "loss": 0.8307,
+      "step": 5645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.407223871531683,
+      "learning_rate": 7.386666683442936e-06,
+      "loss": 0.9951,
+      "step": 5646
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6331179546586285,
+      "learning_rate": 7.3857558477642966e-06,
+      "loss": 0.8509,
+      "step": 5647
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9359537861321914,
+      "learning_rate": 7.3848449095644155e-06,
+      "loss": 0.9221,
+      "step": 5648
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4062131484868785,
+      "learning_rate": 7.383933868882438e-06,
+      "loss": 0.6893,
+      "step": 5649
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0522709772051693,
+      "learning_rate": 7.383022725757514e-06,
+      "loss": 0.7284,
+      "step": 5650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.210554912138151,
+      "learning_rate": 7.382111480228793e-06,
+      "loss": 0.7738,
+      "step": 5651
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9551846448931185,
+      "learning_rate": 7.381200132335438e-06,
+      "loss": 0.8331,
+      "step": 5652
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.412812847322408,
+      "learning_rate": 7.38028868211661e-06,
+      "loss": 0.9111,
+      "step": 5653
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0587437308002325,
+      "learning_rate": 7.379377129611478e-06,
+      "loss": 1.0231,
+      "step": 5654
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7598261455543285,
+      "learning_rate": 7.378465474859211e-06,
+      "loss": 1.0429,
+      "step": 5655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4060394859127452,
+      "learning_rate": 7.377553717898983e-06,
+      "loss": 0.7692,
+      "step": 5656
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2288911723743734,
+      "learning_rate": 7.376641858769978e-06,
+      "loss": 0.9101,
+      "step": 5657
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.163195679082153,
+      "learning_rate": 7.375729897511379e-06,
+      "loss": 0.7216,
+      "step": 5658
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.215985919314139,
+      "learning_rate": 7.374817834162378e-06,
+      "loss": 0.8098,
+      "step": 5659
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2714930104567244,
+      "learning_rate": 7.3739056687621616e-06,
+      "loss": 0.8912,
+      "step": 5660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0764935424206623,
+      "learning_rate": 7.3729934013499335e-06,
+      "loss": 0.7196,
+      "step": 5661
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.358393004307054,
+      "learning_rate": 7.372081031964893e-06,
+      "loss": 0.906,
+      "step": 5662
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.303548826584258,
+      "learning_rate": 7.37116856064625e-06,
+      "loss": 0.699,
+      "step": 5663
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1695536181444472,
+      "learning_rate": 7.3702559874332125e-06,
+      "loss": 0.816,
+      "step": 5664
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0680406528759765,
+      "learning_rate": 7.369343312364994e-06,
+      "loss": 0.5842,
+      "step": 5665
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.35283783675017,
+      "learning_rate": 7.368430535480818e-06,
+      "loss": 0.9588,
+      "step": 5666
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.415511997470617,
+      "learning_rate": 7.367517656819907e-06,
+      "loss": 0.9189,
+      "step": 5667
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0865565810777227,
+      "learning_rate": 7.366604676421491e-06,
+      "loss": 0.8589,
+      "step": 5668
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.110370592729691,
+      "learning_rate": 7.3656915943247984e-06,
+      "loss": 0.985,
+      "step": 5669
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8313377855835828,
+      "learning_rate": 7.364778410569071e-06,
+      "loss": 0.7159,
+      "step": 5670
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.96619280851054,
+      "learning_rate": 7.363865125193549e-06,
+      "loss": 0.7828,
+      "step": 5671
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0488569084451296,
+      "learning_rate": 7.362951738237477e-06,
+      "loss": 0.8753,
+      "step": 5672
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9920590399430307,
+      "learning_rate": 7.362038249740108e-06,
+      "loss": 0.7881,
+      "step": 5673
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9708271960242363,
+      "learning_rate": 7.3611246597406925e-06,
+      "loss": 0.7581,
+      "step": 5674
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1983566045887932,
+      "learning_rate": 7.360210968278494e-06,
+      "loss": 0.7477,
+      "step": 5675
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1165960380074682,
+      "learning_rate": 7.359297175392772e-06,
+      "loss": 0.7753,
+      "step": 5676
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9150965179474806,
+      "learning_rate": 7.358383281122797e-06,
+      "loss": 0.7396,
+      "step": 5677
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5568862270728148,
+      "learning_rate": 7.35746928550784e-06,
+      "loss": 0.8088,
+      "step": 5678
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.040507115262966,
+      "learning_rate": 7.356555188587178e-06,
+      "loss": 0.7389,
+      "step": 5679
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9828930473411273,
+      "learning_rate": 7.355640990400089e-06,
+      "loss": 0.935,
+      "step": 5680
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9176035653871848,
+      "learning_rate": 7.354726690985862e-06,
+      "loss": 0.7631,
+      "step": 5681
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.180739357438346,
+      "learning_rate": 7.353812290383785e-06,
+      "loss": 0.7198,
+      "step": 5682
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7367130397578283,
+      "learning_rate": 7.352897788633151e-06,
+      "loss": 0.7372,
+      "step": 5683
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1745909992074752,
+      "learning_rate": 7.351983185773259e-06,
+      "loss": 0.6828,
+      "step": 5684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1679755505335194,
+      "learning_rate": 7.351068481843411e-06,
+      "loss": 0.8029,
+      "step": 5685
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.550881447257238,
+      "learning_rate": 7.3501536768829154e-06,
+      "loss": 0.8247,
+      "step": 5686
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4918875807970844,
+      "learning_rate": 7.349238770931081e-06,
+      "loss": 0.68,
+      "step": 5687
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0007358186303859,
+      "learning_rate": 7.348323764027225e-06,
+      "loss": 0.7119,
+      "step": 5688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0962228491137098,
+      "learning_rate": 7.347408656210666e-06,
+      "loss": 0.5607,
+      "step": 5689
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3615190468553717,
+      "learning_rate": 7.346493447520731e-06,
+      "loss": 0.5991,
+      "step": 5690
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8510058365424413,
+      "learning_rate": 7.345578137996745e-06,
+      "loss": 0.7831,
+      "step": 5691
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0465648544565616,
+      "learning_rate": 7.344662727678043e-06,
+      "loss": 0.6924,
+      "step": 5692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.281651673562725,
+      "learning_rate": 7.343747216603962e-06,
+      "loss": 0.7498,
+      "step": 5693
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.161613522728383,
+      "learning_rate": 7.342831604813844e-06,
+      "loss": 0.9063,
+      "step": 5694
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.202363017362974,
+      "learning_rate": 7.341915892347033e-06,
+      "loss": 0.7786,
+      "step": 5695
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8893483194809576,
+      "learning_rate": 7.341000079242881e-06,
+      "loss": 0.6027,
+      "step": 5696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.443415583284235,
+      "learning_rate": 7.340084165540742e-06,
+      "loss": 0.9303,
+      "step": 5697
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2715607944492566,
+      "learning_rate": 7.339168151279974e-06,
+      "loss": 0.7361,
+      "step": 5698
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3988910497250275,
+      "learning_rate": 7.338252036499941e-06,
+      "loss": 0.8759,
+      "step": 5699
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7539374410654978,
+      "learning_rate": 7.3373358212400104e-06,
+      "loss": 0.7212,
+      "step": 5700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8381340691851575,
+      "learning_rate": 7.336419505539555e-06,
+      "loss": 0.7618,
+      "step": 5701
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2025160418322045,
+      "learning_rate": 7.33550308943795e-06,
+      "loss": 0.9285,
+      "step": 5702
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8604051610186996,
+      "learning_rate": 7.334586572974573e-06,
+      "loss": 0.9435,
+      "step": 5703
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8113054106127637,
+      "learning_rate": 7.333669956188815e-06,
+      "loss": 0.9784,
+      "step": 5704
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9977361856123316,
+      "learning_rate": 7.332753239120061e-06,
+      "loss": 0.5701,
+      "step": 5705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0265153854976656,
+      "learning_rate": 7.331836421807704e-06,
+      "loss": 0.9077,
+      "step": 5706
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9384645435860808,
+      "learning_rate": 7.330919504291143e-06,
+      "loss": 0.7271,
+      "step": 5707
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.219565135156038,
+      "learning_rate": 7.33000248660978e-06,
+      "loss": 0.7993,
+      "step": 5708
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.030794472158252,
+      "learning_rate": 7.3290853688030196e-06,
+      "loss": 0.7895,
+      "step": 5709
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8674030522366452,
+      "learning_rate": 7.328168150910275e-06,
+      "loss": 0.8243,
+      "step": 5710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.188832688733305,
+      "learning_rate": 7.32725083297096e-06,
+      "loss": 0.6869,
+      "step": 5711
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4992342257726827,
+      "learning_rate": 7.326333415024494e-06,
+      "loss": 0.819,
+      "step": 5712
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.437068708814835,
+      "learning_rate": 7.325415897110299e-06,
+      "loss": 0.9538,
+      "step": 5713
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2897339904995637,
+      "learning_rate": 7.324498279267803e-06,
+      "loss": 1.0148,
+      "step": 5714
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2715043351219868,
+      "learning_rate": 7.323580561536442e-06,
+      "loss": 0.7061,
+      "step": 5715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9731396599876962,
+      "learning_rate": 7.3226627439556466e-06,
+      "loss": 0.6836,
+      "step": 5716
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2085711569502289,
+      "learning_rate": 7.321744826564862e-06,
+      "loss": 0.6747,
+      "step": 5717
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2652386691967976,
+      "learning_rate": 7.3208268094035295e-06,
+      "loss": 0.9664,
+      "step": 5718
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.874974398497163,
+      "learning_rate": 7.319908692511103e-06,
+      "loss": 0.8187,
+      "step": 5719
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1418277854258894,
+      "learning_rate": 7.318990475927033e-06,
+      "loss": 0.7781,
+      "step": 5720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.555492085067305,
+      "learning_rate": 7.318072159690775e-06,
+      "loss": 0.9663,
+      "step": 5721
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0989418817501444,
+      "learning_rate": 7.317153743841795e-06,
+      "loss": 0.9386,
+      "step": 5722
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0813997708845458,
+      "learning_rate": 7.316235228419557e-06,
+      "loss": 0.6308,
+      "step": 5723
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7493413110445313,
+      "learning_rate": 7.315316613463535e-06,
+      "loss": 0.9974,
+      "step": 5724
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0148138501969104,
+      "learning_rate": 7.314397899013199e-06,
+      "loss": 0.6619,
+      "step": 5725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.147587669966531,
+      "learning_rate": 7.313479085108033e-06,
+      "loss": 0.8409,
+      "step": 5726
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9725522933834565,
+      "learning_rate": 7.312560171787518e-06,
+      "loss": 0.8784,
+      "step": 5727
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3129749858179145,
+      "learning_rate": 7.311641159091141e-06,
+      "loss": 0.8799,
+      "step": 5728
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2338892581157346,
+      "learning_rate": 7.310722047058396e-06,
+      "loss": 0.9242,
+      "step": 5729
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5649797581885556,
+      "learning_rate": 7.309802835728779e-06,
+      "loss": 1.0033,
+      "step": 5730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9895626326251987,
+      "learning_rate": 7.308883525141788e-06,
+      "loss": 0.8017,
+      "step": 5731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4998008231848994,
+      "learning_rate": 7.307964115336931e-06,
+      "loss": 0.693,
+      "step": 5732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2168468090709843,
+      "learning_rate": 7.307044606353715e-06,
+      "loss": 0.7724,
+      "step": 5733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.046483135822517,
+      "learning_rate": 7.306124998231655e-06,
+      "loss": 0.8745,
+      "step": 5734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1143626014123267,
+      "learning_rate": 7.305205291010269e-06,
+      "loss": 0.6156,
+      "step": 5735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4032858881681083,
+      "learning_rate": 7.304285484729074e-06,
+      "loss": 0.8073,
+      "step": 5736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.287323682827452,
+      "learning_rate": 7.303365579427603e-06,
+      "loss": 0.8697,
+      "step": 5737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9962481661662654,
+      "learning_rate": 7.302445575145383e-06,
+      "loss": 0.8356,
+      "step": 5738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.187817404953391,
+      "learning_rate": 7.301525471921949e-06,
+      "loss": 0.694,
+      "step": 5739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1010488058205405,
+      "learning_rate": 7.300605269796839e-06,
+      "loss": 0.8126,
+      "step": 5740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0008840944707225,
+      "learning_rate": 7.299684968809595e-06,
+      "loss": 0.6455,
+      "step": 5741
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.011750790502609,
+      "learning_rate": 7.298764568999768e-06,
+      "loss": 0.722,
+      "step": 5742
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0349509450475192,
+      "learning_rate": 7.297844070406908e-06,
+      "loss": 0.5802,
+      "step": 5743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9205678665557386,
+      "learning_rate": 7.296923473070571e-06,
+      "loss": 0.8031,
+      "step": 5744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.228890678537286,
+      "learning_rate": 7.296002777030316e-06,
+      "loss": 0.7934,
+      "step": 5745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2446898239420823,
+      "learning_rate": 7.295081982325708e-06,
+      "loss": 0.9105,
+      "step": 5746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.702133265262266,
+      "learning_rate": 7.2941610889963164e-06,
+      "loss": 0.819,
+      "step": 5747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9498513671987805,
+      "learning_rate": 7.293240097081714e-06,
+      "loss": 0.9258,
+      "step": 5748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.401437650874789,
+      "learning_rate": 7.292319006621477e-06,
+      "loss": 0.9111,
+      "step": 5749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5469596073635854,
+      "learning_rate": 7.291397817655186e-06,
+      "loss": 0.7258,
+      "step": 5750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2490302256093484,
+      "learning_rate": 7.2904765302224276e-06,
+      "loss": 0.7837,
+      "step": 5751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.121988645231223,
+      "learning_rate": 7.289555144362793e-06,
+      "loss": 0.8535,
+      "step": 5752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0134443688213772,
+      "learning_rate": 7.288633660115874e-06,
+      "loss": 0.6971,
+      "step": 5753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.778066115269923,
+      "learning_rate": 7.2877120775212685e-06,
+      "loss": 0.7988,
+      "step": 5754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.215708236804809,
+      "learning_rate": 7.286790396618582e-06,
+      "loss": 0.8427,
+      "step": 5755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3839358816377767,
+      "learning_rate": 7.285868617447418e-06,
+      "loss": 0.7097,
+      "step": 5756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.938440952042823,
+      "learning_rate": 7.28494674004739e-06,
+      "loss": 0.7278,
+      "step": 5757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1909945836885423,
+      "learning_rate": 7.28402476445811e-06,
+      "loss": 0.9553,
+      "step": 5758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4654324800040293,
+      "learning_rate": 7.283102690719198e-06,
+      "loss": 0.8567,
+      "step": 5759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.171386218764972,
+      "learning_rate": 7.282180518870281e-06,
+      "loss": 0.8737,
+      "step": 5760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8305381656648536,
+      "learning_rate": 7.2812582489509844e-06,
+      "loss": 1.026,
+      "step": 5761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1891786784309715,
+      "learning_rate": 7.28033588100094e-06,
+      "loss": 0.7683,
+      "step": 5762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.095679920308812,
+      "learning_rate": 7.279413415059782e-06,
+      "loss": 0.7504,
+      "step": 5763
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3541874798131524,
+      "learning_rate": 7.278490851167155e-06,
+      "loss": 0.8321,
+      "step": 5764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1019352484736475,
+      "learning_rate": 7.277568189362701e-06,
+      "loss": 0.9168,
+      "step": 5765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7239136368721724,
+      "learning_rate": 7.276645429686071e-06,
+      "loss": 1.0016,
+      "step": 5766
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3308752593703534,
+      "learning_rate": 7.275722572176915e-06,
+      "loss": 0.8048,
+      "step": 5767
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8996762135426517,
+      "learning_rate": 7.2747996168748915e-06,
+      "loss": 0.8975,
+      "step": 5768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9222663206632506,
+      "learning_rate": 7.2738765638196625e-06,
+      "loss": 0.9167,
+      "step": 5769
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1364245956713037,
+      "learning_rate": 7.272953413050894e-06,
+      "loss": 0.8518,
+      "step": 5770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.500935577624563,
+      "learning_rate": 7.272030164608256e-06,
+      "loss": 1.0555,
+      "step": 5771
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8569182601312124,
+      "learning_rate": 7.27110681853142e-06,
+      "loss": 0.6885,
+      "step": 5772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.192547135871792,
+      "learning_rate": 7.270183374860066e-06,
+      "loss": 0.6887,
+      "step": 5773
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0807885269907063,
+      "learning_rate": 7.269259833633877e-06,
+      "loss": 0.8616,
+      "step": 5774
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8863589820102638,
+      "learning_rate": 7.26833619489254e-06,
+      "loss": 1.0585,
+      "step": 5775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5438789092658625,
+      "learning_rate": 7.267412458675744e-06,
+      "loss": 0.9821,
+      "step": 5776
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.603719834356826,
+      "learning_rate": 7.266488625023184e-06,
+      "loss": 0.8369,
+      "step": 5777
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.473637390675325,
+      "learning_rate": 7.265564693974561e-06,
+      "loss": 0.9645,
+      "step": 5778
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.149053731904028,
+      "learning_rate": 7.264640665569577e-06,
+      "loss": 0.7502,
+      "step": 5779
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.29685949513018,
+      "learning_rate": 7.2637165398479405e-06,
+      "loss": 1.038,
+      "step": 5780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.125922608458984,
+      "learning_rate": 7.262792316849362e-06,
+      "loss": 0.6912,
+      "step": 5781
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.743110167128506,
+      "learning_rate": 7.261867996613559e-06,
+      "loss": 0.8114,
+      "step": 5782
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2051643544656727,
+      "learning_rate": 7.26094357918025e-06,
+      "loss": 0.9082,
+      "step": 5783
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4979898527647277,
+      "learning_rate": 7.26001906458916e-06,
+      "loss": 0.8992,
+      "step": 5784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2285609052909723,
+      "learning_rate": 7.259094452880019e-06,
+      "loss": 0.6491,
+      "step": 5785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.980864914145008,
+      "learning_rate": 7.258169744092556e-06,
+      "loss": 1.0705,
+      "step": 5786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.239390938091425,
+      "learning_rate": 7.257244938266511e-06,
+      "loss": 0.9502,
+      "step": 5787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.157610454552901,
+      "learning_rate": 7.256320035441623e-06,
+      "loss": 1.0987,
+      "step": 5788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.421816517460992,
+      "learning_rate": 7.255395035657639e-06,
+      "loss": 0.8222,
+      "step": 5789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3771200832819637,
+      "learning_rate": 7.254469938954307e-06,
+      "loss": 1.0469,
+      "step": 5790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.357424137734064,
+      "learning_rate": 7.25354474537138e-06,
+      "loss": 0.9106,
+      "step": 5791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8785116981519392,
+      "learning_rate": 7.252619454948617e-06,
+      "loss": 0.9314,
+      "step": 5792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0719240400484855,
+      "learning_rate": 7.251694067725779e-06,
+      "loss": 0.8367,
+      "step": 5793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8998872542685912,
+      "learning_rate": 7.250768583742634e-06,
+      "loss": 0.8141,
+      "step": 5794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4841794810061586,
+      "learning_rate": 7.249843003038948e-06,
+      "loss": 0.8715,
+      "step": 5795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9510652325543603,
+      "learning_rate": 7.2489173256544975e-06,
+      "loss": 0.9032,
+      "step": 5796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4542833196367386,
+      "learning_rate": 7.247991551629061e-06,
+      "loss": 0.8303,
+      "step": 5797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3843039848921577,
+      "learning_rate": 7.247065681002421e-06,
+      "loss": 0.657,
+      "step": 5798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1008225667622085,
+      "learning_rate": 7.246139713814365e-06,
+      "loss": 0.702,
+      "step": 5799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9132992476789563,
+      "learning_rate": 7.245213650104683e-06,
+      "loss": 0.8961,
+      "step": 5800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.199768736147951,
+      "learning_rate": 7.2442874899131685e-06,
+      "loss": 0.7445,
+      "step": 5801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.308753990294628,
+      "learning_rate": 7.2433612332796245e-06,
+      "loss": 0.7856,
+      "step": 5802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.060485240686809,
+      "learning_rate": 7.242434880243851e-06,
+      "loss": 0.8542,
+      "step": 5803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.433783943976579,
+      "learning_rate": 7.241508430845656e-06,
+      "loss": 0.9435,
+      "step": 5804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1012042016844326,
+      "learning_rate": 7.2405818851248535e-06,
+      "loss": 0.6746,
+      "step": 5805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5126101728615446,
+      "learning_rate": 7.239655243121255e-06,
+      "loss": 0.9942,
+      "step": 5806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.06551996197297,
+      "learning_rate": 7.238728504874683e-06,
+      "loss": 0.6489,
+      "step": 5807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.185268562672364,
+      "learning_rate": 7.237801670424963e-06,
+      "loss": 0.6956,
+      "step": 5808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.161032559122812,
+      "learning_rate": 7.236874739811921e-06,
+      "loss": 0.8539,
+      "step": 5809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0950246807203585,
+      "learning_rate": 7.235947713075389e-06,
+      "loss": 0.7367,
+      "step": 5810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1443688680149835,
+      "learning_rate": 7.235020590255204e-06,
+      "loss": 0.72,
+      "step": 5811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.197778966748444,
+      "learning_rate": 7.234093371391206e-06,
+      "loss": 0.7555,
+      "step": 5812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9973401812765108,
+      "learning_rate": 7.233166056523242e-06,
+      "loss": 0.6012,
+      "step": 5813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.448449843559809,
+      "learning_rate": 7.232238645691157e-06,
+      "loss": 0.8019,
+      "step": 5814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8230007059504245,
+      "learning_rate": 7.231311138934807e-06,
+      "loss": 0.7169,
+      "step": 5815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.188303977721419,
+      "learning_rate": 7.230383536294048e-06,
+      "loss": 0.8129,
+      "step": 5816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9919589917814302,
+      "learning_rate": 7.229455837808741e-06,
+      "loss": 0.9404,
+      "step": 5817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2305813821917346,
+      "learning_rate": 7.228528043518752e-06,
+      "loss": 0.6819,
+      "step": 5818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9926365796223084,
+      "learning_rate": 7.227600153463947e-06,
+      "loss": 0.985,
+      "step": 5819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.708142258836712,
+      "learning_rate": 7.226672167684205e-06,
+      "loss": 0.8109,
+      "step": 5820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9372293907042144,
+      "learning_rate": 7.225744086219402e-06,
+      "loss": 0.7256,
+      "step": 5821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.391438647738153,
+      "learning_rate": 7.224815909109416e-06,
+      "loss": 0.9307,
+      "step": 5822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0731130120268513,
+      "learning_rate": 7.223887636394137e-06,
+      "loss": 0.8833,
+      "step": 5823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.167991420959418,
+      "learning_rate": 7.222959268113452e-06,
+      "loss": 0.5101,
+      "step": 5824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.846072052855019,
+      "learning_rate": 7.222030804307257e-06,
+      "loss": 0.8661,
+      "step": 5825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9819851209633783,
+      "learning_rate": 7.22110224501545e-06,
+      "loss": 0.7488,
+      "step": 5826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.069588695911669,
+      "learning_rate": 7.220173590277933e-06,
+      "loss": 0.9502,
+      "step": 5827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0264399246811875,
+      "learning_rate": 7.219244840134612e-06,
+      "loss": 0.7552,
+      "step": 5828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.051480633194552,
+      "learning_rate": 7.218315994625397e-06,
+      "loss": 0.915,
+      "step": 5829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.809595103275291,
+      "learning_rate": 7.217387053790204e-06,
+      "loss": 0.9731,
+      "step": 5830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.859878295423208,
+      "learning_rate": 7.216458017668951e-06,
+      "loss": 0.8746,
+      "step": 5831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.275926041444523,
+      "learning_rate": 7.21552888630156e-06,
+      "loss": 0.7591,
+      "step": 5832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0047163935397827,
+      "learning_rate": 7.214599659727959e-06,
+      "loss": 0.8778,
+      "step": 5833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1055803532226274,
+      "learning_rate": 7.213670337988079e-06,
+      "loss": 0.6374,
+      "step": 5834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2052217655644364,
+      "learning_rate": 7.2127409211218526e-06,
+      "loss": 0.6575,
+      "step": 5835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3428969710707066,
+      "learning_rate": 7.211811409169222e-06,
+      "loss": 0.5345,
+      "step": 5836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8813335148810943,
+      "learning_rate": 7.21088180217013e-06,
+      "loss": 0.995,
+      "step": 5837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0691178722566195,
+      "learning_rate": 7.2099521001645225e-06,
+      "loss": 0.6647,
+      "step": 5838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0064087414436114,
+      "learning_rate": 7.209022303192351e-06,
+      "loss": 0.8089,
+      "step": 5839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8325496838072686,
+      "learning_rate": 7.208092411293572e-06,
+      "loss": 0.7343,
+      "step": 5840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.083567793587159,
+      "learning_rate": 7.207162424508144e-06,
+      "loss": 0.8169,
+      "step": 5841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.086437825709854,
+      "learning_rate": 7.206232342876031e-06,
+      "loss": 1.0902,
+      "step": 5842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8036398354315328,
+      "learning_rate": 7.2053021664372e-06,
+      "loss": 0.7133,
+      "step": 5843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.246518697039027,
+      "learning_rate": 7.204371895231623e-06,
+      "loss": 0.7741,
+      "step": 5844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6131386842590194,
+      "learning_rate": 7.20344152929928e-06,
+      "loss": 0.8992,
+      "step": 5845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1511771136973357,
+      "learning_rate": 7.202511068680143e-06,
+      "loss": 0.961,
+      "step": 5846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9779618636827232,
+      "learning_rate": 7.201580513414201e-06,
+      "loss": 0.6674,
+      "step": 5847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.074466391727464,
+      "learning_rate": 7.200649863541442e-06,
+      "loss": 0.7974,
+      "step": 5848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.040922260981622,
+      "learning_rate": 7.199719119101858e-06,
+      "loss": 0.6618,
+      "step": 5849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0001180956781828,
+      "learning_rate": 7.198788280135443e-06,
+      "loss": 0.7733,
+      "step": 5850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9970688052250112,
+      "learning_rate": 7.197857346682199e-06,
+      "loss": 0.8353,
+      "step": 5851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.322676631089809,
+      "learning_rate": 7.19692631878213e-06,
+      "loss": 0.6923,
+      "step": 5852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0814474154679243,
+      "learning_rate": 7.195995196475244e-06,
+      "loss": 0.7032,
+      "step": 5853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6659440221881767,
+      "learning_rate": 7.195063979801554e-06,
+      "loss": 0.9575,
+      "step": 5854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8368463137361195,
+      "learning_rate": 7.194132668801077e-06,
+      "loss": 0.8016,
+      "step": 5855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2260608851498147,
+      "learning_rate": 7.193201263513831e-06,
+      "loss": 0.6986,
+      "step": 5856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.242584696693612,
+      "learning_rate": 7.192269763979844e-06,
+      "loss": 0.9269,
+      "step": 5857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5795753111272237,
+      "learning_rate": 7.1913381702391405e-06,
+      "loss": 0.7532,
+      "step": 5858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.073622966942047,
+      "learning_rate": 7.190406482331757e-06,
+      "loss": 0.8141,
+      "step": 5859
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.579953547016698,
+      "learning_rate": 7.189474700297728e-06,
+      "loss": 0.9242,
+      "step": 5860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.444244604654204,
+      "learning_rate": 7.188542824177096e-06,
+      "loss": 0.8705,
+      "step": 5861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.323736771657211,
+      "learning_rate": 7.187610854009903e-06,
+      "loss": 0.8545,
+      "step": 5862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8983433914495114,
+      "learning_rate": 7.186678789836201e-06,
+      "loss": 0.8541,
+      "step": 5863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0684453834031253,
+      "learning_rate": 7.18574663169604e-06,
+      "loss": 0.8761,
+      "step": 5864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9518457602159234,
+      "learning_rate": 7.18481437962948e-06,
+      "loss": 0.8066,
+      "step": 5865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.048671563639842,
+      "learning_rate": 7.183882033676579e-06,
+      "loss": 0.9479,
+      "step": 5866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3642800622402858,
+      "learning_rate": 7.182949593877404e-06,
+      "loss": 0.8198,
+      "step": 5867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1394756780581483,
+      "learning_rate": 7.182017060272023e-06,
+      "loss": 0.64,
+      "step": 5868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4528053573880606,
+      "learning_rate": 7.1810844329005095e-06,
+      "loss": 0.7573,
+      "step": 5869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.182501647330213,
+      "learning_rate": 7.180151711802941e-06,
+      "loss": 0.6469,
+      "step": 5870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.018766929456008,
+      "learning_rate": 7.179218897019395e-06,
+      "loss": 0.9346,
+      "step": 5871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.756178065517746,
+      "learning_rate": 7.178285988589963e-06,
+      "loss": 0.7376,
+      "step": 5872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9036599253224646,
+      "learning_rate": 7.177352986554729e-06,
+      "loss": 1.0255,
+      "step": 5873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.060294326445592,
+      "learning_rate": 7.176419890953788e-06,
+      "loss": 0.9445,
+      "step": 5874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4961590516487715,
+      "learning_rate": 7.175486701827239e-06,
+      "loss": 0.7632,
+      "step": 5875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1503379406633645,
+      "learning_rate": 7.174553419215179e-06,
+      "loss": 0.7459,
+      "step": 5876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3107126573482346,
+      "learning_rate": 7.173620043157716e-06,
+      "loss": 0.9059,
+      "step": 5877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3450763778522408,
+      "learning_rate": 7.17268657369496e-06,
+      "loss": 0.8572,
+      "step": 5878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2930733848583653,
+      "learning_rate": 7.171753010867023e-06,
+      "loss": 0.7618,
+      "step": 5879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6957707154734702,
+      "learning_rate": 7.1708193547140205e-06,
+      "loss": 0.873,
+      "step": 5880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3316934427816283,
+      "learning_rate": 7.169885605276077e-06,
+      "loss": 0.7749,
+      "step": 5881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.139203028559523,
+      "learning_rate": 7.1689517625933165e-06,
+      "loss": 0.7132,
+      "step": 5882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1045003137192477,
+      "learning_rate": 7.168017826705869e-06,
+      "loss": 0.7066,
+      "step": 5883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9689580923778407,
+      "learning_rate": 7.167083797653866e-06,
+      "loss": 0.7111,
+      "step": 5884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0263874310095753,
+      "learning_rate": 7.166149675477445e-06,
+      "loss": 0.6596,
+      "step": 5885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.137746736374165,
+      "learning_rate": 7.16521546021675e-06,
+      "loss": 0.7856,
+      "step": 5886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.175785580019997,
+      "learning_rate": 7.164281151911923e-06,
+      "loss": 0.6369,
+      "step": 5887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1222753216611236,
+      "learning_rate": 7.163346750603117e-06,
+      "loss": 0.7358,
+      "step": 5888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.983624232130168,
+      "learning_rate": 7.162412256330481e-06,
+      "loss": 0.6328,
+      "step": 5889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2332989611408558,
+      "learning_rate": 7.161477669134174e-06,
+      "loss": 0.6176,
+      "step": 5890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2267346887140165,
+      "learning_rate": 7.160542989054359e-06,
+      "loss": 0.9184,
+      "step": 5891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2599173579894716,
+      "learning_rate": 7.1596082161312005e-06,
+      "loss": 0.9464,
+      "step": 5892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.150792957779602,
+      "learning_rate": 7.158673350404866e-06,
+      "loss": 0.7309,
+      "step": 5893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7760763743439663,
+      "learning_rate": 7.157738391915531e-06,
+      "loss": 0.7807,
+      "step": 5894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0249718190354595,
+      "learning_rate": 7.156803340703371e-06,
+      "loss": 0.6652,
+      "step": 5895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7862685601811972,
+      "learning_rate": 7.155868196808569e-06,
+      "loss": 0.8403,
+      "step": 5896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0570119938659635,
+      "learning_rate": 7.154932960271308e-06,
+      "loss": 1.0103,
+      "step": 5897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.116351985554313,
+      "learning_rate": 7.153997631131778e-06,
+      "loss": 0.7038,
+      "step": 5898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4712411048032217,
+      "learning_rate": 7.153062209430174e-06,
+      "loss": 0.7358,
+      "step": 5899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6542018695127179,
+      "learning_rate": 7.152126695206689e-06,
+      "loss": 0.7441,
+      "step": 5900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7003807736066903,
+      "learning_rate": 7.151191088501531e-06,
+      "loss": 0.8585,
+      "step": 5901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.086300314185237,
+      "learning_rate": 7.150255389354899e-06,
+      "loss": 0.8049,
+      "step": 5902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.73295290456299,
+      "learning_rate": 7.149319597807003e-06,
+      "loss": 0.915,
+      "step": 5903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 19.725207062124962,
+      "learning_rate": 7.148383713898058e-06,
+      "loss": 0.8997,
+      "step": 5904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2158929132816936,
+      "learning_rate": 7.14744773766828e-06,
+      "loss": 0.8534,
+      "step": 5905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0552834775407116,
+      "learning_rate": 7.146511669157892e-06,
+      "loss": 0.7029,
+      "step": 5906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5036562567759075,
+      "learning_rate": 7.145575508407114e-06,
+      "loss": 0.7947,
+      "step": 5907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9718483778574412,
+      "learning_rate": 7.14463925545618e-06,
+      "loss": 0.7503,
+      "step": 5908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9940784683692947,
+      "learning_rate": 7.143702910345318e-06,
+      "loss": 0.8753,
+      "step": 5909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6839861580744953,
+      "learning_rate": 7.142766473114771e-06,
+      "loss": 0.7478,
+      "step": 5910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9498450209413383,
+      "learning_rate": 7.141829943804777e-06,
+      "loss": 0.8122,
+      "step": 5911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.10610273639892,
+      "learning_rate": 7.1408933224555765e-06,
+      "loss": 0.6023,
+      "step": 5912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.043421284739623,
+      "learning_rate": 7.139956609107425e-06,
+      "loss": 0.7775,
+      "step": 5913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1131896692978294,
+      "learning_rate": 7.139019803800569e-06,
+      "loss": 0.8918,
+      "step": 5914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.224379451917857,
+      "learning_rate": 7.138082906575271e-06,
+      "loss": 0.7686,
+      "step": 5915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.376905530377773,
+      "learning_rate": 7.137145917471788e-06,
+      "loss": 0.8531,
+      "step": 5916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3878701186999773,
+      "learning_rate": 7.136208836530385e-06,
+      "loss": 0.9806,
+      "step": 5917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8014954869063111,
+      "learning_rate": 7.13527166379133e-06,
+      "loss": 0.8141,
+      "step": 5918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.229751500348188,
+      "learning_rate": 7.134334399294897e-06,
+      "loss": 0.9249,
+      "step": 5919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3594467095685903,
+      "learning_rate": 7.1333970430813615e-06,
+      "loss": 0.9212,
+      "step": 5920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0462452833708313,
+      "learning_rate": 7.132459595191003e-06,
+      "loss": 0.6272,
+      "step": 5921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.514034998951008,
+      "learning_rate": 7.131522055664109e-06,
+      "loss": 1.0117,
+      "step": 5922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.332856719386903,
+      "learning_rate": 7.13058442454096e-06,
+      "loss": 1.0578,
+      "step": 5923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4395771085365667,
+      "learning_rate": 7.129646701861858e-06,
+      "loss": 0.7918,
+      "step": 5924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9301458951168984,
+      "learning_rate": 7.128708887667093e-06,
+      "loss": 0.9921,
+      "step": 5925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.954270656446494,
+      "learning_rate": 7.127770981996966e-06,
+      "loss": 0.9307,
+      "step": 5926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.008381267989516,
+      "learning_rate": 7.1268329848917795e-06,
+      "loss": 0.7253,
+      "step": 5927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6017829057376747,
+      "learning_rate": 7.125894896391845e-06,
+      "loss": 0.7656,
+      "step": 5928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0994296177172322,
+      "learning_rate": 7.124956716537471e-06,
+      "loss": 0.6843,
+      "step": 5929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.319872767369973,
+      "learning_rate": 7.124018445368975e-06,
+      "loss": 0.9177,
+      "step": 5930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.425747892448023,
+      "learning_rate": 7.123080082926677e-06,
+      "loss": 0.7471,
+      "step": 5931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3079881953768546,
+      "learning_rate": 7.122141629250896e-06,
+      "loss": 0.6767,
+      "step": 5932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.386403629622162,
+      "learning_rate": 7.1212030843819656e-06,
+      "loss": 0.9307,
+      "step": 5933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.88394860720681,
+      "learning_rate": 7.120264448360214e-06,
+      "loss": 0.8595,
+      "step": 5934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6121435160126114,
+      "learning_rate": 7.1193257212259784e-06,
+      "loss": 0.8708,
+      "step": 5935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9178319514490612,
+      "learning_rate": 7.118386903019594e-06,
+      "loss": 0.7111,
+      "step": 5936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9845142148067483,
+      "learning_rate": 7.117447993781408e-06,
+      "loss": 0.832,
+      "step": 5937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1745468039753533,
+      "learning_rate": 7.116508993551765e-06,
+      "loss": 0.9595,
+      "step": 5938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4527957261293647,
+      "learning_rate": 7.115569902371018e-06,
+      "loss": 0.8267,
+      "step": 5939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.029456682068129,
+      "learning_rate": 7.114630720279521e-06,
+      "loss": 0.7699,
+      "step": 5940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.508841401442375,
+      "learning_rate": 7.11369144731763e-06,
+      "loss": 0.8274,
+      "step": 5941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.141222964640593,
+      "learning_rate": 7.1127520835257125e-06,
+      "loss": 0.5758,
+      "step": 5942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6770306806831305,
+      "learning_rate": 7.111812628944132e-06,
+      "loss": 0.7967,
+      "step": 5943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.965053037837954,
+      "learning_rate": 7.110873083613259e-06,
+      "loss": 0.7865,
+      "step": 5944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.974771431575786,
+      "learning_rate": 7.109933447573469e-06,
+      "loss": 0.7581,
+      "step": 5945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2379846820726903,
+      "learning_rate": 7.108993720865138e-06,
+      "loss": 0.8654,
+      "step": 5946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1604464911176442,
+      "learning_rate": 7.108053903528652e-06,
+      "loss": 0.6576,
+      "step": 5947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.934635539783559,
+      "learning_rate": 7.107113995604395e-06,
+      "loss": 1.0314,
+      "step": 5948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0500188630000102,
+      "learning_rate": 7.106173997132755e-06,
+      "loss": 0.7165,
+      "step": 5949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.093842624348042,
+      "learning_rate": 7.105233908154128e-06,
+      "loss": 0.7286,
+      "step": 5950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0442235845184187,
+      "learning_rate": 7.104293728708912e-06,
+      "loss": 0.7427,
+      "step": 5951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8376163813736195,
+      "learning_rate": 7.103353458837509e-06,
+      "loss": 0.9255,
+      "step": 5952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0374683215071414,
+      "learning_rate": 7.102413098580322e-06,
+      "loss": 0.7824,
+      "step": 5953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.582464631484896,
+      "learning_rate": 7.101472647977761e-06,
+      "loss": 0.7995,
+      "step": 5954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9238773168512682,
+      "learning_rate": 7.100532107070242e-06,
+      "loss": 0.5445,
+      "step": 5955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2595424197579628,
+      "learning_rate": 7.099591475898178e-06,
+      "loss": 1.0082,
+      "step": 5956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1724781412515655,
+      "learning_rate": 7.098650754501994e-06,
+      "loss": 0.7273,
+      "step": 5957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8880522387127194,
+      "learning_rate": 7.0977099429221145e-06,
+      "loss": 1.0825,
+      "step": 5958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.133070244112314,
+      "learning_rate": 7.096769041198964e-06,
+      "loss": 0.8747,
+      "step": 5959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0710992140514364,
+      "learning_rate": 7.095828049372979e-06,
+      "loss": 0.8117,
+      "step": 5960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.406945032032459,
+      "learning_rate": 7.0948869674845955e-06,
+      "loss": 0.9287,
+      "step": 5961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4965697183682707,
+      "learning_rate": 7.093945795574254e-06,
+      "loss": 0.8885,
+      "step": 5962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4129785581481205,
+      "learning_rate": 7.093004533682397e-06,
+      "loss": 0.8103,
+      "step": 5963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1927233365817633,
+      "learning_rate": 7.0920631818494745e-06,
+      "loss": 0.9114,
+      "step": 5964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4856338500936634,
+      "learning_rate": 7.091121740115936e-06,
+      "loss": 0.7711,
+      "step": 5965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3468269182332424,
+      "learning_rate": 7.090180208522241e-06,
+      "loss": 0.8884,
+      "step": 5966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.968839088492081,
+      "learning_rate": 7.089238587108849e-06,
+      "loss": 0.7546,
+      "step": 5967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6684528633750655,
+      "learning_rate": 7.08829687591622e-06,
+      "loss": 0.7189,
+      "step": 5968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3475149783593845,
+      "learning_rate": 7.087355074984823e-06,
+      "loss": 0.8467,
+      "step": 5969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6056631851508496,
+      "learning_rate": 7.086413184355129e-06,
+      "loss": 0.7971,
+      "step": 5970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2403871969994902,
+      "learning_rate": 7.085471204067616e-06,
+      "loss": 0.7561,
+      "step": 5971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.519556238113195,
+      "learning_rate": 7.084529134162759e-06,
+      "loss": 1.0814,
+      "step": 5972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1721511652236147,
+      "learning_rate": 7.083586974681042e-06,
+      "loss": 0.799,
+      "step": 5973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8815060084791644,
+      "learning_rate": 7.082644725662954e-06,
+      "loss": 0.7795,
+      "step": 5974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.083013584537894,
+      "learning_rate": 7.0817023871489835e-06,
+      "loss": 0.6197,
+      "step": 5975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.329590338992957,
+      "learning_rate": 7.080759959179626e-06,
+      "loss": 0.8132,
+      "step": 5976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1030862249050846,
+      "learning_rate": 7.079817441795378e-06,
+      "loss": 0.8817,
+      "step": 5977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0647291389053002,
+      "learning_rate": 7.078874835036742e-06,
+      "loss": 0.6474,
+      "step": 5978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3251068623128575,
+      "learning_rate": 7.077932138944225e-06,
+      "loss": 0.6941,
+      "step": 5979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3047253168493294,
+      "learning_rate": 7.076989353558335e-06,
+      "loss": 0.8655,
+      "step": 5980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.117379906203778,
+      "learning_rate": 7.076046478919589e-06,
+      "loss": 0.9153,
+      "step": 5981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8784931042205169,
+      "learning_rate": 7.075103515068501e-06,
+      "loss": 0.9171,
+      "step": 5982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.091265010044141,
+      "learning_rate": 7.074160462045593e-06,
+      "loss": 0.6591,
+      "step": 5983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2193309671604613,
+      "learning_rate": 7.073217319891391e-06,
+      "loss": 0.8405,
+      "step": 5984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3113189999796053,
+      "learning_rate": 7.072274088646425e-06,
+      "loss": 1.0578,
+      "step": 5985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.741910848988494,
+      "learning_rate": 7.071330768351223e-06,
+      "loss": 0.8041,
+      "step": 5986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.11550015092832,
+      "learning_rate": 7.070387359046328e-06,
+      "loss": 0.7755,
+      "step": 5987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.076406516971787,
+      "learning_rate": 7.0694438607722746e-06,
+      "loss": 0.9943,
+      "step": 5988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.593654803014013,
+      "learning_rate": 7.068500273569612e-06,
+      "loss": 0.94,
+      "step": 5989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.017165765413488,
+      "learning_rate": 7.067556597478884e-06,
+      "loss": 0.6865,
+      "step": 5990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2704356353380564,
+      "learning_rate": 7.066612832540644e-06,
+      "loss": 0.6902,
+      "step": 5991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.468784874870985,
+      "learning_rate": 7.065668978795449e-06,
+      "loss": 0.81,
+      "step": 5992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5472850055087153,
+      "learning_rate": 7.064725036283857e-06,
+      "loss": 0.7938,
+      "step": 5993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7912927559186944,
+      "learning_rate": 7.063781005046433e-06,
+      "loss": 0.7436,
+      "step": 5994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.350630910422988,
+      "learning_rate": 7.0628368851237424e-06,
+      "loss": 0.8566,
+      "step": 5995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1359989076152144,
+      "learning_rate": 7.061892676556357e-06,
+      "loss": 0.9257,
+      "step": 5996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0886703680424477,
+      "learning_rate": 7.06094837938485e-06,
+      "loss": 0.7691,
+      "step": 5997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2092029793105854,
+      "learning_rate": 7.060003993649802e-06,
+      "loss": 0.921,
+      "step": 5998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.568673546130124,
+      "learning_rate": 7.059059519391794e-06,
+      "loss": 0.8001,
+      "step": 5999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1984263999378457,
+      "learning_rate": 7.058114956651413e-06,
+      "loss": 0.6492,
+      "step": 6000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9155322563403496,
+      "learning_rate": 7.057170305469248e-06,
+      "loss": 0.7164,
+      "step": 6001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0516341003663137,
+      "learning_rate": 7.056225565885895e-06,
+      "loss": 0.647,
+      "step": 6002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6561296282655384,
+      "learning_rate": 7.055280737941949e-06,
+      "loss": 0.8906,
+      "step": 6003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0781130823475045,
+      "learning_rate": 7.054335821678012e-06,
+      "loss": 1.0,
+      "step": 6004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8289361770236128,
+      "learning_rate": 7.05339081713469e-06,
+      "loss": 0.8266,
+      "step": 6005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.437794403917964,
+      "learning_rate": 7.05244572435259e-06,
+      "loss": 0.8805,
+      "step": 6006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.148561744683384,
+      "learning_rate": 7.051500543372328e-06,
+      "loss": 0.8119,
+      "step": 6007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1301083370475453,
+      "learning_rate": 7.050555274234518e-06,
+      "loss": 0.7988,
+      "step": 6008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9609732592340159,
+      "learning_rate": 7.049609916979782e-06,
+      "loss": 0.7902,
+      "step": 6009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.127238352588408,
+      "learning_rate": 7.04866447164874e-06,
+      "loss": 0.6065,
+      "step": 6010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0394239520823914,
+      "learning_rate": 7.0477189382820266e-06,
+      "loss": 0.6954,
+      "step": 6011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.311570994756949,
+      "learning_rate": 7.046773316920269e-06,
+      "loss": 0.8143,
+      "step": 6012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2648143140971446,
+      "learning_rate": 7.045827607604103e-06,
+      "loss": 0.9308,
+      "step": 6013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.572348612846724,
+      "learning_rate": 7.044881810374169e-06,
+      "loss": 0.8635,
+      "step": 6014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.116114844487457,
+      "learning_rate": 7.043935925271108e-06,
+      "loss": 0.6824,
+      "step": 6015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.259688450572108,
+      "learning_rate": 7.042989952335571e-06,
+      "loss": 0.6119,
+      "step": 6016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.309841724361692,
+      "learning_rate": 7.0420438916082024e-06,
+      "loss": 0.8444,
+      "step": 6017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.006884724256038,
+      "learning_rate": 7.041097743129662e-06,
+      "loss": 0.8487,
+      "step": 6018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.008879162693564,
+      "learning_rate": 7.040151506940605e-06,
+      "loss": 0.763,
+      "step": 6019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.293521525660298,
+      "learning_rate": 7.039205183081694e-06,
+      "loss": 0.9464,
+      "step": 6020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.287985632664134,
+      "learning_rate": 7.038258771593596e-06,
+      "loss": 0.8979,
+      "step": 6021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1270948083105843,
+      "learning_rate": 7.037312272516977e-06,
+      "loss": 0.7324,
+      "step": 6022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3517838534565003,
+      "learning_rate": 7.036365685892515e-06,
+      "loss": 0.8393,
+      "step": 6023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8656621137421925,
+      "learning_rate": 7.035419011760882e-06,
+      "loss": 0.8398,
+      "step": 6024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.191121862383731,
+      "learning_rate": 7.0344722501627625e-06,
+      "loss": 0.7486,
+      "step": 6025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2101850586874108,
+      "learning_rate": 7.033525401138838e-06,
+      "loss": 0.7849,
+      "step": 6026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.384783173210846,
+      "learning_rate": 7.0325784647298e-06,
+      "loss": 0.8123,
+      "step": 6027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.048032860858705,
+      "learning_rate": 7.031631440976336e-06,
+      "loss": 0.7416,
+      "step": 6028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8366113503250305,
+      "learning_rate": 7.0306843299191465e-06,
+      "loss": 0.8069,
+      "step": 6029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9040869676077996,
+      "learning_rate": 7.0297371315989284e-06,
+      "loss": 0.9286,
+      "step": 6030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.757469438040376,
+      "learning_rate": 7.028789846056383e-06,
+      "loss": 0.9151,
+      "step": 6031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7773568340730557,
+      "learning_rate": 7.0278424733322225e-06,
+      "loss": 0.9891,
+      "step": 6032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.031143535452242,
+      "learning_rate": 7.026895013467153e-06,
+      "loss": 0.8515,
+      "step": 6033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0165667270406757,
+      "learning_rate": 7.0259474665018915e-06,
+      "loss": 0.7887,
+      "step": 6034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0224768396679527,
+      "learning_rate": 7.024999832477154e-06,
+      "loss": 0.5662,
+      "step": 6035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4062542094794503,
+      "learning_rate": 7.0240521114336655e-06,
+      "loss": 1.0089,
+      "step": 6036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2583925920753187,
+      "learning_rate": 7.023104303412149e-06,
+      "loss": 0.7527,
+      "step": 6037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.208449190749185,
+      "learning_rate": 7.022156408453336e-06,
+      "loss": 0.7307,
+      "step": 6038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7305805133908532,
+      "learning_rate": 7.0212084265979575e-06,
+      "loss": 0.6572,
+      "step": 6039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.981687033672612,
+      "learning_rate": 7.02026035788675e-06,
+      "loss": 0.8934,
+      "step": 6040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3416554997905092,
+      "learning_rate": 7.019312202360457e-06,
+      "loss": 0.827,
+      "step": 6041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5296207145408736,
+      "learning_rate": 7.018363960059823e-06,
+      "loss": 0.906,
+      "step": 6042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.321621146202022,
+      "learning_rate": 7.017415631025593e-06,
+      "loss": 0.9235,
+      "step": 6043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8191393722120939,
+      "learning_rate": 7.016467215298519e-06,
+      "loss": 0.8843,
+      "step": 6044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8590758627965354,
+      "learning_rate": 7.01551871291936e-06,
+      "loss": 0.8228,
+      "step": 6045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3717693574554777,
+      "learning_rate": 7.014570123928872e-06,
+      "loss": 0.7701,
+      "step": 6046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2668067380688595,
+      "learning_rate": 7.01362144836782e-06,
+      "loss": 0.6054,
+      "step": 6047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.921136859807103,
+      "learning_rate": 7.012672686276969e-06,
+      "loss": 0.7336,
+      "step": 6048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.635327681898881,
+      "learning_rate": 7.011723837697091e-06,
+      "loss": 0.8841,
+      "step": 6049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.173226949791364,
+      "learning_rate": 7.010774902668958e-06,
+      "loss": 0.726,
+      "step": 6050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1528475362277013,
+      "learning_rate": 7.009825881233351e-06,
+      "loss": 0.9951,
+      "step": 6051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0453690001535554,
+      "learning_rate": 7.008876773431048e-06,
+      "loss": 0.7638,
+      "step": 6052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9548321025914714,
+      "learning_rate": 7.007927579302837e-06,
+      "loss": 0.8434,
+      "step": 6053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0578506260857488,
+      "learning_rate": 7.0069782988895056e-06,
+      "loss": 0.8437,
+      "step": 6054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2532208242406155,
+      "learning_rate": 7.006028932231847e-06,
+      "loss": 0.7759,
+      "step": 6055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.710409147459792,
+      "learning_rate": 7.005079479370658e-06,
+      "loss": 0.8296,
+      "step": 6056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.376425792385226,
+      "learning_rate": 7.0041299403467365e-06,
+      "loss": 0.7162,
+      "step": 6057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9968572840537346,
+      "learning_rate": 7.003180315200891e-06,
+      "loss": 0.6744,
+      "step": 6058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.218583149368076,
+      "learning_rate": 7.002230603973924e-06,
+      "loss": 0.9144,
+      "step": 6059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.139257036021591,
+      "learning_rate": 7.001280806706649e-06,
+      "loss": 0.744,
+      "step": 6060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0272789000210962,
+      "learning_rate": 7.0003309234398805e-06,
+      "loss": 0.8674,
+      "step": 6061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.062679132807201,
+      "learning_rate": 6.999380954214438e-06,
+      "loss": 0.8633,
+      "step": 6062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9064752932923685,
+      "learning_rate": 6.9984308990711425e-06,
+      "loss": 0.9256,
+      "step": 6063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.576139460578992,
+      "learning_rate": 6.9974807580508205e-06,
+      "loss": 0.8007,
+      "step": 6064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.36531480558515,
+      "learning_rate": 6.996530531194302e-06,
+      "loss": 0.8096,
+      "step": 6065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0519519100640677,
+      "learning_rate": 6.995580218542419e-06,
+      "loss": 0.6385,
+      "step": 6066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9838808498561078,
+      "learning_rate": 6.994629820136011e-06,
+      "loss": 0.7108,
+      "step": 6067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.076199578112525,
+      "learning_rate": 6.9936793360159175e-06,
+      "loss": 0.8444,
+      "step": 6068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.167047958578297,
+      "learning_rate": 6.992728766222982e-06,
+      "loss": 0.7815,
+      "step": 6069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1932229654058788,
+      "learning_rate": 6.991778110798053e-06,
+      "loss": 0.7619,
+      "step": 6070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3543374724003625,
+      "learning_rate": 6.990827369781983e-06,
+      "loss": 0.7409,
+      "step": 6071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.978737292475163,
+      "learning_rate": 6.9898765432156266e-06,
+      "loss": 0.719,
+      "step": 6072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.60723555574351,
+      "learning_rate": 6.988925631139843e-06,
+      "loss": 0.8509,
+      "step": 6073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.321338359903755,
+      "learning_rate": 6.987974633595498e-06,
+      "loss": 0.7341,
+      "step": 6074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.954184542796795,
+      "learning_rate": 6.987023550623453e-06,
+      "loss": 0.9096,
+      "step": 6075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2387087274439605,
+      "learning_rate": 6.9860723822645825e-06,
+      "loss": 0.8199,
+      "step": 6076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8922400765203014,
+      "learning_rate": 6.985121128559757e-06,
+      "loss": 0.8664,
+      "step": 6077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2646576553830973,
+      "learning_rate": 6.984169789549857e-06,
+      "loss": 0.5986,
+      "step": 6078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9564513807554085,
+      "learning_rate": 6.9832183652757625e-06,
+      "loss": 0.7998,
+      "step": 6079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1857571388682557,
+      "learning_rate": 6.982266855778356e-06,
+      "loss": 0.8169,
+      "step": 6080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8050742303521008,
+      "learning_rate": 6.981315261098529e-06,
+      "loss": 0.6835,
+      "step": 6081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.020378541375612,
+      "learning_rate": 6.980363581277172e-06,
+      "loss": 0.7422,
+      "step": 6082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7197269521348948,
+      "learning_rate": 6.979411816355183e-06,
+      "loss": 0.8742,
+      "step": 6083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1748047615373853,
+      "learning_rate": 6.978459966373458e-06,
+      "loss": 0.5789,
+      "step": 6084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.707078544100491,
+      "learning_rate": 6.9775080313729026e-06,
+      "loss": 0.8884,
+      "step": 6085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0392057724556016,
+      "learning_rate": 6.9765560113944235e-06,
+      "loss": 1.0019,
+      "step": 6086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8359287402522462,
+      "learning_rate": 6.9756039064789295e-06,
+      "loss": 0.8368,
+      "step": 6087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1012882483940687,
+      "learning_rate": 6.9746517166673375e-06,
+      "loss": 0.8745,
+      "step": 6088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.437798838692337,
+      "learning_rate": 6.973699442000561e-06,
+      "loss": 0.7809,
+      "step": 6089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.500099787795502,
+      "learning_rate": 6.972747082519526e-06,
+      "loss": 0.8626,
+      "step": 6090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9012858010136875,
+      "learning_rate": 6.971794638265153e-06,
+      "loss": 0.7273,
+      "step": 6091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1046887519088697,
+      "learning_rate": 6.970842109278375e-06,
+      "loss": 0.9559,
+      "step": 6092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2072627497483293,
+      "learning_rate": 6.969889495600122e-06,
+      "loss": 0.7472,
+      "step": 6093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.168169838424484,
+      "learning_rate": 6.96893679727133e-06,
+      "loss": 0.7256,
+      "step": 6094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0385381126995943,
+      "learning_rate": 6.967984014332938e-06,
+      "loss": 0.7031,
+      "step": 6095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3624901311463873,
+      "learning_rate": 6.96703114682589e-06,
+      "loss": 0.804,
+      "step": 6096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.337342593495474,
+      "learning_rate": 6.966078194791133e-06,
+      "loss": 0.7478,
+      "step": 6097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.83820037536143,
+      "learning_rate": 6.965125158269619e-06,
+      "loss": 0.7849,
+      "step": 6098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9313606019330345,
+      "learning_rate": 6.9641720373022996e-06,
+      "loss": 0.9448,
+      "step": 6099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0877376332357467,
+      "learning_rate": 6.963218831930132e-06,
+      "loss": 1.0096,
+      "step": 6100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.141340791516158,
+      "learning_rate": 6.96226554219408e-06,
+      "loss": 0.6493,
+      "step": 6101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.293148580937075,
+      "learning_rate": 6.961312168135107e-06,
+      "loss": 0.8735,
+      "step": 6102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0731123656139323,
+      "learning_rate": 6.960358709794182e-06,
+      "loss": 0.6709,
+      "step": 6103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.237294324498977,
+      "learning_rate": 6.959405167212278e-06,
+      "loss": 0.8009,
+      "step": 6104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.014805659144825,
+      "learning_rate": 6.9584515404303685e-06,
+      "loss": 0.7323,
+      "step": 6105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9908309779181037,
+      "learning_rate": 6.9574978294894355e-06,
+      "loss": 0.5673,
+      "step": 6106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0988084216375082,
+      "learning_rate": 6.956544034430463e-06,
+      "loss": 0.7083,
+      "step": 6107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6352468794645025,
+      "learning_rate": 6.955590155294434e-06,
+      "loss": 0.7851,
+      "step": 6108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.209151426415368,
+      "learning_rate": 6.954636192122339e-06,
+      "loss": 0.9258,
+      "step": 6109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.924705982153521,
+      "learning_rate": 6.953682144955176e-06,
+      "loss": 0.8364,
+      "step": 6110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.689433014773511,
+      "learning_rate": 6.952728013833941e-06,
+      "loss": 0.8582,
+      "step": 6111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1748709051501582,
+      "learning_rate": 6.951773798799632e-06,
+      "loss": 0.8551,
+      "step": 6112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.724279362470364,
+      "learning_rate": 6.950819499893257e-06,
+      "loss": 0.7461,
+      "step": 6113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7894667181234352,
+      "learning_rate": 6.949865117155823e-06,
+      "loss": 0.8152,
+      "step": 6114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3892961373698505,
+      "learning_rate": 6.948910650628343e-06,
+      "loss": 0.8399,
+      "step": 6115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1087513375861917,
+      "learning_rate": 6.947956100351833e-06,
+      "loss": 0.8533,
+      "step": 6116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5328694735731054,
+      "learning_rate": 6.947001466367309e-06,
+      "loss": 0.9403,
+      "step": 6117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.981603323291565,
+      "learning_rate": 6.946046748715796e-06,
+      "loss": 0.9042,
+      "step": 6118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6691199631386557,
+      "learning_rate": 6.94509194743832e-06,
+      "loss": 0.7964,
+      "step": 6119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6593405209421648,
+      "learning_rate": 6.944137062575913e-06,
+      "loss": 0.7621,
+      "step": 6120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7144975299210965,
+      "learning_rate": 6.943182094169605e-06,
+      "loss": 0.6982,
+      "step": 6121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2496187382192634,
+      "learning_rate": 6.942227042260435e-06,
+      "loss": 0.7637,
+      "step": 6122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0777381795257015,
+      "learning_rate": 6.941271906889443e-06,
+      "loss": 0.9094,
+      "step": 6123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.765877927598399,
+      "learning_rate": 6.940316688097675e-06,
+      "loss": 0.7576,
+      "step": 6124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.034128714262635,
+      "learning_rate": 6.9393613859261755e-06,
+      "loss": 0.7324,
+      "step": 6125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5052774579456774,
+      "learning_rate": 6.938406000416e-06,
+      "loss": 0.7777,
+      "step": 6126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1564974884590438,
+      "learning_rate": 6.9374505316082e-06,
+      "loss": 0.7876,
+      "step": 6127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4100651835632996,
+      "learning_rate": 6.936494979543836e-06,
+      "loss": 0.9025,
+      "step": 6128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2782254790716936,
+      "learning_rate": 6.935539344263971e-06,
+      "loss": 0.8237,
+      "step": 6129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7950782296287344,
+      "learning_rate": 6.93458362580967e-06,
+      "loss": 0.7842,
+      "step": 6130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9941168674082297,
+      "learning_rate": 6.9336278242220015e-06,
+      "loss": 0.6628,
+      "step": 6131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8583974329051172,
+      "learning_rate": 6.932671939542037e-06,
+      "loss": 0.9055,
+      "step": 6132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3213744553585687,
+      "learning_rate": 6.9317159718108575e-06,
+      "loss": 0.8971,
+      "step": 6133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4454955748518006,
+      "learning_rate": 6.93075992106954e-06,
+      "loss": 0.9686,
+      "step": 6134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.164959466478023,
+      "learning_rate": 6.929803787359169e-06,
+      "loss": 0.7712,
+      "step": 6135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.624384573512747,
+      "learning_rate": 6.928847570720829e-06,
+      "loss": 0.8367,
+      "step": 6136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.101578453982976,
+      "learning_rate": 6.927891271195616e-06,
+      "loss": 0.9293,
+      "step": 6137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0751010738881754,
+      "learning_rate": 6.926934888824619e-06,
+      "loss": 0.6125,
+      "step": 6138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2534516685476715,
+      "learning_rate": 6.925978423648941e-06,
+      "loss": 0.6955,
+      "step": 6139
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6710116560881922,
+      "learning_rate": 6.92502187570968e-06,
+      "loss": 0.9409,
+      "step": 6140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.74595173163185,
+      "learning_rate": 6.92406524504794e-06,
+      "loss": 0.9297,
+      "step": 6141
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.220132095175088,
+      "learning_rate": 6.923108531704832e-06,
+      "loss": 0.7303,
+      "step": 6142
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8746639389272808,
+      "learning_rate": 6.9221517357214674e-06,
+      "loss": 0.7342,
+      "step": 6143
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.669826936957494,
+      "learning_rate": 6.921194857138963e-06,
+      "loss": 0.8458,
+      "step": 6144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8564837565517638,
+      "learning_rate": 6.920237895998436e-06,
+      "loss": 0.7793,
+      "step": 6145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0127357709287246,
+      "learning_rate": 6.919280852341011e-06,
+      "loss": 0.8368,
+      "step": 6146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2379938040729976,
+      "learning_rate": 6.91832372620781e-06,
+      "loss": 0.5675,
+      "step": 6147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7406807323375402,
+      "learning_rate": 6.917366517639969e-06,
+      "loss": 0.7375,
+      "step": 6148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.115921911923104,
+      "learning_rate": 6.91640922667862e-06,
+      "loss": 0.8111,
+      "step": 6149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5545533531033926,
+      "learning_rate": 6.9154518533648966e-06,
+      "loss": 0.8435,
+      "step": 6150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0958457206870666,
+      "learning_rate": 6.91449439773994e-06,
+      "loss": 0.7987,
+      "step": 6151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.057232586078699,
+      "learning_rate": 6.913536859844896e-06,
+      "loss": 0.7769,
+      "step": 6152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1496502581779122,
+      "learning_rate": 6.912579239720913e-06,
+      "loss": 0.5892,
+      "step": 6153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.412231563949267,
+      "learning_rate": 6.911621537409139e-06,
+      "loss": 0.9012,
+      "step": 6154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.240007106689766,
+      "learning_rate": 6.9106637529507305e-06,
+      "loss": 0.7255,
+      "step": 6155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1665420079613957,
+      "learning_rate": 6.909705886386844e-06,
+      "loss": 0.7855,
+      "step": 6156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9303352917151426,
+      "learning_rate": 6.908747937758644e-06,
+      "loss": 0.8125,
+      "step": 6157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3713596470804603,
+      "learning_rate": 6.9077899071072954e-06,
+      "loss": 0.7471,
+      "step": 6158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.019011079561419,
+      "learning_rate": 6.906831794473963e-06,
+      "loss": 0.6754,
+      "step": 6159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.723108529886098,
+      "learning_rate": 6.9058735998998224e-06,
+      "loss": 0.6601,
+      "step": 6160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4727824655299173,
+      "learning_rate": 6.9049153234260495e-06,
+      "loss": 0.8095,
+      "step": 6161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.190739490444295,
+      "learning_rate": 6.903956965093821e-06,
+      "loss": 0.7065,
+      "step": 6162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.011571706611777,
+      "learning_rate": 6.902998524944322e-06,
+      "loss": 0.9461,
+      "step": 6163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.983052536402335,
+      "learning_rate": 6.9020400030187394e-06,
+      "loss": 0.8372,
+      "step": 6164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2576083574392,
+      "learning_rate": 6.90108139935826e-06,
+      "loss": 0.7738,
+      "step": 6165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5771602855156925,
+      "learning_rate": 6.90012271400408e-06,
+      "loss": 0.8709,
+      "step": 6166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.288100844621224,
+      "learning_rate": 6.899163946997396e-06,
+      "loss": 0.7071,
+      "step": 6167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1027660015640866,
+      "learning_rate": 6.898205098379407e-06,
+      "loss": 0.8188,
+      "step": 6168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6630822551768745,
+      "learning_rate": 6.897246168191317e-06,
+      "loss": 0.8916,
+      "step": 6169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0929296411797425,
+      "learning_rate": 6.8962871564743325e-06,
+      "loss": 0.9039,
+      "step": 6170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2954625545810512,
+      "learning_rate": 6.895328063269667e-06,
+      "loss": 0.7462,
+      "step": 6171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1751945995448518,
+      "learning_rate": 6.894368888618534e-06,
+      "loss": 0.6354,
+      "step": 6172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0227130604550811,
+      "learning_rate": 6.8934096325621505e-06,
+      "loss": 0.6897,
+      "step": 6173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9731142655695018,
+      "learning_rate": 6.892450295141737e-06,
+      "loss": 0.9141,
+      "step": 6174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8232038999079156,
+      "learning_rate": 6.891490876398521e-06,
+      "loss": 0.8977,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9992242775666815,
+      "learning_rate": 6.890531376373729e-06,
+      "loss": 1.0653,
+      "step": 6176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7952132165552837,
+      "learning_rate": 6.889571795108594e-06,
+      "loss": 0.8797,
+      "step": 6177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5051197587477683,
+      "learning_rate": 6.88861213264435e-06,
+      "loss": 0.7861,
+      "step": 6178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9129401372203332,
+      "learning_rate": 6.887652389022236e-06,
+      "loss": 0.8605,
+      "step": 6179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4903827218438614,
+      "learning_rate": 6.886692564283496e-06,
+      "loss": 0.7894,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8964975709240983,
+      "learning_rate": 6.885732658469374e-06,
+      "loss": 0.849,
+      "step": 6181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.437298155928612,
+      "learning_rate": 6.88477267162112e-06,
+      "loss": 0.8823,
+      "step": 6182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0866282101760287,
+      "learning_rate": 6.883812603779985e-06,
+      "loss": 0.7949,
+      "step": 6183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7244458963791838,
+      "learning_rate": 6.88285245498723e-06,
+      "loss": 0.6764,
+      "step": 6184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9403604378034376,
+      "learning_rate": 6.88189222528411e-06,
+      "loss": 0.8227,
+      "step": 6185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.277938996679647,
+      "learning_rate": 6.8809319147118905e-06,
+      "loss": 0.9918,
+      "step": 6186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9590616202526714,
+      "learning_rate": 6.879971523311837e-06,
+      "loss": 0.7964,
+      "step": 6187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5556397271059144,
+      "learning_rate": 6.8790110511252195e-06,
+      "loss": 0.864,
+      "step": 6188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7409436643848386,
+      "learning_rate": 6.878050498193314e-06,
+      "loss": 0.898,
+      "step": 6189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1095394852739715,
+      "learning_rate": 6.877089864557395e-06,
+      "loss": 0.8832,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.519547021884957,
+      "learning_rate": 6.876129150258744e-06,
+      "loss": 1.0594,
+      "step": 6191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1949852683847675,
+      "learning_rate": 6.875168355338645e-06,
+      "loss": 0.5864,
+      "step": 6192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2902241177054723,
+      "learning_rate": 6.874207479838386e-06,
+      "loss": 0.8531,
+      "step": 6193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.945364210660777,
+      "learning_rate": 6.873246523799256e-06,
+      "loss": 0.9339,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.371004945993436,
+      "learning_rate": 6.872285487262555e-06,
+      "loss": 0.6935,
+      "step": 6195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.540211603846563,
+      "learning_rate": 6.871324370269573e-06,
+      "loss": 0.7937,
+      "step": 6196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8176993788431188,
+      "learning_rate": 6.870363172861616e-06,
+      "loss": 0.7395,
+      "step": 6197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1265368983569957,
+      "learning_rate": 6.869401895079988e-06,
+      "loss": 0.7178,
+      "step": 6198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7175857176857734,
+      "learning_rate": 6.868440536965997e-06,
+      "loss": 0.8749,
+      "step": 6199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.2145702804913014,
+      "learning_rate": 6.867479098560956e-06,
+      "loss": 0.7838,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.459555368283475,
+      "learning_rate": 6.866517579906178e-06,
+      "loss": 0.8338,
+      "step": 6201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.958864359157779,
+      "learning_rate": 6.865555981042983e-06,
+      "loss": 0.9066,
+      "step": 6202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9376892655512132,
+      "learning_rate": 6.864594302012691e-06,
+      "loss": 0.6897,
+      "step": 6203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.194003799139881,
+      "learning_rate": 6.863632542856632e-06,
+      "loss": 1.0062,
+      "step": 6204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1866524437855936,
+      "learning_rate": 6.8626707036161314e-06,
+      "loss": 0.7853,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9214999124484728,
+      "learning_rate": 6.86170878433252e-06,
+      "loss": 0.7834,
+      "step": 6206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0442230917496953,
+      "learning_rate": 6.860746785047139e-06,
+      "loss": 0.7405,
+      "step": 6207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.659895688914554,
+      "learning_rate": 6.859784705801323e-06,
+      "loss": 0.7665,
+      "step": 6208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.071400745115418,
+      "learning_rate": 6.858822546636417e-06,
+      "loss": 0.6353,
+      "step": 6209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8968107517181994,
+      "learning_rate": 6.857860307593764e-06,
+      "loss": 0.7662,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2974572330445224,
+      "learning_rate": 6.856897988714718e-06,
+      "loss": 0.7639,
+      "step": 6211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1615947300325433,
+      "learning_rate": 6.855935590040629e-06,
+      "loss": 0.7834,
+      "step": 6212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.984492867073395,
+      "learning_rate": 6.854973111612855e-06,
+      "loss": 0.9076,
+      "step": 6213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.588677268206023,
+      "learning_rate": 6.854010553472757e-06,
+      "loss": 0.8621,
+      "step": 6214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0731533352174383,
+      "learning_rate": 6.8530479156616925e-06,
+      "loss": 1.0237,
+      "step": 6215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.509437919018239,
+      "learning_rate": 6.852085198221035e-06,
+      "loss": 0.7881,
+      "step": 6216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.094416782018002,
+      "learning_rate": 6.851122401192149e-06,
+      "loss": 0.8555,
+      "step": 6217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6231176260713693,
+      "learning_rate": 6.850159524616413e-06,
+      "loss": 0.6341,
+      "step": 6218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.751884694175054,
+      "learning_rate": 6.849196568535201e-06,
+      "loss": 0.7189,
+      "step": 6219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.15036539204692,
+      "learning_rate": 6.848233532989894e-06,
+      "loss": 0.8752,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.048189532006801,
+      "learning_rate": 6.847270418021876e-06,
+      "loss": 0.9692,
+      "step": 6221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.150382518957472,
+      "learning_rate": 6.8463072236725345e-06,
+      "loss": 0.8112,
+      "step": 6222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.246681132996241,
+      "learning_rate": 6.845343949983258e-06,
+      "loss": 0.7812,
+      "step": 6223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9890007638679332,
+      "learning_rate": 6.8443805969954445e-06,
+      "loss": 0.9746,
+      "step": 6224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.335793810075161,
+      "learning_rate": 6.8434171647504885e-06,
+      "loss": 0.9888,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9508903986635933,
+      "learning_rate": 6.84245365328979e-06,
+      "loss": 0.9445,
+      "step": 6226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1594391451562176,
+      "learning_rate": 6.841490062654757e-06,
+      "loss": 0.7032,
+      "step": 6227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0036666004971417,
+      "learning_rate": 6.840526392886793e-06,
+      "loss": 0.8968,
+      "step": 6228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9608698068042139,
+      "learning_rate": 6.839562644027311e-06,
+      "loss": 0.7814,
+      "step": 6229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2133098092103083,
+      "learning_rate": 6.838598816117725e-06,
+      "loss": 0.8174,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.206892936966067,
+      "learning_rate": 6.837634909199453e-06,
+      "loss": 0.6548,
+      "step": 6231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.844455518339176,
+      "learning_rate": 6.836670923313917e-06,
+      "loss": 0.8059,
+      "step": 6232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9396050339249007,
+      "learning_rate": 6.835706858502542e-06,
+      "loss": 0.9494,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2283658111884186,
+      "learning_rate": 6.834742714806754e-06,
+      "loss": 0.6384,
+      "step": 6234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8316798457373873,
+      "learning_rate": 6.833778492267984e-06,
+      "loss": 0.7205,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1372919025301886,
+      "learning_rate": 6.832814190927669e-06,
+      "loss": 0.7421,
+      "step": 6236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.451810795909319,
+      "learning_rate": 6.831849810827247e-06,
+      "loss": 0.7834,
+      "step": 6237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0484024247011576,
+      "learning_rate": 6.830885352008159e-06,
+      "loss": 0.8918,
+      "step": 6238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6749321501541987,
+      "learning_rate": 6.8299208145118475e-06,
+      "loss": 0.8134,
+      "step": 6239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0178250359448247,
+      "learning_rate": 6.828956198379766e-06,
+      "loss": 0.8107,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.438077081601704,
+      "learning_rate": 6.827991503653363e-06,
+      "loss": 0.7775,
+      "step": 6241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.215512304451096,
+      "learning_rate": 6.827026730374093e-06,
+      "loss": 0.6489,
+      "step": 6242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3848324271169448,
+      "learning_rate": 6.826061878583416e-06,
+      "loss": 0.7307,
+      "step": 6243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1941235753796007,
+      "learning_rate": 6.825096948322791e-06,
+      "loss": 0.8353,
+      "step": 6244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3610043934212803,
+      "learning_rate": 6.824131939633688e-06,
+      "loss": 0.9151,
+      "step": 6245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.975243875311721,
+      "learning_rate": 6.823166852557572e-06,
+      "loss": 0.8762,
+      "step": 6246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.069933718655328,
+      "learning_rate": 6.822201687135916e-06,
+      "loss": 0.7922,
+      "step": 6247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6749680313468933,
+      "learning_rate": 6.821236443410195e-06,
+      "loss": 0.95,
+      "step": 6248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9719956823090741,
+      "learning_rate": 6.820271121421889e-06,
+      "loss": 0.5861,
+      "step": 6249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5730358981949832,
+      "learning_rate": 6.819305721212479e-06,
+      "loss": 0.8638,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.888356841670499,
+      "learning_rate": 6.818340242823449e-06,
+      "loss": 0.6599,
+      "step": 6251
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0287098977374058,
+      "learning_rate": 6.8173746862962904e-06,
+      "loss": 0.5814,
+      "step": 6252
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.209349045665124,
+      "learning_rate": 6.816409051672492e-06,
+      "loss": 0.7865,
+      "step": 6253
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0075066411731495,
+      "learning_rate": 6.815443338993554e-06,
+      "loss": 0.9243,
+      "step": 6254
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9211647278451782,
+      "learning_rate": 6.81447754830097e-06,
+      "loss": 0.7705,
+      "step": 6255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9874670810738986,
+      "learning_rate": 6.813511679636247e-06,
+      "loss": 0.9115,
+      "step": 6256
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.976888759950928,
+      "learning_rate": 6.8125457330408865e-06,
+      "loss": 0.6738,
+      "step": 6257
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9712610091110183,
+      "learning_rate": 6.8115797085564e-06,
+      "loss": 0.6371,
+      "step": 6258
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1412956669819254,
+      "learning_rate": 6.810613606224299e-06,
+      "loss": 0.7233,
+      "step": 6259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.522402410529074,
+      "learning_rate": 6.809647426086098e-06,
+      "loss": 0.8406,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2351175242604846,
+      "learning_rate": 6.808681168183317e-06,
+      "loss": 0.8612,
+      "step": 6261
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3638993496058096,
+      "learning_rate": 6.807714832557479e-06,
+      "loss": 0.7594,
+      "step": 6262
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9789945409663297,
+      "learning_rate": 6.806748419250107e-06,
+      "loss": 0.7036,
+      "step": 6263
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3506515969744495,
+      "learning_rate": 6.805781928302732e-06,
+      "loss": 0.8256,
+      "step": 6264
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.322315867286081,
+      "learning_rate": 6.804815359756887e-06,
+      "loss": 0.7687,
+      "step": 6265
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.127609342026193,
+      "learning_rate": 6.803848713654105e-06,
+      "loss": 0.9344,
+      "step": 6266
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6750975668982453,
+      "learning_rate": 6.8028819900359276e-06,
+      "loss": 0.6665,
+      "step": 6267
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8902726230612426,
+      "learning_rate": 6.801915188943895e-06,
+      "loss": 0.8953,
+      "step": 6268
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.198771304923335,
+      "learning_rate": 6.800948310419554e-06,
+      "loss": 0.8086,
+      "step": 6269
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0543686239911683,
+      "learning_rate": 6.799981354504455e-06,
+      "loss": 0.9616,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8646750454629442,
+      "learning_rate": 6.799014321240145e-06,
+      "loss": 1.0198,
+      "step": 6271
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.099230356789526,
+      "learning_rate": 6.798047210668185e-06,
+      "loss": 0.9422,
+      "step": 6272
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.222752246639794,
+      "learning_rate": 6.797080022830131e-06,
+      "loss": 0.8368,
+      "step": 6273
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5045390361130173,
+      "learning_rate": 6.796112757767547e-06,
+      "loss": 0.8347,
+      "step": 6274
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2501062789874133,
+      "learning_rate": 6.795145415521997e-06,
+      "loss": 0.7035,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.137909531635415,
+      "learning_rate": 6.79417799613505e-06,
+      "loss": 0.9201,
+      "step": 6276
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.667575209097091,
+      "learning_rate": 6.793210499648281e-06,
+      "loss": 0.774,
+      "step": 6277
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8882219459401577,
+      "learning_rate": 6.792242926103261e-06,
+      "loss": 0.9142,
+      "step": 6278
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7437896103165365,
+      "learning_rate": 6.7912752755415716e-06,
+      "loss": 0.7441,
+      "step": 6279
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0683161461935695,
+      "learning_rate": 6.790307548004794e-06,
+      "loss": 0.9291,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2733646255729756,
+      "learning_rate": 6.7893397435345144e-06,
+      "loss": 0.8594,
+      "step": 6281
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7998952706764733,
+      "learning_rate": 6.788371862172319e-06,
+      "loss": 0.8866,
+      "step": 6282
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.942030592831826,
+      "learning_rate": 6.787403903959804e-06,
+      "loss": 0.8755,
+      "step": 6283
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.95785058288242,
+      "learning_rate": 6.786435868938561e-06,
+      "loss": 0.741,
+      "step": 6284
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1333636710313058,
+      "learning_rate": 6.785467757150193e-06,
+      "loss": 0.9461,
+      "step": 6285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.773066258398996,
+      "learning_rate": 6.7844995686362955e-06,
+      "loss": 0.9322,
+      "step": 6286
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.682586323376777,
+      "learning_rate": 6.7835313034384786e-06,
+      "loss": 0.7745,
+      "step": 6287
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8209268067320523,
+      "learning_rate": 6.782562961598349e-06,
+      "loss": 0.7464,
+      "step": 6288
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5908290838854984,
+      "learning_rate": 6.78159454315752e-06,
+      "loss": 0.7633,
+      "step": 6289
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0250185322453387,
+      "learning_rate": 6.780626048157606e-06,
+      "loss": 0.9166,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0206232687536292,
+      "learning_rate": 6.779657476640222e-06,
+      "loss": 0.6436,
+      "step": 6291
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4645231851625313,
+      "learning_rate": 6.778688828646996e-06,
+      "loss": 0.902,
+      "step": 6292
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.477451518712588,
+      "learning_rate": 6.777720104219548e-06,
+      "loss": 0.7261,
+      "step": 6293
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8016082375482956,
+      "learning_rate": 6.776751303399509e-06,
+      "loss": 0.8493,
+      "step": 6294
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1010002830677306,
+      "learning_rate": 6.775782426228509e-06,
+      "loss": 0.7323,
+      "step": 6295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.576062679580143,
+      "learning_rate": 6.774813472748183e-06,
+      "loss": 0.8259,
+      "step": 6296
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9410794272284082,
+      "learning_rate": 6.773844443000169e-06,
+      "loss": 0.9595,
+      "step": 6297
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1948102926033894,
+      "learning_rate": 6.77287533702611e-06,
+      "loss": 0.9558,
+      "step": 6298
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.001597414603817,
+      "learning_rate": 6.771906154867649e-06,
+      "loss": 0.9264,
+      "step": 6299
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.071195595066484,
+      "learning_rate": 6.770936896566434e-06,
+      "loss": 0.8298,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7027499800159966,
+      "learning_rate": 6.769967562164116e-06,
+      "loss": 0.8179,
+      "step": 6301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1424653183042535,
+      "learning_rate": 6.768998151702351e-06,
+      "loss": 0.6333,
+      "step": 6302
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1339626265614224,
+      "learning_rate": 6.768028665222795e-06,
+      "loss": 0.6224,
+      "step": 6303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.255072739255987,
+      "learning_rate": 6.767059102767109e-06,
+      "loss": 0.7208,
+      "step": 6304
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.190165626038479,
+      "learning_rate": 6.766089464376958e-06,
+      "loss": 0.8868,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1156910287562156,
+      "learning_rate": 6.76511975009401e-06,
+      "loss": 0.808,
+      "step": 6306
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9134217011102463,
+      "learning_rate": 6.7641499599599355e-06,
+      "loss": 0.7897,
+      "step": 6307
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.419614292088881,
+      "learning_rate": 6.763180094016409e-06,
+      "loss": 0.7268,
+      "step": 6308
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.693936077306741,
+      "learning_rate": 6.7622101523051045e-06,
+      "loss": 0.9241,
+      "step": 6309
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.378960429375087,
+      "learning_rate": 6.761240134867706e-06,
+      "loss": 0.7483,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8868788943244532,
+      "learning_rate": 6.760270041745898e-06,
+      "loss": 0.7457,
+      "step": 6311
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.05165870116401,
+      "learning_rate": 6.759299872981364e-06,
+      "loss": 0.9192,
+      "step": 6312
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9237515693491944,
+      "learning_rate": 6.758329628615796e-06,
+      "loss": 0.716,
+      "step": 6313
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0523061197595833,
+      "learning_rate": 6.757359308690889e-06,
+      "loss": 0.8512,
+      "step": 6314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.304323819749558,
+      "learning_rate": 6.756388913248339e-06,
+      "loss": 0.9312,
+      "step": 6315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2417602043436644,
+      "learning_rate": 6.755418442329846e-06,
+      "loss": 0.9475,
+      "step": 6316
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8273505990827386,
+      "learning_rate": 6.754447895977113e-06,
+      "loss": 0.8629,
+      "step": 6317
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9340322183099574,
+      "learning_rate": 6.753477274231846e-06,
+      "loss": 1.031,
+      "step": 6318
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.46693431300505,
+      "learning_rate": 6.7525065771357546e-06,
+      "loss": 0.8658,
+      "step": 6319
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.936534082283718,
+      "learning_rate": 6.751535804730553e-06,
+      "loss": 0.7518,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1677089403222616,
+      "learning_rate": 6.750564957057958e-06,
+      "loss": 0.9276,
+      "step": 6321
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8616018849582767,
+      "learning_rate": 6.749594034159687e-06,
+      "loss": 0.7626,
+      "step": 6322
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1154195715406403,
+      "learning_rate": 6.748623036077464e-06,
+      "loss": 0.8697,
+      "step": 6323
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.918255596251332,
+      "learning_rate": 6.7476519628530145e-06,
+      "loss": 0.9805,
+      "step": 6324
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.032300891168417,
+      "learning_rate": 6.7466808145280685e-06,
+      "loss": 0.8022,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1305081016322476,
+      "learning_rate": 6.745709591144358e-06,
+      "loss": 0.5995,
+      "step": 6326
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.468235257821152,
+      "learning_rate": 6.744738292743617e-06,
+      "loss": 0.9153,
+      "step": 6327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2080884967794865,
+      "learning_rate": 6.743766919367588e-06,
+      "loss": 0.6252,
+      "step": 6328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.110394405244301,
+      "learning_rate": 6.742795471058009e-06,
+      "loss": 0.6727,
+      "step": 6329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9917595464278055,
+      "learning_rate": 6.741823947856627e-06,
+      "loss": 0.6225,
+      "step": 6330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.377674913926684,
+      "learning_rate": 6.740852349805193e-06,
+      "loss": 0.7697,
+      "step": 6331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.210015467492271,
+      "learning_rate": 6.739880676945455e-06,
+      "loss": 0.8243,
+      "step": 6332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9432029596981523,
+      "learning_rate": 6.73890892931917e-06,
+      "loss": 0.7061,
+      "step": 6333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4053494164842597,
+      "learning_rate": 6.737937106968094e-06,
+      "loss": 0.8498,
+      "step": 6334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8190684242209536,
+      "learning_rate": 6.736965209933992e-06,
+      "loss": 0.8374,
+      "step": 6335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9736089195671056,
+      "learning_rate": 6.735993238258624e-06,
+      "loss": 0.9549,
+      "step": 6336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9160528968841772,
+      "learning_rate": 6.735021191983763e-06,
+      "loss": 0.8177,
+      "step": 6337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.376950070606331,
+      "learning_rate": 6.734049071151175e-06,
+      "loss": 0.6457,
+      "step": 6338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.587455250911704,
+      "learning_rate": 6.7330768758026374e-06,
+      "loss": 0.8097,
+      "step": 6339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6656608122119327,
+      "learning_rate": 6.732104605979928e-06,
+      "loss": 0.8621,
+      "step": 6340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.885937392729318,
+      "learning_rate": 6.7311322617248246e-06,
+      "loss": 0.86,
+      "step": 6341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1131358766477146,
+      "learning_rate": 6.730159843079113e-06,
+      "loss": 0.7579,
+      "step": 6342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.712032293420833,
+      "learning_rate": 6.729187350084579e-06,
+      "loss": 0.9955,
+      "step": 6343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7868227170888926,
+      "learning_rate": 6.728214782783013e-06,
+      "loss": 0.8103,
+      "step": 6344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2068056323462164,
+      "learning_rate": 6.72724214121621e-06,
+      "loss": 0.8882,
+      "step": 6345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.247685375686582,
+      "learning_rate": 6.726269425425967e-06,
+      "loss": 0.6841,
+      "step": 6346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.813977834114683,
+      "learning_rate": 6.725296635454079e-06,
+      "loss": 0.8515,
+      "step": 6347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0058381148999413,
+      "learning_rate": 6.724323771342354e-06,
+      "loss": 0.8638,
+      "step": 6348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.020117892766964,
+      "learning_rate": 6.723350833132596e-06,
+      "loss": 0.8245,
+      "step": 6349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.975667390800692,
+      "learning_rate": 6.722377820866614e-06,
+      "loss": 0.759,
+      "step": 6350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.901294562467036,
+      "learning_rate": 6.721404734586221e-06,
+      "loss": 0.7322,
+      "step": 6351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.236438151330182,
+      "learning_rate": 6.7204315743332325e-06,
+      "loss": 0.6078,
+      "step": 6352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9079707109876103,
+      "learning_rate": 6.71945834014947e-06,
+      "loss": 0.7609,
+      "step": 6353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3277813359445663,
+      "learning_rate": 6.7184850320767505e-06,
+      "loss": 0.9808,
+      "step": 6354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3689879688304507,
+      "learning_rate": 6.717511650156903e-06,
+      "loss": 0.8712,
+      "step": 6355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.250559183371506,
+      "learning_rate": 6.716538194431754e-06,
+      "loss": 0.7825,
+      "step": 6356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.764323882059061,
+      "learning_rate": 6.715564664943138e-06,
+      "loss": 0.6697,
+      "step": 6357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.586766977581921,
+      "learning_rate": 6.714591061732886e-06,
+      "loss": 0.9545,
+      "step": 6358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.512080664829798,
+      "learning_rate": 6.7136173848428375e-06,
+      "loss": 0.9796,
+      "step": 6359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8099662977072315,
+      "learning_rate": 6.712643634314834e-06,
+      "loss": 0.9435,
+      "step": 6360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.736875320786672,
+      "learning_rate": 6.711669810190717e-06,
+      "loss": 0.6379,
+      "step": 6361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.264834396013666,
+      "learning_rate": 6.710695912512339e-06,
+      "loss": 0.8918,
+      "step": 6362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1614489917093236,
+      "learning_rate": 6.7097219413215474e-06,
+      "loss": 0.9291,
+      "step": 6363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0312265988356555,
+      "learning_rate": 6.708747896660196e-06,
+      "loss": 0.8611,
+      "step": 6364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.365455343077016,
+      "learning_rate": 6.70777377857014e-06,
+      "loss": 0.8898,
+      "step": 6365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4385263575467033,
+      "learning_rate": 6.706799587093243e-06,
+      "loss": 0.9622,
+      "step": 6366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5221004371391746,
+      "learning_rate": 6.705825322271367e-06,
+      "loss": 0.8613,
+      "step": 6367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0125694763037822,
+      "learning_rate": 6.7048509841463775e-06,
+      "loss": 0.8125,
+      "step": 6368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5500554384130694,
+      "learning_rate": 6.703876572760144e-06,
+      "loss": 0.801,
+      "step": 6369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2273569571907423,
+      "learning_rate": 6.702902088154539e-06,
+      "loss": 0.5848,
+      "step": 6370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9412536768610507,
+      "learning_rate": 6.701927530371439e-06,
+      "loss": 0.7994,
+      "step": 6371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.454350617278811,
+      "learning_rate": 6.700952899452723e-06,
+      "loss": 0.7869,
+      "step": 6372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9293185084207305,
+      "learning_rate": 6.699978195440273e-06,
+      "loss": 0.9343,
+      "step": 6373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.195054958006777,
+      "learning_rate": 6.6990034183759726e-06,
+      "loss": 0.8999,
+      "step": 6374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1568629169493407,
+      "learning_rate": 6.698028568301712e-06,
+      "loss": 0.8348,
+      "step": 6375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5765670062493906,
+      "learning_rate": 6.697053645259382e-06,
+      "loss": 0.9865,
+      "step": 6376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3765160460651784,
+      "learning_rate": 6.696078649290878e-06,
+      "loss": 0.7154,
+      "step": 6377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7628782611516287,
+      "learning_rate": 6.695103580438097e-06,
+      "loss": 0.8574,
+      "step": 6378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3181443781434687,
+      "learning_rate": 6.694128438742939e-06,
+      "loss": 0.8219,
+      "step": 6379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0710618253408013,
+      "learning_rate": 6.693153224247309e-06,
+      "loss": 0.8573,
+      "step": 6380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.275105016543878,
+      "learning_rate": 6.692177936993114e-06,
+      "loss": 0.7779,
+      "step": 6381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.582240396126286,
+      "learning_rate": 6.691202577022267e-06,
+      "loss": 0.8731,
+      "step": 6382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.777914000353625,
+      "learning_rate": 6.690227144376676e-06,
+      "loss": 0.8727,
+      "step": 6383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8031255428692705,
+      "learning_rate": 6.689251639098261e-06,
+      "loss": 0.8039,
+      "step": 6384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2679675953528387,
+      "learning_rate": 6.68827606122894e-06,
+      "loss": 0.9591,
+      "step": 6385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.253132953385873,
+      "learning_rate": 6.687300410810638e-06,
+      "loss": 0.9811,
+      "step": 6386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3643583681011537,
+      "learning_rate": 6.686324687885281e-06,
+      "loss": 0.7925,
+      "step": 6387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.100889217209221,
+      "learning_rate": 6.685348892494793e-06,
+      "loss": 0.9731,
+      "step": 6388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3640067845561807,
+      "learning_rate": 6.684373024681112e-06,
+      "loss": 1.0102,
+      "step": 6389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0319899643550172,
+      "learning_rate": 6.683397084486169e-06,
+      "loss": 0.5669,
+      "step": 6390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.155296395800635,
+      "learning_rate": 6.682421071951907e-06,
+      "loss": 0.6993,
+      "step": 6391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0285696017149846,
+      "learning_rate": 6.681444987120262e-06,
+      "loss": 0.8603,
+      "step": 6392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.299441977148562,
+      "learning_rate": 6.680468830033182e-06,
+      "loss": 1.0298,
+      "step": 6393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8300420175560426,
+      "learning_rate": 6.679492600732614e-06,
+      "loss": 0.8019,
+      "step": 6394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0773127562478164,
+      "learning_rate": 6.678516299260507e-06,
+      "loss": 0.8086,
+      "step": 6395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.123822151096538,
+      "learning_rate": 6.677539925658819e-06,
+      "loss": 0.7889,
+      "step": 6396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.227243652566738,
+      "learning_rate": 6.676563479969502e-06,
+      "loss": 1.0667,
+      "step": 6397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.086262758306438,
+      "learning_rate": 6.67558696223452e-06,
+      "loss": 0.7605,
+      "step": 6398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.269857211870047,
+      "learning_rate": 6.674610372495832e-06,
+      "loss": 0.8777,
+      "step": 6399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.726109459498516,
+      "learning_rate": 6.673633710795409e-06,
+      "loss": 0.8439,
+      "step": 6400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5821885771348376,
+      "learning_rate": 6.672656977175218e-06,
+      "loss": 0.8698,
+      "step": 6401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0054126897702467,
+      "learning_rate": 6.671680171677232e-06,
+      "loss": 0.6199,
+      "step": 6402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5882539284407837,
+      "learning_rate": 6.670703294343423e-06,
+      "loss": 0.8793,
+      "step": 6403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1322333554184487,
+      "learning_rate": 6.669726345215776e-06,
+      "loss": 0.8536,
+      "step": 6404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.347432972860685,
+      "learning_rate": 6.668749324336268e-06,
+      "loss": 0.7484,
+      "step": 6405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.7760616337646775,
+      "learning_rate": 6.667772231746885e-06,
+      "loss": 0.9225,
+      "step": 6406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.263163962457332,
+      "learning_rate": 6.666795067489615e-06,
+      "loss": 0.8338,
+      "step": 6407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3117953502428747,
+      "learning_rate": 6.665817831606447e-06,
+      "loss": 0.7212,
+      "step": 6408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.014873812914088,
+      "learning_rate": 6.66484052413938e-06,
+      "loss": 0.982,
+      "step": 6409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1224909725249455,
+      "learning_rate": 6.663863145130406e-06,
+      "loss": 0.6025,
+      "step": 6410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4532968638702313,
+      "learning_rate": 6.662885694621527e-06,
+      "loss": 0.9217,
+      "step": 6411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.353894892928616,
+      "learning_rate": 6.661908172654746e-06,
+      "loss": 0.9591,
+      "step": 6412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.111305447475508,
+      "learning_rate": 6.6609305792720705e-06,
+      "loss": 0.8162,
+      "step": 6413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.526940720647731,
+      "learning_rate": 6.659952914515508e-06,
+      "loss": 0.8889,
+      "step": 6414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8910796057027155,
+      "learning_rate": 6.658975178427073e-06,
+      "loss": 0.8949,
+      "step": 6415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7270968110138547,
+      "learning_rate": 6.65799737104878e-06,
+      "loss": 0.5331,
+      "step": 6416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.726488638568398,
+      "learning_rate": 6.657019492422645e-06,
+      "loss": 0.8234,
+      "step": 6417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.172826761263437,
+      "learning_rate": 6.656041542590694e-06,
+      "loss": 0.7158,
+      "step": 6418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.255601095117392,
+      "learning_rate": 6.65506352159495e-06,
+      "loss": 0.8668,
+      "step": 6419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.036909281246078,
+      "learning_rate": 6.65408542947744e-06,
+      "loss": 1.0222,
+      "step": 6420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8473752137087704,
+      "learning_rate": 6.653107266280194e-06,
+      "loss": 0.8013,
+      "step": 6421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7962310134981008,
+      "learning_rate": 6.652129032045249e-06,
+      "loss": 0.8909,
+      "step": 6422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4433954784211385,
+      "learning_rate": 6.651150726814639e-06,
+      "loss": 0.8588,
+      "step": 6423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.338629865662459,
+      "learning_rate": 6.650172350630406e-06,
+      "loss": 0.8532,
+      "step": 6424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.336222600696049,
+      "learning_rate": 6.649193903534591e-06,
+      "loss": 0.7957,
+      "step": 6425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1725521548052047,
+      "learning_rate": 6.6482153855692395e-06,
+      "loss": 0.8388,
+      "step": 6426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.354620620678434,
+      "learning_rate": 6.647236796776405e-06,
+      "loss": 0.7587,
+      "step": 6427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.784604239028471,
+      "learning_rate": 6.646258137198136e-06,
+      "loss": 0.9136,
+      "step": 6428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0357725194859309,
+      "learning_rate": 6.645279406876488e-06,
+      "loss": 0.7677,
+      "step": 6429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8075634921160697,
+      "learning_rate": 6.644300605853518e-06,
+      "loss": 0.875,
+      "step": 6430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.135725455128749,
+      "learning_rate": 6.643321734171291e-06,
+      "loss": 0.9077,
+      "step": 6431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1088410090194922,
+      "learning_rate": 6.642342791871868e-06,
+      "loss": 0.8813,
+      "step": 6432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9428415822810523,
+      "learning_rate": 6.64136377899732e-06,
+      "loss": 0.6549,
+      "step": 6433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7620178783222131,
+      "learning_rate": 6.640384695589714e-06,
+      "loss": 0.7327,
+      "step": 6434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8950898769747178,
+      "learning_rate": 6.639405541691122e-06,
+      "loss": 0.8887,
+      "step": 6435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8043790166384408,
+      "learning_rate": 6.638426317343624e-06,
+      "loss": 0.8077,
+      "step": 6436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8348191791483397,
+      "learning_rate": 6.6374470225892985e-06,
+      "loss": 0.8692,
+      "step": 6437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1832703219173375,
+      "learning_rate": 6.6364676574702284e-06,
+      "loss": 0.6218,
+      "step": 6438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.043542201929468,
+      "learning_rate": 6.635488222028497e-06,
+      "loss": 0.6058,
+      "step": 6439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1458917318052717,
+      "learning_rate": 6.6345087163061935e-06,
+      "loss": 0.6897,
+      "step": 6440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0178328344942944,
+      "learning_rate": 6.633529140345412e-06,
+      "loss": 0.7486,
+      "step": 6441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1315033679868893,
+      "learning_rate": 6.632549494188245e-06,
+      "loss": 0.6408,
+      "step": 6442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3678240900405196,
+      "learning_rate": 6.631569777876791e-06,
+      "loss": 1.0124,
+      "step": 6443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.365505877552722,
+      "learning_rate": 6.630589991453148e-06,
+      "loss": 0.8578,
+      "step": 6444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4937179358557837,
+      "learning_rate": 6.629610134959423e-06,
+      "loss": 0.7728,
+      "step": 6445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.574237939800109,
+      "learning_rate": 6.628630208437719e-06,
+      "loss": 0.8022,
+      "step": 6446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.065572728383007,
+      "learning_rate": 6.627650211930152e-06,
+      "loss": 0.9644,
+      "step": 6447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3102758474450904,
+      "learning_rate": 6.626670145478826e-06,
+      "loss": 0.9549,
+      "step": 6448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0826295480516812,
+      "learning_rate": 6.6256900091258644e-06,
+      "loss": 0.5689,
+      "step": 6449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0600877037763068,
+      "learning_rate": 6.62470980291338e-06,
+      "loss": 0.708,
+      "step": 6450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.255476278591482,
+      "learning_rate": 6.623729526883498e-06,
+      "loss": 0.9041,
+      "step": 6451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.296572790546804,
+      "learning_rate": 6.622749181078344e-06,
+      "loss": 0.8376,
+      "step": 6452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9638809719267079,
+      "learning_rate": 6.621768765540041e-06,
+      "loss": 0.9906,
+      "step": 6453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.315628578104158,
+      "learning_rate": 6.620788280310722e-06,
+      "loss": 0.7507,
+      "step": 6454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3710232175183164,
+      "learning_rate": 6.619807725432522e-06,
+      "loss": 0.7425,
+      "step": 6455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6376671819199613,
+      "learning_rate": 6.618827100947577e-06,
+      "loss": 0.835,
+      "step": 6456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7021268684455664,
+      "learning_rate": 6.6178464068980256e-06,
+      "loss": 0.8422,
+      "step": 6457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5066039214441775,
+      "learning_rate": 6.616865643326011e-06,
+      "loss": 0.8769,
+      "step": 6458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.193143283664724,
+      "learning_rate": 6.615884810273678e-06,
+      "loss": 0.6939,
+      "step": 6459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4072690031482686,
+      "learning_rate": 6.614903907783177e-06,
+      "loss": 0.8587,
+      "step": 6460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.223326552822601,
+      "learning_rate": 6.613922935896659e-06,
+      "loss": 0.8721,
+      "step": 6461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.743567009847034,
+      "learning_rate": 6.612941894656277e-06,
+      "loss": 0.8499,
+      "step": 6462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4608380384695536,
+      "learning_rate": 6.61196078410419e-06,
+      "loss": 0.8657,
+      "step": 6463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3476009275801726,
+      "learning_rate": 6.610979604282557e-06,
+      "loss": 0.7919,
+      "step": 6464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3920693189788382,
+      "learning_rate": 6.609998355233543e-06,
+      "loss": 0.7407,
+      "step": 6465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.128477901008596,
+      "learning_rate": 6.609017036999314e-06,
+      "loss": 0.8982,
+      "step": 6466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2658693788972686,
+      "learning_rate": 6.608035649622041e-06,
+      "loss": 0.9078,
+      "step": 6467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4746215079810456,
+      "learning_rate": 6.607054193143894e-06,
+      "loss": 0.7871,
+      "step": 6468
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.044322133273274,
+      "learning_rate": 6.606072667607048e-06,
+      "loss": 0.7936,
+      "step": 6469
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2591463747060865,
+      "learning_rate": 6.605091073053685e-06,
+      "loss": 0.785,
+      "step": 6470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1167965563332167,
+      "learning_rate": 6.604109409525982e-06,
+      "loss": 0.7705,
+      "step": 6471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.021848056015594,
+      "learning_rate": 6.603127677066126e-06,
+      "loss": 0.8994,
+      "step": 6472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2036718313156207,
+      "learning_rate": 6.602145875716301e-06,
+      "loss": 0.7264,
+      "step": 6473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5900902846113656,
+      "learning_rate": 6.601164005518702e-06,
+      "loss": 0.616,
+      "step": 6474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.043362275818194,
+      "learning_rate": 6.600182066515519e-06,
+      "loss": 0.9445,
+      "step": 6475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.90394957423996,
+      "learning_rate": 6.599200058748949e-06,
+      "loss": 0.857,
+      "step": 6476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.967072171661711,
+      "learning_rate": 6.59821798226119e-06,
+      "loss": 0.8835,
+      "step": 6477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0937033308776616,
+      "learning_rate": 6.597235837094446e-06,
+      "loss": 0.7963,
+      "step": 6478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.204170302065137,
+      "learning_rate": 6.59625362329092e-06,
+      "loss": 0.749,
+      "step": 6479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0834041425778875,
+      "learning_rate": 6.595271340892823e-06,
+      "loss": 0.8089,
+      "step": 6480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.12527046056751,
+      "learning_rate": 6.594288989942361e-06,
+      "loss": 0.8025,
+      "step": 6481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.07041198364226,
+      "learning_rate": 6.593306570481751e-06,
+      "loss": 0.9706,
+      "step": 6482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9599541110299739,
+      "learning_rate": 6.592324082553209e-06,
+      "loss": 0.7083,
+      "step": 6483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9586779617021919,
+      "learning_rate": 6.591341526198955e-06,
+      "loss": 0.717,
+      "step": 6484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.335177303818045,
+      "learning_rate": 6.590358901461213e-06,
+      "loss": 0.6533,
+      "step": 6485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3658076902391745,
+      "learning_rate": 6.589376208382206e-06,
+      "loss": 0.8792,
+      "step": 6486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.894700527826415,
+      "learning_rate": 6.588393447004163e-06,
+      "loss": 0.932,
+      "step": 6487
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.099350282885331,
+      "learning_rate": 6.587410617369316e-06,
+      "loss": 0.7909,
+      "step": 6488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.52935518052607,
+      "learning_rate": 6.586427719519901e-06,
+      "loss": 0.7055,
+      "step": 6489
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2900140723363895,
+      "learning_rate": 6.585444753498153e-06,
+      "loss": 0.7702,
+      "step": 6490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.456883191877623,
+      "learning_rate": 6.584461719346312e-06,
+      "loss": 0.8418,
+      "step": 6491
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.606043746603003,
+      "learning_rate": 6.583478617106623e-06,
+      "loss": 0.7856,
+      "step": 6492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2424897282763228,
+      "learning_rate": 6.582495446821331e-06,
+      "loss": 0.8124,
+      "step": 6493
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8328304950275114,
+      "learning_rate": 6.581512208532685e-06,
+      "loss": 0.9275,
+      "step": 6494
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.551768405938063,
+      "learning_rate": 6.580528902282938e-06,
+      "loss": 0.9051,
+      "step": 6495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1571354348682408,
+      "learning_rate": 6.579545528114344e-06,
+      "loss": 0.5617,
+      "step": 6496
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.141726065380536,
+      "learning_rate": 6.57856208606916e-06,
+      "loss": 0.7688,
+      "step": 6497
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.128867225473676,
+      "learning_rate": 6.577578576189648e-06,
+      "loss": 0.6574,
+      "step": 6498
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8840390570843528,
+      "learning_rate": 6.576594998518071e-06,
+      "loss": 0.912,
+      "step": 6499
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.284422716061072,
+      "learning_rate": 6.575611353096695e-06,
+      "loss": 0.8768,
+      "step": 6500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9669783090806818,
+      "learning_rate": 6.574627639967791e-06,
+      "loss": 0.9193,
+      "step": 6501
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7336785058460242,
+      "learning_rate": 6.57364385917363e-06,
+      "loss": 0.7944,
+      "step": 6502
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6669727131760967,
+      "learning_rate": 6.572660010756489e-06,
+      "loss": 0.8092,
+      "step": 6503
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.411825235501468,
+      "learning_rate": 6.5716760947586425e-06,
+      "loss": 0.6114,
+      "step": 6504
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1091294661446442,
+      "learning_rate": 6.570692111222375e-06,
+      "loss": 0.6371,
+      "step": 6505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.480543253477148,
+      "learning_rate": 6.56970806018997e-06,
+      "loss": 0.7717,
+      "step": 6506
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.762283527266388,
+      "learning_rate": 6.568723941703712e-06,
+      "loss": 0.8834,
+      "step": 6507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2406595736875778,
+      "learning_rate": 6.567739755805895e-06,
+      "loss": 0.6933,
+      "step": 6508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.20913226747092,
+      "learning_rate": 6.566755502538806e-06,
+      "loss": 0.6124,
+      "step": 6509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.117234844401068,
+      "learning_rate": 6.565771181944747e-06,
+      "loss": 0.7169,
+      "step": 6510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2920325851531205,
+      "learning_rate": 6.5647867940660105e-06,
+      "loss": 0.8757,
+      "step": 6511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7592260613411943,
+      "learning_rate": 6.563802338944902e-06,
+      "loss": 0.7135,
+      "step": 6512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0943011975986683,
+      "learning_rate": 6.562817816623724e-06,
+      "loss": 0.666,
+      "step": 6513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6223154211006796,
+      "learning_rate": 6.561833227144784e-06,
+      "loss": 0.8454,
+      "step": 6514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.078359042232934,
+      "learning_rate": 6.560848570550391e-06,
+      "loss": 0.8574,
+      "step": 6515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3728113218535096,
+      "learning_rate": 6.559863846882859e-06,
+      "loss": 0.7902,
+      "step": 6516
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1247473718085192,
+      "learning_rate": 6.558879056184505e-06,
+      "loss": 0.9203,
+      "step": 6517
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8050620603037357,
+      "learning_rate": 6.557894198497644e-06,
+      "loss": 0.9767,
+      "step": 6518
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.253257605620509,
+      "learning_rate": 6.556909273864601e-06,
+      "loss": 0.7808,
+      "step": 6519
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2080880286725915,
+      "learning_rate": 6.555924282327698e-06,
+      "loss": 0.8017,
+      "step": 6520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3855154053469017,
+      "learning_rate": 6.554939223929264e-06,
+      "loss": 0.7794,
+      "step": 6521
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.271595083733517,
+      "learning_rate": 6.553954098711629e-06,
+      "loss": 0.669,
+      "step": 6522
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.004525268913268,
+      "learning_rate": 6.5529689067171255e-06,
+      "loss": 0.6488,
+      "step": 6523
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8887994669547026,
+      "learning_rate": 6.551983647988089e-06,
+      "loss": 0.7131,
+      "step": 6524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4220963765759276,
+      "learning_rate": 6.550998322566857e-06,
+      "loss": 0.9423,
+      "step": 6525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.478732675100328,
+      "learning_rate": 6.550012930495775e-06,
+      "loss": 1.0548,
+      "step": 6526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.339786581275418,
+      "learning_rate": 6.549027471817186e-06,
+      "loss": 0.7724,
+      "step": 6527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2330218398441073,
+      "learning_rate": 6.5480419465734355e-06,
+      "loss": 0.8096,
+      "step": 6528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2362451412202202,
+      "learning_rate": 6.547056354806874e-06,
+      "loss": 0.8164,
+      "step": 6529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.919903946476421,
+      "learning_rate": 6.546070696559856e-06,
+      "loss": 0.7903,
+      "step": 6530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.592749550989067,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 1.0564,
+      "step": 6531
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.26779003621699,
+      "learning_rate": 6.544099180793877e-06,
+      "loss": 0.8756,
+      "step": 6532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1089503921834765,
+      "learning_rate": 6.543113323359635e-06,
+      "loss": 0.6465,
+      "step": 6533
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6999535979123954,
+      "learning_rate": 6.542127399614376e-06,
+      "loss": 0.8315,
+      "step": 6534
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9263605095671372,
+      "learning_rate": 6.54114140960047e-06,
+      "loss": 1.0886,
+      "step": 6535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0442113272891975,
+      "learning_rate": 6.540155353360284e-06,
+      "loss": 0.7423,
+      "step": 6536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9812434012961143,
+      "learning_rate": 6.539169230936193e-06,
+      "loss": 0.5593,
+      "step": 6537
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9443281401454868,
+      "learning_rate": 6.5381830423705714e-06,
+      "loss": 0.6139,
+      "step": 6538
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9744182572429255,
+      "learning_rate": 6.5371967877058e-06,
+      "loss": 0.7517,
+      "step": 6539
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.089215390716909,
+      "learning_rate": 6.536210466984259e-06,
+      "loss": 0.7374,
+      "step": 6540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9254832760773646,
+      "learning_rate": 6.535224080248332e-06,
+      "loss": 0.729,
+      "step": 6541
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9188792198948679,
+      "learning_rate": 6.534237627540408e-06,
+      "loss": 0.8226,
+      "step": 6542
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.940935099436746,
+      "learning_rate": 6.533251108902875e-06,
+      "loss": 0.7472,
+      "step": 6543
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0350112211388613,
+      "learning_rate": 6.532264524378128e-06,
+      "loss": 0.8714,
+      "step": 6544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.320916790995536,
+      "learning_rate": 6.531277874008562e-06,
+      "loss": 1.0365,
+      "step": 6545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2130206886520694,
+      "learning_rate": 6.5302911578365755e-06,
+      "loss": 0.8928,
+      "step": 6546
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.115722617102325,
+      "learning_rate": 6.529304375904568e-06,
+      "loss": 0.8472,
+      "step": 6547
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2330464044353877,
+      "learning_rate": 6.528317528254948e-06,
+      "loss": 0.7095,
+      "step": 6548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4440447748558802,
+      "learning_rate": 6.52733061493012e-06,
+      "loss": 0.6614,
+      "step": 6549
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1383139782630494,
+      "learning_rate": 6.526343635972492e-06,
+      "loss": 0.8713,
+      "step": 6550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8660370727023692,
+      "learning_rate": 6.525356591424479e-06,
+      "loss": 0.9213,
+      "step": 6551
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0887758132391383,
+      "learning_rate": 6.5243694813284975e-06,
+      "loss": 0.901,
+      "step": 6552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.347724031008665,
+      "learning_rate": 6.523382305726964e-06,
+      "loss": 0.9127,
+      "step": 6553
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8134741050567538,
+      "learning_rate": 6.522395064662299e-06,
+      "loss": 0.8432,
+      "step": 6554
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1763590394376857,
+      "learning_rate": 6.521407758176929e-06,
+      "loss": 0.9119,
+      "step": 6555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1471608596544334,
+      "learning_rate": 6.520420386313277e-06,
+      "loss": 1.049,
+      "step": 6556
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.35247937187677,
+      "learning_rate": 6.5194329491137775e-06,
+      "loss": 0.9358,
+      "step": 6557
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4698497833414765,
+      "learning_rate": 6.518445446620858e-06,
+      "loss": 0.8264,
+      "step": 6558
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8153891358355358,
+      "learning_rate": 6.517457878876958e-06,
+      "loss": 0.7241,
+      "step": 6559
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9144189096773485,
+      "learning_rate": 6.516470245924511e-06,
+      "loss": 0.6364,
+      "step": 6560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4970478456746696,
+      "learning_rate": 6.515482547805962e-06,
+      "loss": 0.7731,
+      "step": 6561
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0631831448117188,
+      "learning_rate": 6.514494784563753e-06,
+      "loss": 0.7024,
+      "step": 6562
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.067078225954953,
+      "learning_rate": 6.5135069562403285e-06,
+      "loss": 0.8702,
+      "step": 6563
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2082577604587827,
+      "learning_rate": 6.512519062878142e-06,
+      "loss": 0.9644,
+      "step": 6564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.290085550227101,
+      "learning_rate": 6.511531104519639e-06,
+      "loss": 0.6999,
+      "step": 6565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8245802670574045,
+      "learning_rate": 6.510543081207281e-06,
+      "loss": 0.7431,
+      "step": 6566
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2141065702012424,
+      "learning_rate": 6.509554992983521e-06,
+      "loss": 0.8383,
+      "step": 6567
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6862299216840313,
+      "learning_rate": 6.508566839890823e-06,
+      "loss": 0.8455,
+      "step": 6568
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.828367318515681,
+      "learning_rate": 6.507578621971646e-06,
+      "loss": 0.6581,
+      "step": 6569
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.187679234451297,
+      "learning_rate": 6.506590339268459e-06,
+      "loss": 0.747,
+      "step": 6570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.864048193424272,
+      "learning_rate": 6.50560199182373e-06,
+      "loss": 0.6252,
+      "step": 6571
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.667295578943196,
+      "learning_rate": 6.504613579679929e-06,
+      "loss": 0.9229,
+      "step": 6572
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.001401135154571,
+      "learning_rate": 6.503625102879534e-06,
+      "loss": 0.9626,
+      "step": 6573
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8517818865574742,
+      "learning_rate": 6.502636561465018e-06,
+      "loss": 0.8355,
+      "step": 6574
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.004373964955205,
+      "learning_rate": 6.501647955478863e-06,
+      "loss": 0.7597,
+      "step": 6575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.164240558245079,
+      "learning_rate": 6.500659284963549e-06,
+      "loss": 0.8707,
+      "step": 6576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.241456147894553,
+      "learning_rate": 6.499670549961565e-06,
+      "loss": 0.7086,
+      "step": 6577
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.322846898925281,
+      "learning_rate": 6.498681750515398e-06,
+      "loss": 0.9264,
+      "step": 6578
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2358963935640173,
+      "learning_rate": 6.497692886667537e-06,
+      "loss": 0.7492,
+      "step": 6579
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6869481704052607,
+      "learning_rate": 6.496703958460479e-06,
+      "loss": 0.7713,
+      "step": 6580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.045761191757835,
+      "learning_rate": 6.495714965936717e-06,
+      "loss": 0.785,
+      "step": 6581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0298870747958513,
+      "learning_rate": 6.494725909138751e-06,
+      "loss": 0.6396,
+      "step": 6582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.006991409492692,
+      "learning_rate": 6.493736788109086e-06,
+      "loss": 0.8226,
+      "step": 6583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0058157955967673,
+      "learning_rate": 6.492747602890223e-06,
+      "loss": 0.7704,
+      "step": 6584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9881147287131977,
+      "learning_rate": 6.491758353524672e-06,
+      "loss": 0.717,
+      "step": 6585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.443933459462508,
+      "learning_rate": 6.490769040054942e-06,
+      "loss": 0.7584,
+      "step": 6586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.13500207259539,
+      "learning_rate": 6.489779662523545e-06,
+      "loss": 0.8997,
+      "step": 6587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.13642460882119,
+      "learning_rate": 6.4887902209730004e-06,
+      "loss": 0.9612,
+      "step": 6588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1024491221050146,
+      "learning_rate": 6.487800715445822e-06,
+      "loss": 0.5867,
+      "step": 6589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9741993663115447,
+      "learning_rate": 6.486811145984534e-06,
+      "loss": 0.7828,
+      "step": 6590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.409995723463331,
+      "learning_rate": 6.48582151263166e-06,
+      "loss": 0.7868,
+      "step": 6591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.128026899504153,
+      "learning_rate": 6.484831815429726e-06,
+      "loss": 0.9012,
+      "step": 6592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.705601706582096,
+      "learning_rate": 6.483842054421263e-06,
+      "loss": 0.6549,
+      "step": 6593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3649209214428573,
+      "learning_rate": 6.4828522296488014e-06,
+      "loss": 0.6876,
+      "step": 6594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.436620563036828,
+      "learning_rate": 6.481862341154878e-06,
+      "loss": 0.8954,
+      "step": 6595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1255402832371852,
+      "learning_rate": 6.480872388982028e-06,
+      "loss": 0.8116,
+      "step": 6596
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9343486468765676,
+      "learning_rate": 6.479882373172795e-06,
+      "loss": 0.7424,
+      "step": 6597
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9315217322385034,
+      "learning_rate": 6.47889229376972e-06,
+      "loss": 0.9299,
+      "step": 6598
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2463781779407506,
+      "learning_rate": 6.477902150815347e-06,
+      "loss": 0.8761,
+      "step": 6599
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.116839405751012,
+      "learning_rate": 6.476911944352231e-06,
+      "loss": 0.8735,
+      "step": 6600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.932699255309146,
+      "learning_rate": 6.475921674422917e-06,
+      "loss": 0.7746,
+      "step": 6601
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3199656375930269,
+      "learning_rate": 6.474931341069963e-06,
+      "loss": 0.7766,
+      "step": 6602
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0989298835934795,
+      "learning_rate": 6.473940944335923e-06,
+      "loss": 0.6131,
+      "step": 6603
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.906302857972085,
+      "learning_rate": 6.472950484263359e-06,
+      "loss": 0.6227,
+      "step": 6604
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6882629434521452,
+      "learning_rate": 6.471959960894832e-06,
+      "loss": 0.9044,
+      "step": 6605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9302584286743631,
+      "learning_rate": 6.470969374272908e-06,
+      "loss": 0.935,
+      "step": 6606
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7310039109953503,
+      "learning_rate": 6.469978724440153e-06,
+      "loss": 0.767,
+      "step": 6607
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.333580229316664,
+      "learning_rate": 6.4689880114391375e-06,
+      "loss": 0.813,
+      "step": 6608
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1666509813554233,
+      "learning_rate": 6.467997235312437e-06,
+      "loss": 0.8574,
+      "step": 6609
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0376625936681994,
+      "learning_rate": 6.4670063961026265e-06,
+      "loss": 0.5312,
+      "step": 6610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8879202128132,
+      "learning_rate": 6.466015493852283e-06,
+      "loss": 0.9088,
+      "step": 6611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.265159430495573,
+      "learning_rate": 6.465024528603987e-06,
+      "loss": 0.9618,
+      "step": 6612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0223833755691643,
+      "learning_rate": 6.464033500400327e-06,
+      "loss": 0.7509,
+      "step": 6613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.329669951681842,
+      "learning_rate": 6.463042409283885e-06,
+      "loss": 0.8919,
+      "step": 6614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2630411368955454,
+      "learning_rate": 6.462051255297255e-06,
+      "loss": 0.8924,
+      "step": 6615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0919531508182123,
+      "learning_rate": 6.461060038483025e-06,
+      "loss": 0.7324,
+      "step": 6616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.832788619466443,
+      "learning_rate": 6.460068758883789e-06,
+      "loss": 0.8998,
+      "step": 6617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.501401992120107,
+      "learning_rate": 6.4590774165421485e-06,
+      "loss": 0.5828,
+      "step": 6618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9120445987085003,
+      "learning_rate": 6.458086011500703e-06,
+      "loss": 0.5963,
+      "step": 6619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7049255668564633,
+      "learning_rate": 6.457094543802053e-06,
+      "loss": 0.9562,
+      "step": 6620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.72935843784799,
+      "learning_rate": 6.456103013488804e-06,
+      "loss": 0.7397,
+      "step": 6621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.246338163988665,
+      "learning_rate": 6.455111420603568e-06,
+      "loss": 0.9297,
+      "step": 6622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5688900234376972,
+      "learning_rate": 6.454119765188951e-06,
+      "loss": 0.8784,
+      "step": 6623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.077309858403266,
+      "learning_rate": 6.453128047287573e-06,
+      "loss": 0.793,
+      "step": 6624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.434272226559143,
+      "learning_rate": 6.452136266942044e-06,
+      "loss": 0.8663,
+      "step": 6625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.090986727095526,
+      "learning_rate": 6.451144424194985e-06,
+      "loss": 0.6955,
+      "step": 6626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2335658514056083,
+      "learning_rate": 6.450152519089019e-06,
+      "loss": 0.7369,
+      "step": 6627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8761576556232573,
+      "learning_rate": 6.449160551666769e-06,
+      "loss": 0.8958,
+      "step": 6628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.883680080746067,
+      "learning_rate": 6.448168521970865e-06,
+      "loss": 0.8481,
+      "step": 6629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2289430264227286,
+      "learning_rate": 6.447176430043931e-06,
+      "loss": 0.6945,
+      "step": 6630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9846338121589393,
+      "learning_rate": 6.446184275928604e-06,
+      "loss": 0.8666,
+      "step": 6631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3780403550745683,
+      "learning_rate": 6.445192059667518e-06,
+      "loss": 0.8821,
+      "step": 6632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.044769893807534,
+      "learning_rate": 6.444199781303311e-06,
+      "loss": 1.0317,
+      "step": 6633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5648391618067543,
+      "learning_rate": 6.443207440878624e-06,
+      "loss": 0.8468,
+      "step": 6634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.323982071962355,
+      "learning_rate": 6.442215038436097e-06,
+      "loss": 0.8747,
+      "step": 6635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1093543048837438,
+      "learning_rate": 6.441222574018378e-06,
+      "loss": 0.8862,
+      "step": 6636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5518904471895634,
+      "learning_rate": 6.440230047668116e-06,
+      "loss": 0.9014,
+      "step": 6637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.453187608037801,
+      "learning_rate": 6.4392374594279614e-06,
+      "loss": 0.7163,
+      "step": 6638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.92720321969246,
+      "learning_rate": 6.438244809340568e-06,
+      "loss": 0.7239,
+      "step": 6639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9768681421101995,
+      "learning_rate": 6.4372520974485916e-06,
+      "loss": 0.7639,
+      "step": 6640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.211523150379543,
+      "learning_rate": 6.436259323794691e-06,
+      "loss": 0.6764,
+      "step": 6641
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2180176039303383,
+      "learning_rate": 6.43526648842153e-06,
+      "loss": 0.6637,
+      "step": 6642
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8439515812291927,
+      "learning_rate": 6.434273591371771e-06,
+      "loss": 0.801,
+      "step": 6643
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2197625968710166,
+      "learning_rate": 6.43328063268808e-06,
+      "loss": 0.5946,
+      "step": 6644
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7556258702880245,
+      "learning_rate": 6.432287612413131e-06,
+      "loss": 0.7857,
+      "step": 6645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8147444759597382,
+      "learning_rate": 6.431294530589589e-06,
+      "loss": 0.6851,
+      "step": 6646
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8439365739498952,
+      "learning_rate": 6.4303013872601375e-06,
+      "loss": 0.786,
+      "step": 6647
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.201311944854229,
+      "learning_rate": 6.4293081824674485e-06,
+      "loss": 0.6594,
+      "step": 6648
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.575706105407568,
+      "learning_rate": 6.428314916254203e-06,
+      "loss": 0.8862,
+      "step": 6649
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.404262702554771,
+      "learning_rate": 6.427321588663085e-06,
+      "loss": 0.8453,
+      "step": 6650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.223779551325084,
+      "learning_rate": 6.42632819973678e-06,
+      "loss": 0.9188,
+      "step": 6651
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6427246485921057,
+      "learning_rate": 6.425334749517975e-06,
+      "loss": 0.7346,
+      "step": 6652
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8691223126963148,
+      "learning_rate": 6.424341238049362e-06,
+      "loss": 0.9689,
+      "step": 6653
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0519130154234957,
+      "learning_rate": 6.423347665373633e-06,
+      "loss": 0.9281,
+      "step": 6654
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3274054437210596,
+      "learning_rate": 6.422354031533485e-06,
+      "loss": 0.8359,
+      "step": 6655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9824973749988968,
+      "learning_rate": 6.421360336571618e-06,
+      "loss": 0.9536,
+      "step": 6656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9638425728674813,
+      "learning_rate": 6.42036658053073e-06,
+      "loss": 0.9725,
+      "step": 6657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7266296789924966,
+      "learning_rate": 6.419372763453529e-06,
+      "loss": 0.7007,
+      "step": 6658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.887329782771604,
+      "learning_rate": 6.418378885382716e-06,
+      "loss": 0.8237,
+      "step": 6659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.541274527648749,
+      "learning_rate": 6.417384946361005e-06,
+      "loss": 0.845,
+      "step": 6660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.158583369996596,
+      "learning_rate": 6.416390946431107e-06,
+      "loss": 0.785,
+      "step": 6661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1362351045233665,
+      "learning_rate": 6.4153968856357355e-06,
+      "loss": 0.6313,
+      "step": 6662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.480810547392666,
+      "learning_rate": 6.414402764017607e-06,
+      "loss": 0.9762,
+      "step": 6663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.071432906256072,
+      "learning_rate": 6.41340858161944e-06,
+      "loss": 0.7047,
+      "step": 6664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0883122447080043,
+      "learning_rate": 6.412414338483961e-06,
+      "loss": 0.8424,
+      "step": 6665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1156990853079334,
+      "learning_rate": 6.411420034653891e-06,
+      "loss": 0.7186,
+      "step": 6666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3412722152560064,
+      "learning_rate": 6.410425670171959e-06,
+      "loss": 0.8811,
+      "step": 6667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.772082033097107,
+      "learning_rate": 6.409431245080894e-06,
+      "loss": 0.6807,
+      "step": 6668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.308054452931174,
+      "learning_rate": 6.408436759423431e-06,
+      "loss": 0.7333,
+      "step": 6669
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.007754470729929,
+      "learning_rate": 6.4074422132423035e-06,
+      "loss": 0.7106,
+      "step": 6670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7847605238014135,
+      "learning_rate": 6.406447606580248e-06,
+      "loss": 0.7156,
+      "step": 6671
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0391961665165157,
+      "learning_rate": 6.405452939480008e-06,
+      "loss": 0.7455,
+      "step": 6672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7060826132933298,
+      "learning_rate": 6.404458211984324e-06,
+      "loss": 0.9054,
+      "step": 6673
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.26724806668429,
+      "learning_rate": 6.403463424135943e-06,
+      "loss": 0.8219,
+      "step": 6674
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3631929087388843,
+      "learning_rate": 6.402468575977614e-06,
+      "loss": 0.6242,
+      "step": 6675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2824524389167498,
+      "learning_rate": 6.401473667552085e-06,
+      "loss": 0.798,
+      "step": 6676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9494800408943918,
+      "learning_rate": 6.400478698902112e-06,
+      "loss": 0.8809,
+      "step": 6677
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6965407265195274,
+      "learning_rate": 6.399483670070451e-06,
+      "loss": 0.7751,
+      "step": 6678
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0792997394751427,
+      "learning_rate": 6.398488581099859e-06,
+      "loss": 0.9033,
+      "step": 6679
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.661405253879516,
+      "learning_rate": 6.3974934320330996e-06,
+      "loss": 0.8441,
+      "step": 6680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1612877538527706,
+      "learning_rate": 6.396498222912933e-06,
+      "loss": 0.6431,
+      "step": 6681
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9451078041618999,
+      "learning_rate": 6.395502953782129e-06,
+      "loss": 0.9893,
+      "step": 6682
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7306741661834233,
+      "learning_rate": 6.394507624683455e-06,
+      "loss": 0.8021,
+      "step": 6683
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9164025397634517,
+      "learning_rate": 6.393512235659681e-06,
+      "loss": 0.7206,
+      "step": 6684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1203189547904957,
+      "learning_rate": 6.392516786753586e-06,
+      "loss": 0.7009,
+      "step": 6685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.213390997339428,
+      "learning_rate": 6.391521278007939e-06,
+      "loss": 0.7943,
+      "step": 6686
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7135990616474177,
+      "learning_rate": 6.390525709465527e-06,
+      "loss": 0.9561,
+      "step": 6687
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9672684513505866,
+      "learning_rate": 6.389530081169126e-06,
+      "loss": 0.692,
+      "step": 6688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.0012049337212305,
+      "learning_rate": 6.388534393161525e-06,
+      "loss": 0.8806,
+      "step": 6689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.279941604977279,
+      "learning_rate": 6.38753864548551e-06,
+      "loss": 0.8623,
+      "step": 6690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.152159343549849,
+      "learning_rate": 6.386542838183865e-06,
+      "loss": 0.6474,
+      "step": 6691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.246424931003352,
+      "learning_rate": 6.385546971299389e-06,
+      "loss": 0.6815,
+      "step": 6692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0028839430136578,
+      "learning_rate": 6.384551044874871e-06,
+      "loss": 0.63,
+      "step": 6693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.86178162335091,
+      "learning_rate": 6.383555058953115e-06,
+      "loss": 0.977,
+      "step": 6694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7952078186441414,
+      "learning_rate": 6.382559013576914e-06,
+      "loss": 0.8186,
+      "step": 6695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1426640973333066,
+      "learning_rate": 6.381562908789075e-06,
+      "loss": 0.6866,
+      "step": 6696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1416790499316718,
+      "learning_rate": 6.380566744632399e-06,
+      "loss": 0.6711,
+      "step": 6697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.51698189600209,
+      "learning_rate": 6.379570521149696e-06,
+      "loss": 0.7704,
+      "step": 6698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.983096217386921,
+      "learning_rate": 6.378574238383776e-06,
+      "loss": 0.7195,
+      "step": 6699
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2623755009193576,
+      "learning_rate": 6.377577896377451e-06,
+      "loss": 0.7771,
+      "step": 6700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0787093374487062,
+      "learning_rate": 6.3765814951735346e-06,
+      "loss": 0.8124,
+      "step": 6701
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1002168363525824,
+      "learning_rate": 6.375585034814845e-06,
+      "loss": 1.0641,
+      "step": 6702
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.113614116792777,
+      "learning_rate": 6.374588515344204e-06,
+      "loss": 0.8909,
+      "step": 6703
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2458938854592927,
+      "learning_rate": 6.373591936804433e-06,
+      "loss": 0.881,
+      "step": 6704
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.649200052378964,
+      "learning_rate": 6.372595299238357e-06,
+      "loss": 0.715,
+      "step": 6705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.837760413448331,
+      "learning_rate": 6.3715986026888046e-06,
+      "loss": 0.8027,
+      "step": 6706
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1485791271280816,
+      "learning_rate": 6.370601847198606e-06,
+      "loss": 0.7736,
+      "step": 6707
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.049327371009982,
+      "learning_rate": 6.3696050328105915e-06,
+      "loss": 0.6192,
+      "step": 6708
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2386643596189502,
+      "learning_rate": 6.3686081595676e-06,
+      "loss": 0.6454,
+      "step": 6709
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0208421894786524,
+      "learning_rate": 6.367611227512468e-06,
+      "loss": 0.9422,
+      "step": 6710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1736669328620883,
+      "learning_rate": 6.366614236688034e-06,
+      "loss": 0.7588,
+      "step": 6711
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9103144539571948,
+      "learning_rate": 6.365617187137144e-06,
+      "loss": 1.0054,
+      "step": 6712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.18169034007297,
+      "learning_rate": 6.3646200789026426e-06,
+      "loss": 0.9464,
+      "step": 6713
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.077746082413079,
+      "learning_rate": 6.3636229120273766e-06,
+      "loss": 0.6275,
+      "step": 6714
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1323193514077894,
+      "learning_rate": 6.362625686554196e-06,
+      "loss": 1.011,
+      "step": 6715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.664452781899257,
+      "learning_rate": 6.361628402525956e-06,
+      "loss": 0.7363,
+      "step": 6716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2566260831631866,
+      "learning_rate": 6.360631059985512e-06,
+      "loss": 0.6271,
+      "step": 6717
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.341173961713595,
+      "learning_rate": 6.35963365897572e-06,
+      "loss": 0.899,
+      "step": 6718
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.113925951030236,
+      "learning_rate": 6.3586361995394415e-06,
+      "loss": 0.9843,
+      "step": 6719
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.848451684102044,
+      "learning_rate": 6.35763868171954e-06,
+      "loss": 0.6654,
+      "step": 6720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.107145807323324,
+      "learning_rate": 6.3566411055588806e-06,
+      "loss": 0.8801,
+      "step": 6721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1029513775504662,
+      "learning_rate": 6.355643471100332e-06,
+      "loss": 0.7063,
+      "step": 6722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.357084993727709,
+      "learning_rate": 6.354645778386763e-06,
+      "loss": 0.8788,
+      "step": 6723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.822608764821642,
+      "learning_rate": 6.353648027461048e-06,
+      "loss": 0.8259,
+      "step": 6724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6161069429332438,
+      "learning_rate": 6.3526502183660646e-06,
+      "loss": 0.8369,
+      "step": 6725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.195234129408785,
+      "learning_rate": 6.351652351144687e-06,
+      "loss": 0.8588,
+      "step": 6726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1803574146393316,
+      "learning_rate": 6.350654425839799e-06,
+      "loss": 0.7566,
+      "step": 6727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.16336094807374,
+      "learning_rate": 6.349656442494281e-06,
+      "loss": 0.6732,
+      "step": 6728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0716718019867595,
+      "learning_rate": 6.348658401151018e-06,
+      "loss": 0.8972,
+      "step": 6729
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.022412016635643,
+      "learning_rate": 6.347660301852903e-06,
+      "loss": 0.8955,
+      "step": 6730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8521085527899346,
+      "learning_rate": 6.346662144642823e-06,
+      "loss": 0.6882,
+      "step": 6731
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7674736632694152,
+      "learning_rate": 6.345663929563672e-06,
+      "loss": 0.5168,
+      "step": 6732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9491187935440482,
+      "learning_rate": 6.344665656658342e-06,
+      "loss": 0.8741,
+      "step": 6733
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1996336852644034,
+      "learning_rate": 6.343667325969736e-06,
+      "loss": 0.8924,
+      "step": 6734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0688038033795502,
+      "learning_rate": 6.342668937540752e-06,
+      "loss": 0.8261,
+      "step": 6735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5804936009910846,
+      "learning_rate": 6.341670491414294e-06,
+      "loss": 0.8559,
+      "step": 6736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5867227516157167,
+      "learning_rate": 6.3406719876332675e-06,
+      "loss": 0.8583,
+      "step": 6737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0762549606922438,
+      "learning_rate": 6.339673426240578e-06,
+      "loss": 0.7348,
+      "step": 6738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2467134430574673,
+      "learning_rate": 6.3386748072791395e-06,
+      "loss": 0.8015,
+      "step": 6739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3151064972779523,
+      "learning_rate": 6.337676130791861e-06,
+      "loss": 0.8398,
+      "step": 6740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5165798615375494,
+      "learning_rate": 6.336677396821663e-06,
+      "loss": 0.8112,
+      "step": 6741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0877632438506644,
+      "learning_rate": 6.3356786054114585e-06,
+      "loss": 0.7278,
+      "step": 6742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.688385681333364,
+      "learning_rate": 6.3346797566041695e-06,
+      "loss": 0.8824,
+      "step": 6743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9683214483976808,
+      "learning_rate": 6.33368085044272e-06,
+      "loss": 0.8155,
+      "step": 6744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9419375754436434,
+      "learning_rate": 6.332681886970032e-06,
+      "loss": 0.9649,
+      "step": 6745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.022732711460188,
+      "learning_rate": 6.331682866229039e-06,
+      "loss": 0.7457,
+      "step": 6746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9387564136330286,
+      "learning_rate": 6.3306837882626635e-06,
+      "loss": 0.9326,
+      "step": 6747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.432251575776431,
+      "learning_rate": 6.3296846531138445e-06,
+      "loss": 0.8256,
+      "step": 6748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8511668155839573,
+      "learning_rate": 6.328685460825512e-06,
+      "loss": 0.8086,
+      "step": 6749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.378289463732398,
+      "learning_rate": 6.327686211440609e-06,
+      "loss": 0.8462,
+      "step": 6750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2197299232383876,
+      "learning_rate": 6.32668690500207e-06,
+      "loss": 1.0092,
+      "step": 6751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.883253304146279,
+      "learning_rate": 6.325687541552843e-06,
+      "loss": 0.8019,
+      "step": 6752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3188164430248204,
+      "learning_rate": 6.324688121135868e-06,
+      "loss": 0.7514,
+      "step": 6753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0755518494870002,
+      "learning_rate": 6.323688643794094e-06,
+      "loss": 0.6766,
+      "step": 6754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2615931233017625,
+      "learning_rate": 6.322689109570472e-06,
+      "loss": 0.9933,
+      "step": 6755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.826501719161511,
+      "learning_rate": 6.3216895185079515e-06,
+      "loss": 0.9513,
+      "step": 6756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0857756664438996,
+      "learning_rate": 6.3206898706494915e-06,
+      "loss": 0.7561,
+      "step": 6757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.260293675677505,
+      "learning_rate": 6.3196901660380436e-06,
+      "loss": 0.7936,
+      "step": 6758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2557326805945537,
+      "learning_rate": 6.318690404716572e-06,
+      "loss": 0.7046,
+      "step": 6759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.931848808864083,
+      "learning_rate": 6.317690586728036e-06,
+      "loss": 0.8303,
+      "step": 6760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9168805216515743,
+      "learning_rate": 6.316690712115401e-06,
+      "loss": 0.7885,
+      "step": 6761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4556528359980585,
+      "learning_rate": 6.315690780921634e-06,
+      "loss": 0.9003,
+      "step": 6762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8137915053976132,
+      "learning_rate": 6.314690793189703e-06,
+      "loss": 0.7795,
+      "step": 6763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.0268215246641255,
+      "learning_rate": 6.313690748962582e-06,
+      "loss": 0.7052,
+      "step": 6764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5584729806428697,
+      "learning_rate": 6.312690648283243e-06,
+      "loss": 0.7595,
+      "step": 6765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0331070554179151,
+      "learning_rate": 6.311690491194664e-06,
+      "loss": 0.6225,
+      "step": 6766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.36767275173424,
+      "learning_rate": 6.310690277739821e-06,
+      "loss": 1.0065,
+      "step": 6767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0555131417177748,
+      "learning_rate": 6.3096900079617e-06,
+      "loss": 0.6758,
+      "step": 6768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.202754676351193,
+      "learning_rate": 6.3086896819032814e-06,
+      "loss": 0.7071,
+      "step": 6769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3762777519546634,
+      "learning_rate": 6.307689299607555e-06,
+      "loss": 0.8864,
+      "step": 6770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.023539534245134,
+      "learning_rate": 6.306688861117505e-06,
+      "loss": 0.8671,
+      "step": 6771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5656347001297175,
+      "learning_rate": 6.305688366476124e-06,
+      "loss": 0.8094,
+      "step": 6772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.63002275917374,
+      "learning_rate": 6.304687815726406e-06,
+      "loss": 0.7281,
+      "step": 6773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.181068972885403,
+      "learning_rate": 6.303687208911348e-06,
+      "loss": 0.8677,
+      "step": 6774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5866669073133344,
+      "learning_rate": 6.302686546073946e-06,
+      "loss": 0.9349,
+      "step": 6775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0777450353331126,
+      "learning_rate": 6.301685827257202e-06,
+      "loss": 0.9,
+      "step": 6776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.672788176981851,
+      "learning_rate": 6.300685052504119e-06,
+      "loss": 0.9051,
+      "step": 6777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8351070272944545,
+      "learning_rate": 6.299684221857705e-06,
+      "loss": 0.8268,
+      "step": 6778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2269642549911997,
+      "learning_rate": 6.298683335360962e-06,
+      "loss": 0.6131,
+      "step": 6779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.696786336418103,
+      "learning_rate": 6.297682393056906e-06,
+      "loss": 1.0461,
+      "step": 6780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1144459261819586,
+      "learning_rate": 6.296681394988546e-06,
+      "loss": 0.7881,
+      "step": 6781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.336367324097868,
+      "learning_rate": 6.2956803411989e-06,
+      "loss": 0.8342,
+      "step": 6782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.355883394374581,
+      "learning_rate": 6.294679231730983e-06,
+      "loss": 0.8618,
+      "step": 6783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1384685393731324,
+      "learning_rate": 6.293678066627816e-06,
+      "loss": 0.9207,
+      "step": 6784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0959367677983036,
+      "learning_rate": 6.292676845932422e-06,
+      "loss": 0.7951,
+      "step": 6785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6609418205730857,
+      "learning_rate": 6.291675569687824e-06,
+      "loss": 0.973,
+      "step": 6786
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.253130169019891,
+      "learning_rate": 6.290674237937051e-06,
+      "loss": 0.8997,
+      "step": 6787
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.390934641493984,
+      "learning_rate": 6.289672850723132e-06,
+      "loss": 1.0802,
+      "step": 6788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.87108410527697,
+      "learning_rate": 6.288671408089098e-06,
+      "loss": 0.7647,
+      "step": 6789
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7703487248877714,
+      "learning_rate": 6.2876699100779815e-06,
+      "loss": 0.829,
+      "step": 6790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3131797826029623,
+      "learning_rate": 6.2866683567328235e-06,
+      "loss": 0.594,
+      "step": 6791
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0282086636161236,
+      "learning_rate": 6.285666748096659e-06,
+      "loss": 0.801,
+      "step": 6792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0783062224801707,
+      "learning_rate": 6.284665084212533e-06,
+      "loss": 0.8481,
+      "step": 6793
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4518459085442275,
+      "learning_rate": 6.283663365123486e-06,
+      "loss": 0.8806,
+      "step": 6794
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2730505165895565,
+      "learning_rate": 6.282661590872565e-06,
+      "loss": 0.9286,
+      "step": 6795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3126874721403468,
+      "learning_rate": 6.281659761502819e-06,
+      "loss": 0.9038,
+      "step": 6796
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1950049100840365,
+      "learning_rate": 6.2806578770573e-06,
+      "loss": 0.7797,
+      "step": 6797
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0308991046716334,
+      "learning_rate": 6.279655937579058e-06,
+      "loss": 0.8448,
+      "step": 6798
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2894525087865,
+      "learning_rate": 6.278653943111152e-06,
+      "loss": 0.8692,
+      "step": 6799
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.066483158638273,
+      "learning_rate": 6.277651893696637e-06,
+      "loss": 0.8445,
+      "step": 6800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1552955237644997,
+      "learning_rate": 6.276649789378575e-06,
+      "loss": 0.708,
+      "step": 6801
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.695453171058064,
+      "learning_rate": 6.275647630200029e-06,
+      "loss": 0.8584,
+      "step": 6802
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8815851701909208,
+      "learning_rate": 6.274645416204062e-06,
+      "loss": 0.8222,
+      "step": 6803
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0942114678600006,
+      "learning_rate": 6.273643147433743e-06,
+      "loss": 0.9157,
+      "step": 6804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5593350822525873,
+      "learning_rate": 6.272640823932141e-06,
+      "loss": 0.9001,
+      "step": 6805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.521083398049253,
+      "learning_rate": 6.271638445742332e-06,
+      "loss": 0.6955,
+      "step": 6806
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.223249246089247,
+      "learning_rate": 6.270636012907383e-06,
+      "loss": 0.9672,
+      "step": 6807
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.246541949051622,
+      "learning_rate": 6.269633525470376e-06,
+      "loss": 0.8366,
+      "step": 6808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.289607273508345,
+      "learning_rate": 6.268630983474388e-06,
+      "loss": 0.7601,
+      "step": 6809
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8518624194896576,
+      "learning_rate": 6.267628386962502e-06,
+      "loss": 0.7395,
+      "step": 6810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.806270088767103,
+      "learning_rate": 6.266625735977802e-06,
+      "loss": 0.7394,
+      "step": 6811
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.551953046444896,
+      "learning_rate": 6.265623030563372e-06,
+      "loss": 0.8752,
+      "step": 6812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0829680456938364,
+      "learning_rate": 6.264620270762302e-06,
+      "loss": 0.9285,
+      "step": 6813
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2365188674020167,
+      "learning_rate": 6.263617456617681e-06,
+      "loss": 0.6832,
+      "step": 6814
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2594355926537486,
+      "learning_rate": 6.262614588172607e-06,
+      "loss": 0.924,
+      "step": 6815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3424015518718475,
+      "learning_rate": 6.261611665470168e-06,
+      "loss": 0.8522,
+      "step": 6816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1311246865449474,
+      "learning_rate": 6.260608688553469e-06,
+      "loss": 0.7498,
+      "step": 6817
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.344420342156125,
+      "learning_rate": 6.259605657465607e-06,
+      "loss": 0.8234,
+      "step": 6818
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5483607484374877,
+      "learning_rate": 6.258602572249683e-06,
+      "loss": 0.6576,
+      "step": 6819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0003514343298003,
+      "learning_rate": 6.2575994329488045e-06,
+      "loss": 0.9355,
+      "step": 6820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8293761559687807,
+      "learning_rate": 6.2565962396060774e-06,
+      "loss": 0.7823,
+      "step": 6821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3513360579211273,
+      "learning_rate": 6.255592992264612e-06,
+      "loss": 1.0285,
+      "step": 6822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2069786290285254,
+      "learning_rate": 6.254589690967517e-06,
+      "loss": 0.6228,
+      "step": 6823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0497857686864145,
+      "learning_rate": 6.2535863357579105e-06,
+      "loss": 0.5293,
+      "step": 6824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.347943761017813,
+      "learning_rate": 6.252582926678908e-06,
+      "loss": 0.9247,
+      "step": 6825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.244628135427385,
+      "learning_rate": 6.2515794637736275e-06,
+      "loss": 0.888,
+      "step": 6826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6627160298713193,
+      "learning_rate": 6.250575947085189e-06,
+      "loss": 1.0345,
+      "step": 6827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8433351908406976,
+      "learning_rate": 6.2495723766567165e-06,
+      "loss": 0.7965,
+      "step": 6828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1640460907979702,
+      "learning_rate": 6.248568752531337e-06,
+      "loss": 0.6436,
+      "step": 6829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6764816012351673,
+      "learning_rate": 6.247565074752177e-06,
+      "loss": 0.6006,
+      "step": 6830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.238549263607674,
+      "learning_rate": 6.246561343362368e-06,
+      "loss": 0.8287,
+      "step": 6831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.937906162982686,
+      "learning_rate": 6.24555755840504e-06,
+      "loss": 0.9546,
+      "step": 6832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.964888887228792,
+      "learning_rate": 6.244553719923332e-06,
+      "loss": 0.7006,
+      "step": 6833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8450720405603396,
+      "learning_rate": 6.243549827960378e-06,
+      "loss": 0.7994,
+      "step": 6834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.642756742263657,
+      "learning_rate": 6.242545882559318e-06,
+      "loss": 0.9059,
+      "step": 6835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.317117131931308,
+      "learning_rate": 6.241541883763294e-06,
+      "loss": 0.8212,
+      "step": 6836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0705288009662257,
+      "learning_rate": 6.240537831615451e-06,
+      "loss": 0.9096,
+      "step": 6837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6453179389974504,
+      "learning_rate": 6.239533726158934e-06,
+      "loss": 0.8202,
+      "step": 6838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1150758573768738,
+      "learning_rate": 6.238529567436892e-06,
+      "loss": 0.7221,
+      "step": 6839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.037497119832672,
+      "learning_rate": 6.2375253554924766e-06,
+      "loss": 0.8211,
+      "step": 6840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0929111964322809,
+      "learning_rate": 6.23652109036884e-06,
+      "loss": 0.7229,
+      "step": 6841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1688891858844834,
+      "learning_rate": 6.235516772109138e-06,
+      "loss": 0.9423,
+      "step": 6842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2129522285332732,
+      "learning_rate": 6.23451240075653e-06,
+      "loss": 0.9151,
+      "step": 6843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.956216912037655,
+      "learning_rate": 6.233507976354174e-06,
+      "loss": 0.8937,
+      "step": 6844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.547942527460975,
+      "learning_rate": 6.2325034989452335e-06,
+      "loss": 0.8374,
+      "step": 6845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8790536961472493,
+      "learning_rate": 6.231498968572872e-06,
+      "loss": 0.8186,
+      "step": 6846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2932570586003598,
+      "learning_rate": 6.230494385280257e-06,
+      "loss": 0.825,
+      "step": 6847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2558441480851992,
+      "learning_rate": 6.229489749110559e-06,
+      "loss": 0.645,
+      "step": 6848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3336036059643326,
+      "learning_rate": 6.228485060106948e-06,
+      "loss": 0.8605,
+      "step": 6849
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8260960163164426,
+      "learning_rate": 6.227480318312597e-06,
+      "loss": 0.8863,
+      "step": 6850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9098955490841134,
+      "learning_rate": 6.226475523770683e-06,
+      "loss": 0.5992,
+      "step": 6851
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8925032966666648,
+      "learning_rate": 6.225470676524385e-06,
+      "loss": 0.874,
+      "step": 6852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4551912494986337,
+      "learning_rate": 6.224465776616883e-06,
+      "loss": 0.6171,
+      "step": 6853
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7293668235454407,
+      "learning_rate": 6.223460824091358e-06,
+      "loss": 0.8761,
+      "step": 6854
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.346315826462712,
+      "learning_rate": 6.222455818990998e-06,
+      "loss": 0.9622,
+      "step": 6855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0390798644882846,
+      "learning_rate": 6.221450761358988e-06,
+      "loss": 0.7997,
+      "step": 6856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2604722162553166,
+      "learning_rate": 6.220445651238519e-06,
+      "loss": 0.7041,
+      "step": 6857
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0577104051729815,
+      "learning_rate": 6.2194404886727835e-06,
+      "loss": 0.7937,
+      "step": 6858
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.959375152098361,
+      "learning_rate": 6.218435273704973e-06,
+      "loss": 0.9215,
+      "step": 6859
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.676133928561902,
+      "learning_rate": 6.217430006378285e-06,
+      "loss": 0.7039,
+      "step": 6860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3236434365756846,
+      "learning_rate": 6.216424686735917e-06,
+      "loss": 0.8873,
+      "step": 6861
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.088056418789541,
+      "learning_rate": 6.215419314821074e-06,
+      "loss": 0.8586,
+      "step": 6862
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5038353097633106,
+      "learning_rate": 6.214413890676956e-06,
+      "loss": 0.8317,
+      "step": 6863
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5924907642990447,
+      "learning_rate": 6.213408414346765e-06,
+      "loss": 0.9211,
+      "step": 6864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1089892332561178,
+      "learning_rate": 6.212402885873715e-06,
+      "loss": 0.8013,
+      "step": 6865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.448918102619776,
+      "learning_rate": 6.211397305301011e-06,
+      "loss": 0.967,
+      "step": 6866
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9607222855315785,
+      "learning_rate": 6.210391672671869e-06,
+      "loss": 0.9697,
+      "step": 6867
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.162551116898402,
+      "learning_rate": 6.2093859880295e-06,
+      "loss": 0.8737,
+      "step": 6868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2374089041289424,
+      "learning_rate": 6.208380251417122e-06,
+      "loss": 0.7123,
+      "step": 6869
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3261822880930647,
+      "learning_rate": 6.207374462877953e-06,
+      "loss": 0.84,
+      "step": 6870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.12908456699904,
+      "learning_rate": 6.2063686224552144e-06,
+      "loss": 0.849,
+      "step": 6871
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.155210257575057,
+      "learning_rate": 6.205362730192131e-06,
+      "loss": 0.8251,
+      "step": 6872
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8082560929165152,
+      "learning_rate": 6.2043567861319245e-06,
+      "loss": 0.8525,
+      "step": 6873
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1650844316191975,
+      "learning_rate": 6.203350790317825e-06,
+      "loss": 0.9502,
+      "step": 6874
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.370819295436287,
+      "learning_rate": 6.2023447427930615e-06,
+      "loss": 0.9506,
+      "step": 6875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9696622307951461,
+      "learning_rate": 6.201338643600869e-06,
+      "loss": 0.6773,
+      "step": 6876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0560487278207384,
+      "learning_rate": 6.2003324927844765e-06,
+      "loss": 0.9504,
+      "step": 6877
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3509197187694824,
+      "learning_rate": 6.199326290387126e-06,
+      "loss": 0.6649,
+      "step": 6878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.925110358947684,
+      "learning_rate": 6.198320036452051e-06,
+      "loss": 0.7567,
+      "step": 6879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7123807650214617,
+      "learning_rate": 6.197313731022497e-06,
+      "loss": 0.8816,
+      "step": 6880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7371707478681084,
+      "learning_rate": 6.196307374141707e-06,
+      "loss": 0.7197,
+      "step": 6881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.870781016489268,
+      "learning_rate": 6.195300965852923e-06,
+      "loss": 0.8643,
+      "step": 6882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.988293718652391,
+      "learning_rate": 6.194294506199394e-06,
+      "loss": 0.6199,
+      "step": 6883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8144247140451641,
+      "learning_rate": 6.193287995224371e-06,
+      "loss": 0.8448,
+      "step": 6884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.389907577991923,
+      "learning_rate": 6.192281432971104e-06,
+      "loss": 0.7633,
+      "step": 6885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2718786448788406,
+      "learning_rate": 6.19127481948285e-06,
+      "loss": 0.7193,
+      "step": 6886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7509158414810784,
+      "learning_rate": 6.1902681548028646e-06,
+      "loss": 0.6909,
+      "step": 6887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.067821000798556,
+      "learning_rate": 6.189261438974403e-06,
+      "loss": 0.9715,
+      "step": 6888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.886446022312674,
+      "learning_rate": 6.18825467204073e-06,
+      "loss": 0.6962,
+      "step": 6889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.965297938412317,
+      "learning_rate": 6.187247854045109e-06,
+      "loss": 0.7765,
+      "step": 6890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1262509702530124,
+      "learning_rate": 6.186240985030802e-06,
+      "loss": 0.6973,
+      "step": 6891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9459036621903943,
+      "learning_rate": 6.1852340650410784e-06,
+      "loss": 0.6025,
+      "step": 6892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.983597917542116,
+      "learning_rate": 6.184227094119207e-06,
+      "loss": 0.7565,
+      "step": 6893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1517484008491647,
+      "learning_rate": 6.183220072308459e-06,
+      "loss": 0.741,
+      "step": 6894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.370075980411061,
+      "learning_rate": 6.1822129996521105e-06,
+      "loss": 0.7866,
+      "step": 6895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0735253682488053,
+      "learning_rate": 6.181205876193437e-06,
+      "loss": 0.7818,
+      "step": 6896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0028247119038745,
+      "learning_rate": 6.180198701975715e-06,
+      "loss": 0.6635,
+      "step": 6897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.196442993543736,
+      "learning_rate": 6.179191477042227e-06,
+      "loss": 0.8105,
+      "step": 6898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.224379178512585,
+      "learning_rate": 6.178184201436256e-06,
+      "loss": 0.6633,
+      "step": 6899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.479582909335687,
+      "learning_rate": 6.177176875201086e-06,
+      "loss": 0.9109,
+      "step": 6900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1269326972112808,
+      "learning_rate": 6.176169498380001e-06,
+      "loss": 0.8422,
+      "step": 6901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5359163173665396,
+      "learning_rate": 6.175162071016295e-06,
+      "loss": 0.6575,
+      "step": 6902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9999298093578242,
+      "learning_rate": 6.174154593153258e-06,
+      "loss": 0.7067,
+      "step": 6903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.85582167104969,
+      "learning_rate": 6.173147064834183e-06,
+      "loss": 0.7564,
+      "step": 6904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.020042384533157,
+      "learning_rate": 6.1721394861023665e-06,
+      "loss": 0.7739,
+      "step": 6905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.261398394788983,
+      "learning_rate": 6.171131857001104e-06,
+      "loss": 0.6859,
+      "step": 6906
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.568355774161419,
+      "learning_rate": 6.170124177573698e-06,
+      "loss": 0.7181,
+      "step": 6907
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1860883256344887,
+      "learning_rate": 6.169116447863451e-06,
+      "loss": 0.7121,
+      "step": 6908
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0510655945521745,
+      "learning_rate": 6.168108667913666e-06,
+      "loss": 0.7112,
+      "step": 6909
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.01018813420199,
+      "learning_rate": 6.16710083776765e-06,
+      "loss": 0.969,
+      "step": 6910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.033768579663392,
+      "learning_rate": 6.1660929574687115e-06,
+      "loss": 0.6449,
+      "step": 6911
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7395044515937206,
+      "learning_rate": 6.165085027060162e-06,
+      "loss": 0.8384,
+      "step": 6912
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1519752864357513,
+      "learning_rate": 6.164077046585314e-06,
+      "loss": 0.8674,
+      "step": 6913
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9328362231998868,
+      "learning_rate": 6.163069016087483e-06,
+      "loss": 0.7671,
+      "step": 6914
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.728066523153197,
+      "learning_rate": 6.162060935609985e-06,
+      "loss": 0.8587,
+      "step": 6915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1049497219196556,
+      "learning_rate": 6.161052805196141e-06,
+      "loss": 0.8093,
+      "step": 6916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7547971223667376,
+      "learning_rate": 6.160044624889272e-06,
+      "loss": 0.84,
+      "step": 6917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4529542282521195,
+      "learning_rate": 6.1590363947327025e-06,
+      "loss": 0.9284,
+      "step": 6918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.087505803666383,
+      "learning_rate": 6.158028114769758e-06,
+      "loss": 0.8999,
+      "step": 6919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8347273512738873,
+      "learning_rate": 6.157019785043765e-06,
+      "loss": 0.8327,
+      "step": 6920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1787113097788353,
+      "learning_rate": 6.156011405598057e-06,
+      "loss": 0.8615,
+      "step": 6921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.836607872231491,
+      "learning_rate": 6.155002976475962e-06,
+      "loss": 0.8477,
+      "step": 6922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.376410758719311,
+      "learning_rate": 6.15399449772082e-06,
+      "loss": 0.7824,
+      "step": 6923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.949230470170662,
+      "learning_rate": 6.152985969375962e-06,
+      "loss": 0.8352,
+      "step": 6924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.901670392369535,
+      "learning_rate": 6.151977391484729e-06,
+      "loss": 0.8406,
+      "step": 6925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.082699344792226,
+      "learning_rate": 6.150968764090461e-06,
+      "loss": 0.6752,
+      "step": 6926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3298169597992238,
+      "learning_rate": 6.149960087236505e-06,
+      "loss": 0.8597,
+      "step": 6927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1440234547341768,
+      "learning_rate": 6.1489513609662025e-06,
+      "loss": 0.6423,
+      "step": 6928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1658163134322015,
+      "learning_rate": 6.147942585322898e-06,
+      "loss": 0.8778,
+      "step": 6929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9473266372673983,
+      "learning_rate": 6.146933760349947e-06,
+      "loss": 0.9974,
+      "step": 6930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.642751870682567,
+      "learning_rate": 6.145924886090695e-06,
+      "loss": 0.9953,
+      "step": 6931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.021036965453424,
+      "learning_rate": 6.144915962588503e-06,
+      "loss": 0.9721,
+      "step": 6932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.931149962159902,
+      "learning_rate": 6.143906989886718e-06,
+      "loss": 0.6675,
+      "step": 6933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.485926427670658,
+      "learning_rate": 6.142897968028704e-06,
+      "loss": 0.9533,
+      "step": 6934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3099063203548849,
+      "learning_rate": 6.141888897057817e-06,
+      "loss": 0.6238,
+      "step": 6935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1708992185647293,
+      "learning_rate": 6.140879777017423e-06,
+      "loss": 0.9599,
+      "step": 6936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2748999345926255,
+      "learning_rate": 6.139870607950885e-06,
+      "loss": 0.9182,
+      "step": 6937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.472676752995953,
+      "learning_rate": 6.138861389901566e-06,
+      "loss": 0.9173,
+      "step": 6938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.425601418681153,
+      "learning_rate": 6.137852122912839e-06,
+      "loss": 0.8561,
+      "step": 6939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.977641625043964,
+      "learning_rate": 6.136842807028069e-06,
+      "loss": 0.7731,
+      "step": 6940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.154363770938921,
+      "learning_rate": 6.135833442290634e-06,
+      "loss": 0.7515,
+      "step": 6941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.266341714085117,
+      "learning_rate": 6.134824028743906e-06,
+      "loss": 0.8268,
+      "step": 6942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.991352230993232,
+      "learning_rate": 6.133814566431262e-06,
+      "loss": 0.7469,
+      "step": 6943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2412128524831028,
+      "learning_rate": 6.1328050553960804e-06,
+      "loss": 0.8979,
+      "step": 6944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0044127437395605,
+      "learning_rate": 6.131795495681743e-06,
+      "loss": 0.9194,
+      "step": 6945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.269697889936579,
+      "learning_rate": 6.130785887331633e-06,
+      "loss": 0.9165,
+      "step": 6946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.199854320726057,
+      "learning_rate": 6.129776230389135e-06,
+      "loss": 0.7996,
+      "step": 6947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.618965453907698,
+      "learning_rate": 6.128766524897636e-06,
+      "loss": 0.6868,
+      "step": 6948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0030421888699619,
+      "learning_rate": 6.1277567709005245e-06,
+      "loss": 0.6925,
+      "step": 6949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1393593556434847,
+      "learning_rate": 6.126746968441195e-06,
+      "loss": 0.8929,
+      "step": 6950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1806759167619472,
+      "learning_rate": 6.1257371175630375e-06,
+      "loss": 0.7208,
+      "step": 6951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9963492587499687,
+      "learning_rate": 6.124727218309449e-06,
+      "loss": 0.7826,
+      "step": 6952
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.623659983066114,
+      "learning_rate": 6.123717270723827e-06,
+      "loss": 0.7492,
+      "step": 6953
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9092329256131104,
+      "learning_rate": 6.122707274849572e-06,
+      "loss": 0.6836,
+      "step": 6954
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3312642459581279,
+      "learning_rate": 6.121697230730084e-06,
+      "loss": 0.6892,
+      "step": 6955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.326291393079857,
+      "learning_rate": 6.120687138408768e-06,
+      "loss": 0.8169,
+      "step": 6956
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5995489835679093,
+      "learning_rate": 6.11967699792903e-06,
+      "loss": 0.7586,
+      "step": 6957
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0806019613369215,
+      "learning_rate": 6.118666809334277e-06,
+      "loss": 0.6927,
+      "step": 6958
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4591262260146913,
+      "learning_rate": 6.117656572667921e-06,
+      "loss": 0.7735,
+      "step": 6959
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.657500651716769,
+      "learning_rate": 6.116646287973374e-06,
+      "loss": 0.4678,
+      "step": 6960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0519215726260347,
+      "learning_rate": 6.115635955294047e-06,
+      "loss": 0.903,
+      "step": 6961
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.291509713183886,
+      "learning_rate": 6.114625574673359e-06,
+      "loss": 0.9024,
+      "step": 6962
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9608126949181173,
+      "learning_rate": 6.1136151461547266e-06,
+      "loss": 0.7916,
+      "step": 6963
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.883939020959031,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.8048,
+      "step": 6964
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2878525809476313,
+      "learning_rate": 6.111594145597319e-06,
+      "loss": 0.7597,
+      "step": 6965
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.67274779238878,
+      "learning_rate": 6.1105835736453874e-06,
+      "loss": 0.657,
+      "step": 6966
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.400664590316094,
+      "learning_rate": 6.109572953969206e-06,
+      "loss": 0.7463,
+      "step": 6967
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.030157020574552,
+      "learning_rate": 6.108562286612204e-06,
+      "loss": 0.6856,
+      "step": 6968
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.882040818299548,
+      "learning_rate": 6.107551571617813e-06,
+      "loss": 0.6515,
+      "step": 6969
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5777737406382926,
+      "learning_rate": 6.106540809029463e-06,
+      "loss": 0.8583,
+      "step": 6970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.689135307800423,
+      "learning_rate": 6.105529998890589e-06,
+      "loss": 0.8282,
+      "step": 6971
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8385825809914833,
+      "learning_rate": 6.104519141244631e-06,
+      "loss": 0.8921,
+      "step": 6972
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8281436191321294,
+      "learning_rate": 6.103508236135025e-06,
+      "loss": 0.9245,
+      "step": 6973
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.38712270765917,
+      "learning_rate": 6.1024972836052135e-06,
+      "loss": 0.8902,
+      "step": 6974
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5459267116036701,
+      "learning_rate": 6.1014862836986364e-06,
+      "loss": 0.6974,
+      "step": 6975
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8605239762930306,
+      "learning_rate": 6.10047523645874e-06,
+      "loss": 0.7709,
+      "step": 6976
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4738276134207653,
+      "learning_rate": 6.099464141928973e-06,
+      "loss": 0.7369,
+      "step": 6977
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.421935877703611,
+      "learning_rate": 6.098453000152784e-06,
+      "loss": 0.7525,
+      "step": 6978
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.927278682903556,
+      "learning_rate": 6.0974418111736235e-06,
+      "loss": 1.0042,
+      "step": 6979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2560291769298546,
+      "learning_rate": 6.096430575034942e-06,
+      "loss": 0.7802,
+      "step": 6980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2415980353760543,
+      "learning_rate": 6.095419291780198e-06,
+      "loss": 0.698,
+      "step": 6981
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.550553513344467,
+      "learning_rate": 6.0944079614528475e-06,
+      "loss": 0.861,
+      "step": 6982
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5050573297504912,
+      "learning_rate": 6.0933965840963494e-06,
+      "loss": 0.7858,
+      "step": 6983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0234144802849356,
+      "learning_rate": 6.092385159754165e-06,
+      "loss": 0.7004,
+      "step": 6984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.270967903087202,
+      "learning_rate": 6.091373688469757e-06,
+      "loss": 0.6705,
+      "step": 6985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2179329008044952,
+      "learning_rate": 6.090362170286591e-06,
+      "loss": 0.7887,
+      "step": 6986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3130302554206863,
+      "learning_rate": 6.089350605248134e-06,
+      "loss": 1.0785,
+      "step": 6987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.395741348964871,
+      "learning_rate": 6.088338993397857e-06,
+      "loss": 0.8471,
+      "step": 6988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.370610541558175,
+      "learning_rate": 6.0873273347792275e-06,
+      "loss": 1.0036,
+      "step": 6989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2547278365936765,
+      "learning_rate": 6.086315629435722e-06,
+      "loss": 0.7942,
+      "step": 6990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4825599321130696,
+      "learning_rate": 6.085303877410816e-06,
+      "loss": 0.7812,
+      "step": 6991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6782363189315088,
+      "learning_rate": 6.084292078747982e-06,
+      "loss": 0.7513,
+      "step": 6992
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1814576949116633,
+      "learning_rate": 6.0832802334907044e-06,
+      "loss": 0.6504,
+      "step": 6993
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9632936195581026,
+      "learning_rate": 6.0822683416824625e-06,
+      "loss": 0.7055,
+      "step": 6994
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.154376302255512,
+      "learning_rate": 6.08125640336674e-06,
+      "loss": 0.8914,
+      "step": 6995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.742194887427733,
+      "learning_rate": 6.08024441858702e-06,
+      "loss": 0.8561,
+      "step": 6996
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.465501060386607,
+      "learning_rate": 6.079232387386793e-06,
+      "loss": 0.8865,
+      "step": 6997
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0442058668590635,
+      "learning_rate": 6.078220309809546e-06,
+      "loss": 0.7529,
+      "step": 6998
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2350003962163387,
+      "learning_rate": 6.077208185898772e-06,
+      "loss": 0.796,
+      "step": 6999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5765338047175277,
+      "learning_rate": 6.076196015697963e-06,
+      "loss": 0.7482,
+      "step": 7000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0059611351014146,
+      "learning_rate": 6.0751837992506126e-06,
+      "loss": 0.9466,
+      "step": 7001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9312963481288183,
+      "learning_rate": 6.074171536600222e-06,
+      "loss": 0.8331,
+      "step": 7002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2922999092679281,
+      "learning_rate": 6.073159227790287e-06,
+      "loss": 0.6229,
+      "step": 7003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.00780006643362,
+      "learning_rate": 6.07214687286431e-06,
+      "loss": 0.8262,
+      "step": 7004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.163081543966406,
+      "learning_rate": 6.0711344718657934e-06,
+      "loss": 0.8215,
+      "step": 7005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.83639880232239,
+      "learning_rate": 6.070122024838245e-06,
+      "loss": 0.7208,
+      "step": 7006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6742483260014365,
+      "learning_rate": 6.069109531825169e-06,
+      "loss": 0.616,
+      "step": 7007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4817074372944434,
+      "learning_rate": 6.068096992870075e-06,
+      "loss": 0.7593,
+      "step": 7008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1247561734542795,
+      "learning_rate": 6.067084408016475e-06,
+      "loss": 0.839,
+      "step": 7009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.190212887140223,
+      "learning_rate": 6.06607177730788e-06,
+      "loss": 0.8399,
+      "step": 7010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.546266897614058,
+      "learning_rate": 6.065059100787808e-06,
+      "loss": 0.7831,
+      "step": 7011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.120313154316604,
+      "learning_rate": 6.064046378499775e-06,
+      "loss": 0.8541,
+      "step": 7012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9326258290864935,
+      "learning_rate": 6.063033610487298e-06,
+      "loss": 0.882,
+      "step": 7013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.094583353248233,
+      "learning_rate": 6.0620207967939e-06,
+      "loss": 0.811,
+      "step": 7014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2508839544357904,
+      "learning_rate": 6.061007937463101e-06,
+      "loss": 0.6875,
+      "step": 7015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9403177073848714,
+      "learning_rate": 6.05999503253843e-06,
+      "loss": 0.7684,
+      "step": 7016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5848144791525973,
+      "learning_rate": 6.058982082063413e-06,
+      "loss": 1.0797,
+      "step": 7017
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.156008254471925,
+      "learning_rate": 6.057969086081575e-06,
+      "loss": 0.5881,
+      "step": 7018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2236861428493087,
+      "learning_rate": 6.0569560446364495e-06,
+      "loss": 0.8429,
+      "step": 7019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8942923528659537,
+      "learning_rate": 6.05594295777157e-06,
+      "loss": 0.86,
+      "step": 7020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.825078393889379,
+      "learning_rate": 6.054929825530469e-06,
+      "loss": 1.0716,
+      "step": 7021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1087336779835328,
+      "learning_rate": 6.0539166479566844e-06,
+      "loss": 0.656,
+      "step": 7022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7797306469443757,
+      "learning_rate": 6.052903425093752e-06,
+      "loss": 0.7883,
+      "step": 7023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.094316727548101,
+      "learning_rate": 6.051890156985217e-06,
+      "loss": 0.8279,
+      "step": 7024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.279608810951086,
+      "learning_rate": 6.050876843674619e-06,
+      "loss": 0.7532,
+      "step": 7025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8284621027121892,
+      "learning_rate": 6.049863485205503e-06,
+      "loss": 0.5952,
+      "step": 7026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.161137996745258,
+      "learning_rate": 6.048850081621415e-06,
+      "loss": 0.595,
+      "step": 7027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0691939340963734,
+      "learning_rate": 6.047836632965901e-06,
+      "loss": 0.6351,
+      "step": 7028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9322551566959527,
+      "learning_rate": 6.046823139282515e-06,
+      "loss": 0.6908,
+      "step": 7029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9478829087694647,
+      "learning_rate": 6.045809600614807e-06,
+      "loss": 0.8888,
+      "step": 7030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7832304692715961,
+      "learning_rate": 6.044796017006331e-06,
+      "loss": 0.9091,
+      "step": 7031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.357723401966448,
+      "learning_rate": 6.043782388500642e-06,
+      "loss": 0.8312,
+      "step": 7032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9215941702375983,
+      "learning_rate": 6.0427687151413004e-06,
+      "loss": 0.8109,
+      "step": 7033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.037046376479214,
+      "learning_rate": 6.041754996971866e-06,
+      "loss": 0.8515,
+      "step": 7034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.094671812624828,
+      "learning_rate": 6.040741234035898e-06,
+      "loss": 0.9199,
+      "step": 7035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8457303371066538,
+      "learning_rate": 6.039727426376961e-06,
+      "loss": 0.7954,
+      "step": 7036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1947261153582465,
+      "learning_rate": 6.038713574038622e-06,
+      "loss": 0.6175,
+      "step": 7037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.235347567132836,
+      "learning_rate": 6.037699677064446e-06,
+      "loss": 0.8484,
+      "step": 7038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.020700101437654,
+      "learning_rate": 6.036685735498004e-06,
+      "loss": 0.8705,
+      "step": 7039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5545093688271496,
+      "learning_rate": 6.035671749382869e-06,
+      "loss": 0.6948,
+      "step": 7040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4071069687171653,
+      "learning_rate": 6.03465771876261e-06,
+      "loss": 0.7445,
+      "step": 7041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0576289728276307,
+      "learning_rate": 6.0336436436808054e-06,
+      "loss": 0.81,
+      "step": 7042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3386218483948804,
+      "learning_rate": 6.032629524181031e-06,
+      "loss": 0.782,
+      "step": 7043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.349294327549229,
+      "learning_rate": 6.031615360306867e-06,
+      "loss": 0.7358,
+      "step": 7044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1231790095318654,
+      "learning_rate": 6.030601152101892e-06,
+      "loss": 0.7536,
+      "step": 7045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1964025206962994,
+      "learning_rate": 6.029586899609692e-06,
+      "loss": 0.8673,
+      "step": 7046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.252389388609614,
+      "learning_rate": 6.02857260287385e-06,
+      "loss": 0.942,
+      "step": 7047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3521664880460946,
+      "learning_rate": 6.02755826193795e-06,
+      "loss": 0.9054,
+      "step": 7048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.130176433458488,
+      "learning_rate": 6.026543876845586e-06,
+      "loss": 0.7105,
+      "step": 7049
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1315515421579847,
+      "learning_rate": 6.025529447640343e-06,
+      "loss": 0.9394,
+      "step": 7050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.03842754974997,
+      "learning_rate": 6.024514974365818e-06,
+      "loss": 0.7476,
+      "step": 7051
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.003060173293735,
+      "learning_rate": 6.0235004570656005e-06,
+      "loss": 0.8252,
+      "step": 7052
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.054350332248656,
+      "learning_rate": 6.02248589578329e-06,
+      "loss": 0.8302,
+      "step": 7053
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0357206678010007,
+      "learning_rate": 6.021471290562484e-06,
+      "loss": 0.8733,
+      "step": 7054
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9438886456040965,
+      "learning_rate": 6.020456641446781e-06,
+      "loss": 0.8242,
+      "step": 7055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5558573392892288,
+      "learning_rate": 6.019441948479784e-06,
+      "loss": 0.909,
+      "step": 7056
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.499584536696195,
+      "learning_rate": 6.018427211705094e-06,
+      "loss": 0.8358,
+      "step": 7057
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9415080507608131,
+      "learning_rate": 6.017412431166322e-06,
+      "loss": 0.7954,
+      "step": 7058
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.92095476577383,
+      "learning_rate": 6.016397606907069e-06,
+      "loss": 1.0312,
+      "step": 7059
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.891826104473833,
+      "learning_rate": 6.015382738970949e-06,
+      "loss": 0.758,
+      "step": 7060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.225010432121776,
+      "learning_rate": 6.01436782740157e-06,
+      "loss": 1.0478,
+      "step": 7061
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.496821430429964,
+      "learning_rate": 6.013352872242546e-06,
+      "loss": 0.6882,
+      "step": 7062
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.287995486642619,
+      "learning_rate": 6.012337873537494e-06,
+      "loss": 0.8182,
+      "step": 7063
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.554574986540968,
+      "learning_rate": 6.011322831330028e-06,
+      "loss": 0.7203,
+      "step": 7064
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1386542841987033,
+      "learning_rate": 6.010307745663768e-06,
+      "loss": 0.674,
+      "step": 7065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1266434779789753,
+      "learning_rate": 6.009292616582333e-06,
+      "loss": 0.7062,
+      "step": 7066
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.767516198397191,
+      "learning_rate": 6.008277444129347e-06,
+      "loss": 0.986,
+      "step": 7067
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5688584259939904,
+      "learning_rate": 6.007262228348434e-06,
+      "loss": 0.7147,
+      "step": 7068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.107729105151662,
+      "learning_rate": 6.0062469692832205e-06,
+      "loss": 0.7634,
+      "step": 7069
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8812422902259502,
+      "learning_rate": 6.005231666977331e-06,
+      "loss": 0.6541,
+      "step": 7070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.666356717433392,
+      "learning_rate": 6.0042163214744006e-06,
+      "loss": 0.6866,
+      "step": 7071
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9775026651571646,
+      "learning_rate": 6.003200932818058e-06,
+      "loss": 0.6566,
+      "step": 7072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9910878950390785,
+      "learning_rate": 6.002185501051937e-06,
+      "loss": 0.8373,
+      "step": 7073
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1607029927697845,
+      "learning_rate": 6.001170026219673e-06,
+      "loss": 0.6128,
+      "step": 7074
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9563627585154203,
+      "learning_rate": 6.000154508364901e-06,
+      "loss": 0.626,
+      "step": 7075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.452033952718118,
+      "learning_rate": 5.999138947531265e-06,
+      "loss": 0.8356,
+      "step": 7076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.133564049117442,
+      "learning_rate": 5.998123343762403e-06,
+      "loss": 0.6998,
+      "step": 7077
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0520937916478792,
+      "learning_rate": 5.997107697101957e-06,
+      "loss": 0.7088,
+      "step": 7078
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1902771864545594,
+      "learning_rate": 5.996092007593572e-06,
+      "loss": 0.9153,
+      "step": 7079
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1592517746194755,
+      "learning_rate": 5.9950762752808965e-06,
+      "loss": 0.5393,
+      "step": 7080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9142710623589039,
+      "learning_rate": 5.994060500207577e-06,
+      "loss": 0.772,
+      "step": 7081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.924298179870183,
+      "learning_rate": 5.993044682417264e-06,
+      "loss": 0.8306,
+      "step": 7082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3175261610337317,
+      "learning_rate": 5.992028821953609e-06,
+      "loss": 0.7895,
+      "step": 7083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.882206466459726,
+      "learning_rate": 5.9910129188602665e-06,
+      "loss": 0.9942,
+      "step": 7084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.109083592272141,
+      "learning_rate": 5.989996973180892e-06,
+      "loss": 0.9577,
+      "step": 7085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3223176962748187,
+      "learning_rate": 5.988980984959142e-06,
+      "loss": 0.8277,
+      "step": 7086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.037087590495736,
+      "learning_rate": 5.987964954238677e-06,
+      "loss": 0.6777,
+      "step": 7087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1925347782640736,
+      "learning_rate": 5.986948881063157e-06,
+      "loss": 0.6054,
+      "step": 7088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.066862026561889,
+      "learning_rate": 5.985932765476246e-06,
+      "loss": 0.8671,
+      "step": 7089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.301574785081129,
+      "learning_rate": 5.984916607521608e-06,
+      "loss": 0.9568,
+      "step": 7090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7831475462989541,
+      "learning_rate": 5.983900407242911e-06,
+      "loss": 0.9344,
+      "step": 7091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.534451961226748,
+      "learning_rate": 5.982884164683821e-06,
+      "loss": 0.8003,
+      "step": 7092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8984787534036116,
+      "learning_rate": 5.9818678798880085e-06,
+      "loss": 0.9329,
+      "step": 7093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.154701584185381,
+      "learning_rate": 5.9808515528991486e-06,
+      "loss": 0.6982,
+      "step": 7094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.97932292824472,
+      "learning_rate": 5.979835183760911e-06,
+      "loss": 0.9688,
+      "step": 7095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8103498683795667,
+      "learning_rate": 5.978818772516976e-06,
+      "loss": 0.851,
+      "step": 7096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2834629343404937,
+      "learning_rate": 5.977802319211016e-06,
+      "loss": 0.7329,
+      "step": 7097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.254459239652591,
+      "learning_rate": 5.976785823886713e-06,
+      "loss": 0.7623,
+      "step": 7098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8848926823657663,
+      "learning_rate": 5.975769286587747e-06,
+      "loss": 0.697,
+      "step": 7099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0671034997352424,
+      "learning_rate": 5.974752707357805e-06,
+      "loss": 1.0206,
+      "step": 7100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0532518378744724,
+      "learning_rate": 5.973736086240566e-06,
+      "loss": 0.7965,
+      "step": 7101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5728009901437723,
+      "learning_rate": 5.9727194232797185e-06,
+      "loss": 0.8321,
+      "step": 7102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4126202043627147,
+      "learning_rate": 5.97170271851895e-06,
+      "loss": 0.8686,
+      "step": 7103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1655876861416794,
+      "learning_rate": 5.970685972001953e-06,
+      "loss": 0.666,
+      "step": 7104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.612972994648439,
+      "learning_rate": 5.96966918377242e-06,
+      "loss": 0.9177,
+      "step": 7105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9505333946823638,
+      "learning_rate": 5.9686523538740385e-06,
+      "loss": 0.787,
+      "step": 7106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.092797445948717,
+      "learning_rate": 5.96763548235051e-06,
+      "loss": 0.8417,
+      "step": 7107
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5652004878013193,
+      "learning_rate": 5.966618569245529e-06,
+      "loss": 0.8955,
+      "step": 7108
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.032364471788794,
+      "learning_rate": 5.965601614602798e-06,
+      "loss": 0.5512,
+      "step": 7109
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.914892155913523,
+      "learning_rate": 5.964584618466014e-06,
+      "loss": 0.7741,
+      "step": 7110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6294080333312526,
+      "learning_rate": 5.9635675808788785e-06,
+      "loss": 0.839,
+      "step": 7111
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.136031456376496,
+      "learning_rate": 5.9625505018851e-06,
+      "loss": 0.6881,
+      "step": 7112
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8648143837150943,
+      "learning_rate": 5.961533381528382e-06,
+      "loss": 0.8344,
+      "step": 7113
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.105237330807271,
+      "learning_rate": 5.960516219852433e-06,
+      "loss": 0.8094,
+      "step": 7114
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6754793470206337,
+      "learning_rate": 5.959499016900963e-06,
+      "loss": 1.0049,
+      "step": 7115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3580640683711085,
+      "learning_rate": 5.9584817727176836e-06,
+      "loss": 0.6458,
+      "step": 7116
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.202298008068992,
+      "learning_rate": 5.9574644873463075e-06,
+      "loss": 0.9474,
+      "step": 7117
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.201688926059618,
+      "learning_rate": 5.95644716083055e-06,
+      "loss": 0.6458,
+      "step": 7118
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2068591292903923,
+      "learning_rate": 5.955429793214129e-06,
+      "loss": 0.9012,
+      "step": 7119
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.267907940654181,
+      "learning_rate": 5.954412384540761e-06,
+      "loss": 0.6736,
+      "step": 7120
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.823050112706061,
+      "learning_rate": 5.953394934854167e-06,
+      "loss": 0.9488,
+      "step": 7121
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0381906007323995,
+      "learning_rate": 5.952377444198068e-06,
+      "loss": 0.6254,
+      "step": 7122
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.535539973241777,
+      "learning_rate": 5.951359912616192e-06,
+      "loss": 0.6907,
+      "step": 7123
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2541917874267132,
+      "learning_rate": 5.950342340152261e-06,
+      "loss": 0.8838,
+      "step": 7124
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0967757013104507,
+      "learning_rate": 5.949324726850002e-06,
+      "loss": 0.8501,
+      "step": 7125
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.205109044318065,
+      "learning_rate": 5.948307072753146e-06,
+      "loss": 0.934,
+      "step": 7126
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.175162811042204,
+      "learning_rate": 5.947289377905422e-06,
+      "loss": 0.9518,
+      "step": 7127
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.731594337537388,
+      "learning_rate": 5.946271642350565e-06,
+      "loss": 0.9035,
+      "step": 7128
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1414391850037386,
+      "learning_rate": 5.945253866132308e-06,
+      "loss": 0.8317,
+      "step": 7129
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1085941872872087,
+      "learning_rate": 5.9442360492943876e-06,
+      "loss": 0.6886,
+      "step": 7130
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.298064365628049,
+      "learning_rate": 5.94321819188054e-06,
+      "loss": 0.757,
+      "step": 7131
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1234024529776585,
+      "learning_rate": 5.942200293934506e-06,
+      "loss": 0.5852,
+      "step": 7132
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.25524890053776,
+      "learning_rate": 5.941182355500028e-06,
+      "loss": 0.9873,
+      "step": 7133
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1096740966267156,
+      "learning_rate": 5.940164376620847e-06,
+      "loss": 0.8821,
+      "step": 7134
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.782757554962732,
+      "learning_rate": 5.93914635734071e-06,
+      "loss": 0.753,
+      "step": 7135
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5644010174416985,
+      "learning_rate": 5.938128297703361e-06,
+      "loss": 0.613,
+      "step": 7136
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2511348924758674,
+      "learning_rate": 5.937110197752551e-06,
+      "loss": 0.821,
+      "step": 7137
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.260386315556345,
+      "learning_rate": 5.936092057532029e-06,
+      "loss": 0.7884,
+      "step": 7138
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3325863463262815,
+      "learning_rate": 5.935073877085546e-06,
+      "loss": 0.7906,
+      "step": 7139
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5086269653139026,
+      "learning_rate": 5.934055656456855e-06,
+      "loss": 0.8507,
+      "step": 7140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.852128369617982,
+      "learning_rate": 5.9330373956897135e-06,
+      "loss": 0.7532,
+      "step": 7141
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.839458084642207,
+      "learning_rate": 5.932019094827877e-06,
+      "loss": 0.9173,
+      "step": 7142
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5520235939529448,
+      "learning_rate": 5.931000753915104e-06,
+      "loss": 0.8746,
+      "step": 7143
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.903218391591565,
+      "learning_rate": 5.9299823729951544e-06,
+      "loss": 0.8457,
+      "step": 7144
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0861721306871874,
+      "learning_rate": 5.928963952111791e-06,
+      "loss": 0.6639,
+      "step": 7145
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3040510626367294,
+      "learning_rate": 5.9279454913087795e-06,
+      "loss": 0.6598,
+      "step": 7146
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9438353996187705,
+      "learning_rate": 5.926926990629883e-06,
+      "loss": 0.7124,
+      "step": 7147
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.182909648484563,
+      "learning_rate": 5.9259084501188695e-06,
+      "loss": 0.9575,
+      "step": 7148
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9074865485460901,
+      "learning_rate": 5.9248898698195054e-06,
+      "loss": 0.8354,
+      "step": 7149
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4810244932406054,
+      "learning_rate": 5.923871249775566e-06,
+      "loss": 0.7823,
+      "step": 7150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.229826781626989,
+      "learning_rate": 5.922852590030821e-06,
+      "loss": 0.7522,
+      "step": 7151
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.005787886826936,
+      "learning_rate": 5.921833890629045e-06,
+      "loss": 0.9784,
+      "step": 7152
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.105083364915735,
+      "learning_rate": 5.920815151614012e-06,
+      "loss": 0.7613,
+      "step": 7153
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.150740419255821,
+      "learning_rate": 5.919796373029504e-06,
+      "loss": 0.7712,
+      "step": 7154
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.004209419407716,
+      "learning_rate": 5.918777554919296e-06,
+      "loss": 0.5952,
+      "step": 7155
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9772546908659383,
+      "learning_rate": 5.917758697327171e-06,
+      "loss": 0.6576,
+      "step": 7156
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7083200910748984,
+      "learning_rate": 5.9167398002969104e-06,
+      "loss": 0.9101,
+      "step": 7157
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9345475165632073,
+      "learning_rate": 5.915720863872297e-06,
+      "loss": 0.9102,
+      "step": 7158
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3351898137450822,
+      "learning_rate": 5.914701888097121e-06,
+      "loss": 0.6736,
+      "step": 7159
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6710019143504065,
+      "learning_rate": 5.913682873015166e-06,
+      "loss": 0.7812,
+      "step": 7160
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2997154446354218,
+      "learning_rate": 5.912663818670224e-06,
+      "loss": 0.8886,
+      "step": 7161
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4107918412071476,
+      "learning_rate": 5.911644725106085e-06,
+      "loss": 0.7551,
+      "step": 7162
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6005857019639658,
+      "learning_rate": 5.910625592366542e-06,
+      "loss": 0.801,
+      "step": 7163
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8814876943775256,
+      "learning_rate": 5.90960642049539e-06,
+      "loss": 0.8502,
+      "step": 7164
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0621165164392496,
+      "learning_rate": 5.9085872095364236e-06,
+      "loss": 0.9056,
+      "step": 7165
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4418584093903806,
+      "learning_rate": 5.907567959533441e-06,
+      "loss": 0.7757,
+      "step": 7166
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1539120337947253,
+      "learning_rate": 5.906548670530241e-06,
+      "loss": 0.874,
+      "step": 7167
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.291737534805227,
+      "learning_rate": 5.905529342570627e-06,
+      "loss": 0.6917,
+      "step": 7168
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.08153385951621,
+      "learning_rate": 5.904509975698399e-06,
+      "loss": 0.8397,
+      "step": 7169
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7602672574023583,
+      "learning_rate": 5.903490569957366e-06,
+      "loss": 0.7828,
+      "step": 7170
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8496734889669102,
+      "learning_rate": 5.902471125391328e-06,
+      "loss": 0.8708,
+      "step": 7171
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3612836158077979,
+      "learning_rate": 5.901451642044097e-06,
+      "loss": 0.7159,
+      "step": 7172
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1410052150017167,
+      "learning_rate": 5.90043211995948e-06,
+      "loss": 0.6415,
+      "step": 7173
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2047501044874087,
+      "learning_rate": 5.8994125591812914e-06,
+      "loss": 0.9076,
+      "step": 7174
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8690981450923119,
+      "learning_rate": 5.898392959753343e-06,
+      "loss": 0.9698,
+      "step": 7175
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2669112885086524,
+      "learning_rate": 5.897373321719445e-06,
+      "loss": 0.7427,
+      "step": 7176
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8649844610690016,
+      "learning_rate": 5.896353645123418e-06,
+      "loss": 0.8398,
+      "step": 7177
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9211652728507906,
+      "learning_rate": 5.895333930009079e-06,
+      "loss": 0.7033,
+      "step": 7178
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9671323579524636,
+      "learning_rate": 5.894314176420247e-06,
+      "loss": 0.7928,
+      "step": 7179
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.247636593140425,
+      "learning_rate": 5.893294384400743e-06,
+      "loss": 0.9239,
+      "step": 7180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.713968021423351,
+      "learning_rate": 5.89227455399439e-06,
+      "loss": 0.8834,
+      "step": 7181
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9683943043479089,
+      "learning_rate": 5.8912546852450116e-06,
+      "loss": 0.811,
+      "step": 7182
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.256191838538833,
+      "learning_rate": 5.890234778196435e-06,
+      "loss": 0.646,
+      "step": 7183
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1190408333973787,
+      "learning_rate": 5.889214832892489e-06,
+      "loss": 1.0103,
+      "step": 7184
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1008782973770388,
+      "learning_rate": 5.888194849376998e-06,
+      "loss": 0.6985,
+      "step": 7185
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.119773045363618,
+      "learning_rate": 5.8871748276937975e-06,
+      "loss": 0.6274,
+      "step": 7186
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1711869752799646,
+      "learning_rate": 5.886154767886719e-06,
+      "loss": 0.7007,
+      "step": 7187
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4717123795045475,
+      "learning_rate": 5.885134669999597e-06,
+      "loss": 0.7344,
+      "step": 7188
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.155493613161562,
+      "learning_rate": 5.8841145340762665e-06,
+      "loss": 0.8202,
+      "step": 7189
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.238432392954669,
+      "learning_rate": 5.8830943601605665e-06,
+      "loss": 0.8281,
+      "step": 7190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.080004275763364,
+      "learning_rate": 5.882074148296332e-06,
+      "loss": 0.6902,
+      "step": 7191
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8660654084087773,
+      "learning_rate": 5.881053898527409e-06,
+      "loss": 0.7652,
+      "step": 7192
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.836217409633428,
+      "learning_rate": 5.880033610897638e-06,
+      "loss": 0.7415,
+      "step": 7193
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9203817231788525,
+      "learning_rate": 5.879013285450863e-06,
+      "loss": 0.7879,
+      "step": 7194
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1416074874680127,
+      "learning_rate": 5.877992922230928e-06,
+      "loss": 0.732,
+      "step": 7195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.632300533511952,
+      "learning_rate": 5.876972521281683e-06,
+      "loss": 0.8008,
+      "step": 7196
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2398954900481596,
+      "learning_rate": 5.875952082646975e-06,
+      "loss": 0.8994,
+      "step": 7197
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9950116432331575,
+      "learning_rate": 5.8749316063706565e-06,
+      "loss": 0.7769,
+      "step": 7198
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6150670986883575,
+      "learning_rate": 5.873911092496577e-06,
+      "loss": 0.7018,
+      "step": 7199
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5931210105517133,
+      "learning_rate": 5.872890541068591e-06,
+      "loss": 0.7192,
+      "step": 7200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7437690774361883,
+      "learning_rate": 5.871869952130556e-06,
+      "loss": 0.7553,
+      "step": 7201
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.92456821569478,
+      "learning_rate": 5.870849325726326e-06,
+      "loss": 0.9159,
+      "step": 7202
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.560605076869119,
+      "learning_rate": 5.869828661899761e-06,
+      "loss": 0.8328,
+      "step": 7203
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4239839758741315,
+      "learning_rate": 5.8688079606947226e-06,
+      "loss": 0.8305,
+      "step": 7204
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3497952330374066,
+      "learning_rate": 5.86778722215507e-06,
+      "loss": 0.8079,
+      "step": 7205
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.686028887008955,
+      "learning_rate": 5.866766446324668e-06,
+      "loss": 0.801,
+      "step": 7206
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.048213748575479,
+      "learning_rate": 5.865745633247381e-06,
+      "loss": 0.8931,
+      "step": 7207
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9302642383818667,
+      "learning_rate": 5.864724782967077e-06,
+      "loss": 0.9498,
+      "step": 7208
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5524662680474206,
+      "learning_rate": 5.8637038955276225e-06,
+      "loss": 0.7585,
+      "step": 7209
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2705953182358125,
+      "learning_rate": 5.862682970972888e-06,
+      "loss": 0.7219,
+      "step": 7210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0406513918314833,
+      "learning_rate": 5.861662009346747e-06,
+      "loss": 0.8799,
+      "step": 7211
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1385368777818659,
+      "learning_rate": 5.8606410106930675e-06,
+      "loss": 0.6703,
+      "step": 7212
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1514204980140097,
+      "learning_rate": 5.85961997505573e-06,
+      "loss": 0.7434,
+      "step": 7213
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.437241574630029,
+      "learning_rate": 5.858598902478604e-06,
+      "loss": 0.8549,
+      "step": 7214
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0529933032541376,
+      "learning_rate": 5.857577793005573e-06,
+      "loss": 0.7493,
+      "step": 7215
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.781163973874967,
+      "learning_rate": 5.856556646680514e-06,
+      "loss": 0.6905,
+      "step": 7216
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8903645885945966,
+      "learning_rate": 5.855535463547309e-06,
+      "loss": 0.7986,
+      "step": 7217
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.259623306109014,
+      "learning_rate": 5.854514243649839e-06,
+      "loss": 0.784,
+      "step": 7218
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4352546095089167,
+      "learning_rate": 5.853492987031989e-06,
+      "loss": 0.9223,
+      "step": 7219
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1751776124677282,
+      "learning_rate": 5.8524716937376446e-06,
+      "loss": 0.7588,
+      "step": 7220
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6925592349441547,
+      "learning_rate": 5.851450363810694e-06,
+      "loss": 0.8248,
+      "step": 7221
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0579224633242035,
+      "learning_rate": 5.850428997295023e-06,
+      "loss": 0.7074,
+      "step": 7222
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3760811013477556,
+      "learning_rate": 5.8494075942345244e-06,
+      "loss": 0.6242,
+      "step": 7223
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1280523157530673,
+      "learning_rate": 5.8483861546730915e-06,
+      "loss": 0.6803,
+      "step": 7224
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0348469017247384,
+      "learning_rate": 5.8473646786546134e-06,
+      "loss": 0.8384,
+      "step": 7225
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.225078647192402,
+      "learning_rate": 5.846343166222991e-06,
+      "loss": 0.8734,
+      "step": 7226
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7437651680025295,
+      "learning_rate": 5.845321617422115e-06,
+      "loss": 0.9979,
+      "step": 7227
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.101421827716693,
+      "learning_rate": 5.8443000322958875e-06,
+      "loss": 0.6875,
+      "step": 7228
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1174718191712207,
+      "learning_rate": 5.843278410888208e-06,
+      "loss": 0.8877,
+      "step": 7229
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1264066130725294,
+      "learning_rate": 5.842256753242975e-06,
+      "loss": 0.7396,
+      "step": 7230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.118740615502813,
+      "learning_rate": 5.841235059404097e-06,
+      "loss": 0.9698,
+      "step": 7231
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.071443101545244,
+      "learning_rate": 5.840213329415471e-06,
+      "loss": 0.6604,
+      "step": 7232
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1879934497598645,
+      "learning_rate": 5.839191563321009e-06,
+      "loss": 0.9473,
+      "step": 7233
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2016229700738243,
+      "learning_rate": 5.838169761164616e-06,
+      "loss": 0.64,
+      "step": 7234
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.301252081297761,
+      "learning_rate": 5.837147922990204e-06,
+      "loss": 0.7524,
+      "step": 7235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1335619362261378,
+      "learning_rate": 5.836126048841678e-06,
+      "loss": 0.8051,
+      "step": 7236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.945554220002482,
+      "learning_rate": 5.8351041387629546e-06,
+      "loss": 0.8044,
+      "step": 7237
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.11410636023722,
+      "learning_rate": 5.834082192797948e-06,
+      "loss": 0.7936,
+      "step": 7238
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2232029708933225,
+      "learning_rate": 5.83306021099057e-06,
+      "loss": 0.6938,
+      "step": 7239
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.259601503085516,
+      "learning_rate": 5.832038193384741e-06,
+      "loss": 0.5972,
+      "step": 7240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.576695185337048,
+      "learning_rate": 5.831016140024377e-06,
+      "loss": 0.9733,
+      "step": 7241
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0126162497443008,
+      "learning_rate": 5.829994050953398e-06,
+      "loss": 0.6428,
+      "step": 7242
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8833895460275982,
+      "learning_rate": 5.828971926215726e-06,
+      "loss": 0.6787,
+      "step": 7243
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.630127698341947,
+      "learning_rate": 5.827949765855285e-06,
+      "loss": 0.8497,
+      "step": 7244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0311469132051068,
+      "learning_rate": 5.826927569915999e-06,
+      "loss": 0.7126,
+      "step": 7245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.515886685457583,
+      "learning_rate": 5.825905338441794e-06,
+      "loss": 0.7921,
+      "step": 7246
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9065540925036013,
+      "learning_rate": 5.824883071476597e-06,
+      "loss": 0.9045,
+      "step": 7247
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7433617950754263,
+      "learning_rate": 5.823860769064335e-06,
+      "loss": 0.8534,
+      "step": 7248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2753971166548004,
+      "learning_rate": 5.822838431248943e-06,
+      "loss": 0.8991,
+      "step": 7249
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8220136241082812,
+      "learning_rate": 5.821816058074351e-06,
+      "loss": 0.7982,
+      "step": 7250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.533370228118074,
+      "learning_rate": 5.820793649584493e-06,
+      "loss": 0.8251,
+      "step": 7251
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0822505858565816,
+      "learning_rate": 5.819771205823303e-06,
+      "loss": 0.6472,
+      "step": 7252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6796142869495636,
+      "learning_rate": 5.818748726834718e-06,
+      "loss": 0.8287,
+      "step": 7253
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2366330601507336,
+      "learning_rate": 5.817726212662678e-06,
+      "loss": 0.9235,
+      "step": 7254
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0127912871673272,
+      "learning_rate": 5.8167036633511206e-06,
+      "loss": 0.7353,
+      "step": 7255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2773497800586138,
+      "learning_rate": 5.815681078943989e-06,
+      "loss": 0.7034,
+      "step": 7256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9206222794875283,
+      "learning_rate": 5.814658459485223e-06,
+      "loss": 0.6748,
+      "step": 7257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6480983299538843,
+      "learning_rate": 5.81363580501877e-06,
+      "loss": 1.0091,
+      "step": 7258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.059899021922197,
+      "learning_rate": 5.812613115588575e-06,
+      "loss": 0.6712,
+      "step": 7259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7407735668881539,
+      "learning_rate": 5.811590391238584e-06,
+      "loss": 0.7883,
+      "step": 7260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.698054009903055,
+      "learning_rate": 5.8105676320127455e-06,
+      "loss": 0.746,
+      "step": 7261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2320165327296513,
+      "learning_rate": 5.809544837955011e-06,
+      "loss": 0.6934,
+      "step": 7262
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.190833575014655,
+      "learning_rate": 5.808522009109333e-06,
+      "loss": 0.8786,
+      "step": 7263
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.127347746091317,
+      "learning_rate": 5.807499145519663e-06,
+      "loss": 1.0273,
+      "step": 7264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1876444199231915,
+      "learning_rate": 5.806476247229956e-06,
+      "loss": 0.9346,
+      "step": 7265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.24460600526845,
+      "learning_rate": 5.805453314284168e-06,
+      "loss": 0.8154,
+      "step": 7266
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9378111865562744,
+      "learning_rate": 5.804430346726258e-06,
+      "loss": 0.9671,
+      "step": 7267
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1525526707344178,
+      "learning_rate": 5.803407344600185e-06,
+      "loss": 0.6937,
+      "step": 7268
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8273494585527448,
+      "learning_rate": 5.802384307949909e-06,
+      "loss": 0.7909,
+      "step": 7269
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0895249012338968,
+      "learning_rate": 5.8013612368193905e-06,
+      "loss": 0.7061,
+      "step": 7270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.159214279606415,
+      "learning_rate": 5.800338131252597e-06,
+      "loss": 0.7773,
+      "step": 7271
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.067895101045239,
+      "learning_rate": 5.79931499129349e-06,
+      "loss": 0.7684,
+      "step": 7272
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8983805733230628,
+      "learning_rate": 5.7982918169860395e-06,
+      "loss": 0.8527,
+      "step": 7273
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3830727539642784,
+      "learning_rate": 5.79726860837421e-06,
+      "loss": 0.934,
+      "step": 7274
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3188954539373623,
+      "learning_rate": 5.796245365501973e-06,
+      "loss": 0.7095,
+      "step": 7275
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4752981400410694,
+      "learning_rate": 5.795222088413299e-06,
+      "loss": 0.8897,
+      "step": 7276
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1855846731620816,
+      "learning_rate": 5.794198777152162e-06,
+      "loss": 0.8087,
+      "step": 7277
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1518357934543952,
+      "learning_rate": 5.793175431762534e-06,
+      "loss": 0.8698,
+      "step": 7278
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4107111450765206,
+      "learning_rate": 5.792152052288391e-06,
+      "loss": 0.8837,
+      "step": 7279
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3641465517554936,
+      "learning_rate": 5.791128638773711e-06,
+      "loss": 0.684,
+      "step": 7280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1336930442024493,
+      "learning_rate": 5.790105191262471e-06,
+      "loss": 0.6027,
+      "step": 7281
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9162098282028466,
+      "learning_rate": 5.789081709798652e-06,
+      "loss": 0.8164,
+      "step": 7282
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.123860241673635,
+      "learning_rate": 5.788058194426234e-06,
+      "loss": 0.9024,
+      "step": 7283
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.249512922554543,
+      "learning_rate": 5.787034645189199e-06,
+      "loss": 0.7318,
+      "step": 7284
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9121984396756935,
+      "learning_rate": 5.786011062131535e-06,
+      "loss": 0.8794,
+      "step": 7285
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.064485411427857,
+      "learning_rate": 5.784987445297222e-06,
+      "loss": 0.915,
+      "step": 7286
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2095591704755186,
+      "learning_rate": 5.783963794730254e-06,
+      "loss": 0.9244,
+      "step": 7287
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1710892554748895,
+      "learning_rate": 5.782940110474612e-06,
+      "loss": 0.8297,
+      "step": 7288
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1114142233599584,
+      "learning_rate": 5.7819163925742915e-06,
+      "loss": 0.6146,
+      "step": 7289
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6968424998864466,
+      "learning_rate": 5.780892641073281e-06,
+      "loss": 0.9167,
+      "step": 7290
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5683043076584258,
+      "learning_rate": 5.779868856015578e-06,
+      "loss": 0.8659,
+      "step": 7291
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7740572819762312,
+      "learning_rate": 5.778845037445171e-06,
+      "loss": 0.8877,
+      "step": 7292
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1576452374441173,
+      "learning_rate": 5.7778211854060575e-06,
+      "loss": 0.8387,
+      "step": 7293
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.005132528618474,
+      "learning_rate": 5.776797299942236e-06,
+      "loss": 0.73,
+      "step": 7294
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1968103416754836,
+      "learning_rate": 5.775773381097705e-06,
+      "loss": 1.0478,
+      "step": 7295
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3120393737175875,
+      "learning_rate": 5.774749428916464e-06,
+      "loss": 0.589,
+      "step": 7296
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9816241811987241,
+      "learning_rate": 5.773725443442514e-06,
+      "loss": 0.702,
+      "step": 7297
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1401731396644346,
+      "learning_rate": 5.7727014247198585e-06,
+      "loss": 0.8735,
+      "step": 7298
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.204828565969872,
+      "learning_rate": 5.771677372792502e-06,
+      "loss": 0.7892,
+      "step": 7299
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4227970783130597,
+      "learning_rate": 5.770653287704451e-06,
+      "loss": 0.8283,
+      "step": 7300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8898667523566899,
+      "learning_rate": 5.7696291694997105e-06,
+      "loss": 0.9229,
+      "step": 7301
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6681393399878957,
+      "learning_rate": 5.768605018222292e-06,
+      "loss": 0.8533,
+      "step": 7302
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.805262265789683,
+      "learning_rate": 5.767580833916203e-06,
+      "loss": 0.8292,
+      "step": 7303
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0991566643747266,
+      "learning_rate": 5.766556616625456e-06,
+      "loss": 0.9924,
+      "step": 7304
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9297857293052325,
+      "learning_rate": 5.765532366394064e-06,
+      "loss": 0.7632,
+      "step": 7305
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3101486063368344,
+      "learning_rate": 5.764508083266043e-06,
+      "loss": 0.666,
+      "step": 7306
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.377308704914358,
+      "learning_rate": 5.763483767285405e-06,
+      "loss": 0.7751,
+      "step": 7307
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.417295885240577,
+      "learning_rate": 5.762459418496169e-06,
+      "loss": 0.8824,
+      "step": 7308
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2647804164728877,
+      "learning_rate": 5.7614350369423555e-06,
+      "loss": 0.7942,
+      "step": 7309
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0576872343265786,
+      "learning_rate": 5.760410622667981e-06,
+      "loss": 0.9154,
+      "step": 7310
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.300050398285697,
+      "learning_rate": 5.75938617571707e-06,
+      "loss": 0.9143,
+      "step": 7311
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1287343519882884,
+      "learning_rate": 5.758361696133643e-06,
+      "loss": 0.8821,
+      "step": 7312
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9951936960474987,
+      "learning_rate": 5.757337183961724e-06,
+      "loss": 0.8443,
+      "step": 7313
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.176325931086329,
+      "learning_rate": 5.7563126392453415e-06,
+      "loss": 0.7075,
+      "step": 7314
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1648368244002434,
+      "learning_rate": 5.755288062028519e-06,
+      "loss": 0.8329,
+      "step": 7315
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9727161631674388,
+      "learning_rate": 5.754263452355287e-06,
+      "loss": 0.7724,
+      "step": 7316
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2931164759563174,
+      "learning_rate": 5.753238810269674e-06,
+      "loss": 0.7938,
+      "step": 7317
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5530989922677136,
+      "learning_rate": 5.752214135815712e-06,
+      "loss": 0.7396,
+      "step": 7318
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1027529888652787,
+      "learning_rate": 5.751189429037435e-06,
+      "loss": 0.8114,
+      "step": 7319
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7481772001994709,
+      "learning_rate": 5.750164689978873e-06,
+      "loss": 0.8835,
+      "step": 7320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3917535574377657,
+      "learning_rate": 5.7491399186840646e-06,
+      "loss": 0.8695,
+      "step": 7321
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8505295800505879,
+      "learning_rate": 5.748115115197045e-06,
+      "loss": 0.7546,
+      "step": 7322
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9242242178494915,
+      "learning_rate": 5.7470902795618535e-06,
+      "loss": 0.8066,
+      "step": 7323
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6637928524688546,
+      "learning_rate": 5.746065411822528e-06,
+      "loss": 0.914,
+      "step": 7324
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4353016826770357,
+      "learning_rate": 5.745040512023111e-06,
+      "loss": 0.7835,
+      "step": 7325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.413335599542013,
+      "learning_rate": 5.744015580207642e-06,
+      "loss": 0.9607,
+      "step": 7326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8373189414188813,
+      "learning_rate": 5.742990616420169e-06,
+      "loss": 0.7417,
+      "step": 7327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1739313855061475,
+      "learning_rate": 5.741965620704734e-06,
+      "loss": 0.7353,
+      "step": 7328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9960694594195644,
+      "learning_rate": 5.740940593105383e-06,
+      "loss": 0.605,
+      "step": 7329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3363098795962642,
+      "learning_rate": 5.7399155336661645e-06,
+      "loss": 0.7196,
+      "step": 7330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0225798550231615,
+      "learning_rate": 5.738890442431128e-06,
+      "loss": 0.9058,
+      "step": 7331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2634536653328925,
+      "learning_rate": 5.737865319444324e-06,
+      "loss": 0.9586,
+      "step": 7332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7967508862882138,
+      "learning_rate": 5.736840164749803e-06,
+      "loss": 0.8119,
+      "step": 7333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8896482469790925,
+      "learning_rate": 5.73581497839162e-06,
+      "loss": 0.7135,
+      "step": 7334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8771056798424273,
+      "learning_rate": 5.734789760413827e-06,
+      "loss": 0.717,
+      "step": 7335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0214761748849925,
+      "learning_rate": 5.733764510860482e-06,
+      "loss": 0.7204,
+      "step": 7336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.241060929080477,
+      "learning_rate": 5.7327392297756426e-06,
+      "loss": 0.7472,
+      "step": 7337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9187593025852571,
+      "learning_rate": 5.731713917203368e-06,
+      "loss": 0.5704,
+      "step": 7338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.235003277815209,
+      "learning_rate": 5.730688573187715e-06,
+      "loss": 0.8379,
+      "step": 7339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2450335246114137,
+      "learning_rate": 5.729663197772745e-06,
+      "loss": 0.8498,
+      "step": 7340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2929656005208918,
+      "learning_rate": 5.728637791002525e-06,
+      "loss": 0.6907,
+      "step": 7341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.434282614972544,
+      "learning_rate": 5.727612352921116e-06,
+      "loss": 0.9466,
+      "step": 7342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9846163652389901,
+      "learning_rate": 5.726586883572584e-06,
+      "loss": 0.6255,
+      "step": 7343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.447426236301733,
+      "learning_rate": 5.725561383000994e-06,
+      "loss": 0.8456,
+      "step": 7344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.156323146796215,
+      "learning_rate": 5.724535851250416e-06,
+      "loss": 0.6881,
+      "step": 7345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1087304583247626,
+      "learning_rate": 5.723510288364918e-06,
+      "loss": 0.7013,
+      "step": 7346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3256792607899808,
+      "learning_rate": 5.722484694388573e-06,
+      "loss": 0.8691,
+      "step": 7347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.404039699597423,
+      "learning_rate": 5.721459069365452e-06,
+      "loss": 0.6682,
+      "step": 7348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0626365659555799,
+      "learning_rate": 5.720433413339627e-06,
+      "loss": 0.7208,
+      "step": 7349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.703727715658115,
+      "learning_rate": 5.719407726355174e-06,
+      "loss": 0.9587,
+      "step": 7350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6034522368000634,
+      "learning_rate": 5.718382008456168e-06,
+      "loss": 0.8259,
+      "step": 7351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.98957779089815,
+      "learning_rate": 5.717356259686689e-06,
+      "loss": 0.7211,
+      "step": 7352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.064370485067286,
+      "learning_rate": 5.716330480090812e-06,
+      "loss": 0.6593,
+      "step": 7353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.079215775309627,
+      "learning_rate": 5.71530466971262e-06,
+      "loss": 0.8024,
+      "step": 7354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4138132796154004,
+      "learning_rate": 5.714278828596192e-06,
+      "loss": 0.8174,
+      "step": 7355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5436289253110174,
+      "learning_rate": 5.713252956785613e-06,
+      "loss": 0.7214,
+      "step": 7356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2674181753910396,
+      "learning_rate": 5.712227054324968e-06,
+      "loss": 0.8787,
+      "step": 7357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.342329291018848,
+      "learning_rate": 5.711201121258337e-06,
+      "loss": 0.7487,
+      "step": 7358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.397909282240669,
+      "learning_rate": 5.710175157629812e-06,
+      "loss": 0.9369,
+      "step": 7359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7536782177537962,
+      "learning_rate": 5.709149163483476e-06,
+      "loss": 0.8587,
+      "step": 7360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.314686794492153,
+      "learning_rate": 5.708123138863424e-06,
+      "loss": 0.7764,
+      "step": 7361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0012505817667057,
+      "learning_rate": 5.707097083813743e-06,
+      "loss": 0.9013,
+      "step": 7362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6232811185964504,
+      "learning_rate": 5.706070998378524e-06,
+      "loss": 0.7317,
+      "step": 7363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.650609187277256,
+      "learning_rate": 5.705044882601862e-06,
+      "loss": 1.0188,
+      "step": 7364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9755444189853686,
+      "learning_rate": 5.704018736527853e-06,
+      "loss": 0.9427,
+      "step": 7365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9526955350993984,
+      "learning_rate": 5.7029925602005905e-06,
+      "loss": 0.7391,
+      "step": 7366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9732013351581188,
+      "learning_rate": 5.701966353664171e-06,
+      "loss": 0.7657,
+      "step": 7367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9336985188292871,
+      "learning_rate": 5.7009401169626955e-06,
+      "loss": 0.7212,
+      "step": 7368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1466189716078368,
+      "learning_rate": 5.69991385014026e-06,
+      "loss": 0.7032,
+      "step": 7369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0138588906871662,
+      "learning_rate": 5.698887553240968e-06,
+      "loss": 0.756,
+      "step": 7370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.850635280111144,
+      "learning_rate": 5.697861226308923e-06,
+      "loss": 0.662,
+      "step": 7371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9468187935042105,
+      "learning_rate": 5.696834869388226e-06,
+      "loss": 0.5517,
+      "step": 7372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.289866752472956,
+      "learning_rate": 5.695808482522982e-06,
+      "loss": 0.8581,
+      "step": 7373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.297593459596125,
+      "learning_rate": 5.694782065757298e-06,
+      "loss": 0.8098,
+      "step": 7374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.193660359541352,
+      "learning_rate": 5.693755619135282e-06,
+      "loss": 0.9897,
+      "step": 7375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9663312483182203,
+      "learning_rate": 5.692729142701042e-06,
+      "loss": 0.7716,
+      "step": 7376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1167058223248594,
+      "learning_rate": 5.691702636498688e-06,
+      "loss": 0.9321,
+      "step": 7377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.050693274869216,
+      "learning_rate": 5.69067610057233e-06,
+      "loss": 0.8787,
+      "step": 7378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0274249391724062,
+      "learning_rate": 5.689649534966083e-06,
+      "loss": 0.6516,
+      "step": 7379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2324770232199094,
+      "learning_rate": 5.6886229397240585e-06,
+      "loss": 0.6976,
+      "step": 7380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.108689779505813,
+      "learning_rate": 5.6875963148903735e-06,
+      "loss": 0.7253,
+      "step": 7381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2552222795076378,
+      "learning_rate": 5.6865696605091426e-06,
+      "loss": 0.917,
+      "step": 7382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1744735450879453,
+      "learning_rate": 5.685542976624485e-06,
+      "loss": 0.8177,
+      "step": 7383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3775274411761638,
+      "learning_rate": 5.684516263280519e-06,
+      "loss": 0.9121,
+      "step": 7384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.928149895266117,
+      "learning_rate": 5.683489520521365e-06,
+      "loss": 0.7144,
+      "step": 7385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1876199742704783,
+      "learning_rate": 5.682462748391144e-06,
+      "loss": 0.8597,
+      "step": 7386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9589588447646216,
+      "learning_rate": 5.681435946933978e-06,
+      "loss": 0.9383,
+      "step": 7387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2272501221424674,
+      "learning_rate": 5.680409116193991e-06,
+      "loss": 0.8985,
+      "step": 7388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0958532659943392,
+      "learning_rate": 5.679382256215311e-06,
+      "loss": 0.6371,
+      "step": 7389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.112888165950286,
+      "learning_rate": 5.678355367042061e-06,
+      "loss": 0.8954,
+      "step": 7390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5851494115479783,
+      "learning_rate": 5.67732844871837e-06,
+      "loss": 0.9379,
+      "step": 7391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9129071008595466,
+      "learning_rate": 5.6763015012883686e-06,
+      "loss": 0.717,
+      "step": 7392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.639734462552546,
+      "learning_rate": 5.675274524796184e-06,
+      "loss": 0.7638,
+      "step": 7393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.888654876580885,
+      "learning_rate": 5.674247519285951e-06,
+      "loss": 0.7919,
+      "step": 7394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.085155878854846,
+      "learning_rate": 5.673220484801798e-06,
+      "loss": 0.8645,
+      "step": 7395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1903766105941505,
+      "learning_rate": 5.672193421387862e-06,
+      "loss": 0.7704,
+      "step": 7396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7786681525142334,
+      "learning_rate": 5.671166329088278e-06,
+      "loss": 0.7823,
+      "step": 7397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.83333806377774,
+      "learning_rate": 5.6701392079471825e-06,
+      "loss": 0.7918,
+      "step": 7398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9506842356158265,
+      "learning_rate": 5.6691120580087126e-06,
+      "loss": 0.8723,
+      "step": 7399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1538766609528808,
+      "learning_rate": 5.668084879317006e-06,
+      "loss": 0.9073,
+      "step": 7400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1422276855767586,
+      "learning_rate": 5.6670576719162065e-06,
+      "loss": 0.8732,
+      "step": 7401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.021700351236277,
+      "learning_rate": 5.6660304358504524e-06,
+      "loss": 0.7076,
+      "step": 7402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.012632590686403,
+      "learning_rate": 5.665003171163886e-06,
+      "loss": 0.6111,
+      "step": 7403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.13315984948944,
+      "learning_rate": 5.6639758779006535e-06,
+      "loss": 0.8443,
+      "step": 7404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2396423000651615,
+      "learning_rate": 5.662948556104898e-06,
+      "loss": 0.8595,
+      "step": 7405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1678873051911047,
+      "learning_rate": 5.661921205820767e-06,
+      "loss": 0.8406,
+      "step": 7406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.183217750124266,
+      "learning_rate": 5.660893827092407e-06,
+      "loss": 0.7751,
+      "step": 7407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.253695007259191,
+      "learning_rate": 5.659866419963969e-06,
+      "loss": 0.9643,
+      "step": 7408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2292250543761947,
+      "learning_rate": 5.6588389844796e-06,
+      "loss": 0.7412,
+      "step": 7409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.060391642042955,
+      "learning_rate": 5.657811520683454e-06,
+      "loss": 0.8695,
+      "step": 7410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.430176745229646,
+      "learning_rate": 5.656784028619681e-06,
+      "loss": 1.0165,
+      "step": 7411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1873443500952092,
+      "learning_rate": 5.655756508332434e-06,
+      "loss": 0.7664,
+      "step": 7412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1714813491626086,
+      "learning_rate": 5.654728959865872e-06,
+      "loss": 0.7932,
+      "step": 7413
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.116917276965397,
+      "learning_rate": 5.653701383264147e-06,
+      "loss": 0.8148,
+      "step": 7414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1107603118685496,
+      "learning_rate": 5.652673778571418e-06,
+      "loss": 0.6688,
+      "step": 7415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.7863142200396,
+      "learning_rate": 5.651646145831842e-06,
+      "loss": 0.7682,
+      "step": 7416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.357159439927481,
+      "learning_rate": 5.650618485089582e-06,
+      "loss": 0.9544,
+      "step": 7417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.686855362849617,
+      "learning_rate": 5.649590796388794e-06,
+      "loss": 0.8399,
+      "step": 7418
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0664565714564178,
+      "learning_rate": 5.648563079773646e-06,
+      "loss": 0.6363,
+      "step": 7419
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3487001782575647,
+      "learning_rate": 5.647535335288296e-06,
+      "loss": 0.9373,
+      "step": 7420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.889323452403397,
+      "learning_rate": 5.6465075629769095e-06,
+      "loss": 0.8943,
+      "step": 7421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8067951201354204,
+      "learning_rate": 5.645479762883657e-06,
+      "loss": 0.8933,
+      "step": 7422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2761515129519068,
+      "learning_rate": 5.644451935052697e-06,
+      "loss": 0.9897,
+      "step": 7423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4864491741718893,
+      "learning_rate": 5.6434240795282045e-06,
+      "loss": 0.9656,
+      "step": 7424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1960445635118755,
+      "learning_rate": 5.642396196354345e-06,
+      "loss": 0.7143,
+      "step": 7425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.880220152799027,
+      "learning_rate": 5.641368285575292e-06,
+      "loss": 0.7676,
+      "step": 7426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2277177164502764,
+      "learning_rate": 5.640340347235215e-06,
+      "loss": 0.6854,
+      "step": 7427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.059333669467703,
+      "learning_rate": 5.639312381378286e-06,
+      "loss": 1.0767,
+      "step": 7428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0187653525762097,
+      "learning_rate": 5.63828438804868e-06,
+      "loss": 0.7263,
+      "step": 7429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9760425170835323,
+      "learning_rate": 5.637256367290573e-06,
+      "loss": 0.974,
+      "step": 7430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.27589614346359,
+      "learning_rate": 5.636228319148141e-06,
+      "loss": 0.7211,
+      "step": 7431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8964496255994623,
+      "learning_rate": 5.635200243665562e-06,
+      "loss": 0.8077,
+      "step": 7432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.958096534512866,
+      "learning_rate": 5.634172140887013e-06,
+      "loss": 0.8023,
+      "step": 7433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.671275498953294,
+      "learning_rate": 5.6331440108566735e-06,
+      "loss": 0.9865,
+      "step": 7434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6816325535067285,
+      "learning_rate": 5.632115853618727e-06,
+      "loss": 0.7943,
+      "step": 7435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7867778343066907,
+      "learning_rate": 5.631087669217356e-06,
+      "loss": 0.8978,
+      "step": 7436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8658073965716413,
+      "learning_rate": 5.630059457696741e-06,
+      "loss": 0.8703,
+      "step": 7437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.83268158766755,
+      "learning_rate": 5.629031219101065e-06,
+      "loss": 0.785,
+      "step": 7438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5885235511072664,
+      "learning_rate": 5.628002953474521e-06,
+      "loss": 0.8408,
+      "step": 7439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0360253283778635,
+      "learning_rate": 5.626974660861288e-06,
+      "loss": 0.9325,
+      "step": 7440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0393177070997526,
+      "learning_rate": 5.6259463413055604e-06,
+      "loss": 0.742,
+      "step": 7441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.48358060298831,
+      "learning_rate": 5.624917994851523e-06,
+      "loss": 0.8113,
+      "step": 7442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3441965832620393,
+      "learning_rate": 5.6238896215433666e-06,
+      "loss": 0.8456,
+      "step": 7443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.563773641777694,
+      "learning_rate": 5.622861221425286e-06,
+      "loss": 0.9316,
+      "step": 7444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7219296102360206,
+      "learning_rate": 5.6218327945414695e-06,
+      "loss": 0.7449,
+      "step": 7445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1409390111918225,
+      "learning_rate": 5.620804340936114e-06,
+      "loss": 0.6224,
+      "step": 7446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0008944589774442,
+      "learning_rate": 5.619775860653412e-06,
+      "loss": 0.8024,
+      "step": 7447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1602476581530166,
+      "learning_rate": 5.6187473537375635e-06,
+      "loss": 0.8968,
+      "step": 7448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0556596879232316,
+      "learning_rate": 5.617718820232762e-06,
+      "loss": 0.8984,
+      "step": 7449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1544128448687023,
+      "learning_rate": 5.616690260183206e-06,
+      "loss": 0.9432,
+      "step": 7450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5831970979048493,
+      "learning_rate": 5.615661673633098e-06,
+      "loss": 0.8544,
+      "step": 7451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7543524945575693,
+      "learning_rate": 5.614633060626634e-06,
+      "loss": 0.7482,
+      "step": 7452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.042149243600797,
+      "learning_rate": 5.613604421208021e-06,
+      "loss": 1.0198,
+      "step": 7453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.090906096648803,
+      "learning_rate": 5.612575755421459e-06,
+      "loss": 0.6328,
+      "step": 7454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.262037181773028,
+      "learning_rate": 5.611547063311152e-06,
+      "loss": 1.065,
+      "step": 7455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8030405982011768,
+      "learning_rate": 5.610518344921305e-06,
+      "loss": 0.9318,
+      "step": 7456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.208575829841909,
+      "learning_rate": 5.609489600296127e-06,
+      "loss": 0.814,
+      "step": 7457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2533983211638398,
+      "learning_rate": 5.608460829479821e-06,
+      "loss": 0.9046,
+      "step": 7458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8178795419769094,
+      "learning_rate": 5.607432032516601e-06,
+      "loss": 0.7465,
+      "step": 7459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.288836263706184,
+      "learning_rate": 5.606403209450672e-06,
+      "loss": 0.9328,
+      "step": 7460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3476664447596254,
+      "learning_rate": 5.605374360326246e-06,
+      "loss": 0.6752,
+      "step": 7461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.20888977279337,
+      "learning_rate": 5.604345485187535e-06,
+      "loss": 0.9516,
+      "step": 7462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.08000331329736,
+      "learning_rate": 5.603316584078754e-06,
+      "loss": 0.876,
+      "step": 7463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 26.845769488781073,
+      "learning_rate": 5.602287657044116e-06,
+      "loss": 0.7181,
+      "step": 7464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.023030621874956,
+      "learning_rate": 5.601258704127835e-06,
+      "loss": 0.6912,
+      "step": 7465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.158037586584223,
+      "learning_rate": 5.600229725374129e-06,
+      "loss": 0.9471,
+      "step": 7466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.134094779130624,
+      "learning_rate": 5.599200720827215e-06,
+      "loss": 0.8998,
+      "step": 7467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0439321554601926,
+      "learning_rate": 5.598171690531312e-06,
+      "loss": 0.6831,
+      "step": 7468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6988732353268863,
+      "learning_rate": 5.597142634530639e-06,
+      "loss": 0.8225,
+      "step": 7469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9249885479850268,
+      "learning_rate": 5.5961135528694155e-06,
+      "loss": 0.7061,
+      "step": 7470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9089170942053149,
+      "learning_rate": 5.595084445591866e-06,
+      "loss": 0.7962,
+      "step": 7471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.239685366972101,
+      "learning_rate": 5.594055312742213e-06,
+      "loss": 0.8522,
+      "step": 7472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9600121872721805,
+      "learning_rate": 5.59302615436468e-06,
+      "loss": 0.7537,
+      "step": 7473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6513534763180195,
+      "learning_rate": 5.5919969705034914e-06,
+      "loss": 0.6771,
+      "step": 7474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9645539506367542,
+      "learning_rate": 5.590967761202876e-06,
+      "loss": 1.0874,
+      "step": 7475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2738921974411626,
+      "learning_rate": 5.589938526507059e-06,
+      "loss": 0.7899,
+      "step": 7476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0808263429502099,
+      "learning_rate": 5.5889092664602696e-06,
+      "loss": 0.587,
+      "step": 7477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9523906794202757,
+      "learning_rate": 5.5878799811067384e-06,
+      "loss": 0.8951,
+      "step": 7478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8127477986837506,
+      "learning_rate": 5.586850670490694e-06,
+      "loss": 0.8399,
+      "step": 7479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.096591976952861,
+      "learning_rate": 5.58582133465637e-06,
+      "loss": 0.7571,
+      "step": 7480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1906657839945365,
+      "learning_rate": 5.584791973647997e-06,
+      "loss": 0.7362,
+      "step": 7481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8505674247915733,
+      "learning_rate": 5.583762587509812e-06,
+      "loss": 0.7249,
+      "step": 7482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1284203025501616,
+      "learning_rate": 5.582733176286048e-06,
+      "loss": 0.6076,
+      "step": 7483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2085341043268785,
+      "learning_rate": 5.581703740020943e-06,
+      "loss": 0.9484,
+      "step": 7484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2321821682418603,
+      "learning_rate": 5.580674278758732e-06,
+      "loss": 0.5135,
+      "step": 7485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9603855679342104,
+      "learning_rate": 5.579644792543653e-06,
+      "loss": 0.8627,
+      "step": 7486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8454235301131723,
+      "learning_rate": 5.578615281419947e-06,
+      "loss": 0.7003,
+      "step": 7487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.897575322246279,
+      "learning_rate": 5.577585745431854e-06,
+      "loss": 0.8538,
+      "step": 7488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.157173022026394,
+      "learning_rate": 5.576556184623615e-06,
+      "loss": 0.6911,
+      "step": 7489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0999258075435203,
+      "learning_rate": 5.575526599039472e-06,
+      "loss": 0.8634,
+      "step": 7490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7661475127446036,
+      "learning_rate": 5.574496988723669e-06,
+      "loss": 0.7308,
+      "step": 7491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9206334828679674,
+      "learning_rate": 5.573467353720452e-06,
+      "loss": 0.9486,
+      "step": 7492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.451413448551331,
+      "learning_rate": 5.572437694074065e-06,
+      "loss": 0.6162,
+      "step": 7493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3032602604199406,
+      "learning_rate": 5.571408009828757e-06,
+      "loss": 0.9607,
+      "step": 7494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.018101372417441,
+      "learning_rate": 5.570378301028771e-06,
+      "loss": 0.8503,
+      "step": 7495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2980760580017083,
+      "learning_rate": 5.569348567718363e-06,
+      "loss": 0.6916,
+      "step": 7496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.538873435754768,
+      "learning_rate": 5.568318809941777e-06,
+      "loss": 0.7723,
+      "step": 7497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.444075602820668,
+      "learning_rate": 5.567289027743266e-06,
+      "loss": 0.8813,
+      "step": 7498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1269992916208724,
+      "learning_rate": 5.56625922116708e-06,
+      "loss": 0.9132,
+      "step": 7499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9704918576552926,
+      "learning_rate": 5.565229390257478e-06,
+      "loss": 0.5665,
+      "step": 7500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.017365827169991,
+      "learning_rate": 5.564199535058708e-06,
+      "loss": 0.8657,
+      "step": 7501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.739851128932091,
+      "learning_rate": 5.563169655615029e-06,
+      "loss": 1.0981,
+      "step": 7502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.224562926958004,
+      "learning_rate": 5.562139751970694e-06,
+      "loss": 0.7577,
+      "step": 7503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9563580610952077,
+      "learning_rate": 5.561109824169962e-06,
+      "loss": 0.7245,
+      "step": 7504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5108285038726477,
+      "learning_rate": 5.5600798722570925e-06,
+      "loss": 0.8211,
+      "step": 7505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3719677046865755,
+      "learning_rate": 5.559049896276343e-06,
+      "loss": 0.9186,
+      "step": 7506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.3559127256704615,
+      "learning_rate": 5.558019896271975e-06,
+      "loss": 0.8716,
+      "step": 7507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4506243043393345,
+      "learning_rate": 5.556989872288248e-06,
+      "loss": 0.8476,
+      "step": 7508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8368556883012512,
+      "learning_rate": 5.555959824369426e-06,
+      "loss": 0.7172,
+      "step": 7509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0556285099791807,
+      "learning_rate": 5.554929752559772e-06,
+      "loss": 0.6832,
+      "step": 7510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2763404778727843,
+      "learning_rate": 5.553899656903552e-06,
+      "loss": 0.8566,
+      "step": 7511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2154714023943543,
+      "learning_rate": 5.552869537445029e-06,
+      "loss": 0.7882,
+      "step": 7512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0601214394243947,
+      "learning_rate": 5.55183939422847e-06,
+      "loss": 0.6336,
+      "step": 7513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0018831687510596,
+      "learning_rate": 5.550809227298144e-06,
+      "loss": 0.7975,
+      "step": 7514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2138434290269062,
+      "learning_rate": 5.54977903669832e-06,
+      "loss": 0.8974,
+      "step": 7515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.155909314792517,
+      "learning_rate": 5.548748822473265e-06,
+      "loss": 0.6033,
+      "step": 7516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.970195099086099,
+      "learning_rate": 5.54771858466725e-06,
+      "loss": 0.9422,
+      "step": 7517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5734887126677624,
+      "learning_rate": 5.546688323324548e-06,
+      "loss": 0.9561,
+      "step": 7518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.501313111779903,
+      "learning_rate": 5.545658038489433e-06,
+      "loss": 0.8558,
+      "step": 7519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8590267895908468,
+      "learning_rate": 5.544627730206176e-06,
+      "loss": 0.6284,
+      "step": 7520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.41388217639501,
+      "learning_rate": 5.543597398519053e-06,
+      "loss": 0.7885,
+      "step": 7521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.471632748062193,
+      "learning_rate": 5.542567043472337e-06,
+      "loss": 0.8631,
+      "step": 7522
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8966408448695247,
+      "learning_rate": 5.541536665110309e-06,
+      "loss": 1.0121,
+      "step": 7523
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.290426003863555,
+      "learning_rate": 5.540506263477243e-06,
+      "loss": 0.9216,
+      "step": 7524
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3683438157031467,
+      "learning_rate": 5.53947583861742e-06,
+      "loss": 0.9555,
+      "step": 7525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2084494230148781,
+      "learning_rate": 5.5384453905751175e-06,
+      "loss": 0.6597,
+      "step": 7526
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3624089657381195,
+      "learning_rate": 5.537414919394618e-06,
+      "loss": 0.9038,
+      "step": 7527
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6316987832301018,
+      "learning_rate": 5.536384425120202e-06,
+      "loss": 0.8541,
+      "step": 7528
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2399752787584997,
+      "learning_rate": 5.535353907796155e-06,
+      "loss": 0.7573,
+      "step": 7529
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.289140303467432,
+      "learning_rate": 5.534323367466758e-06,
+      "loss": 0.901,
+      "step": 7530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9645102160684917,
+      "learning_rate": 5.5332928041762936e-06,
+      "loss": 0.8719,
+      "step": 7531
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2837692865442105,
+      "learning_rate": 5.5322622179690514e-06,
+      "loss": 0.9904,
+      "step": 7532
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.02159382589415,
+      "learning_rate": 5.531231608889313e-06,
+      "loss": 0.8657,
+      "step": 7533
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.717777650709329,
+      "learning_rate": 5.530200976981375e-06,
+      "loss": 0.703,
+      "step": 7534
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9680799816644947,
+      "learning_rate": 5.529170322289515e-06,
+      "loss": 0.665,
+      "step": 7535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.262676431859735,
+      "learning_rate": 5.528139644858031e-06,
+      "loss": 0.7718,
+      "step": 7536
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0103787436696776,
+      "learning_rate": 5.527108944731207e-06,
+      "loss": 0.8272,
+      "step": 7537
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9395224471441832,
+      "learning_rate": 5.526078221953341e-06,
+      "loss": 0.7469,
+      "step": 7538
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.661108666304138,
+      "learning_rate": 5.525047476568722e-06,
+      "loss": 0.6648,
+      "step": 7539
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.547632354808202,
+      "learning_rate": 5.524016708621641e-06,
+      "loss": 0.8295,
+      "step": 7540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8938290234821293,
+      "learning_rate": 5.522985918156397e-06,
+      "loss": 0.6644,
+      "step": 7541
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1309470420801024,
+      "learning_rate": 5.5219551052172825e-06,
+      "loss": 0.6745,
+      "step": 7542
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2965952672502645,
+      "learning_rate": 5.5209242698485955e-06,
+      "loss": 0.7775,
+      "step": 7543
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.863030662154914,
+      "learning_rate": 5.519893412094631e-06,
+      "loss": 0.6808,
+      "step": 7544
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2455801091009318,
+      "learning_rate": 5.518862531999691e-06,
+      "loss": 0.8085,
+      "step": 7545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.345268226668162,
+      "learning_rate": 5.51783162960807e-06,
+      "loss": 0.8561,
+      "step": 7546
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2271801622790064,
+      "learning_rate": 5.516800704964071e-06,
+      "loss": 0.9716,
+      "step": 7547
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.096304557821954,
+      "learning_rate": 5.515769758111996e-06,
+      "loss": 0.8872,
+      "step": 7548
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7333541998646809,
+      "learning_rate": 5.514738789096146e-06,
+      "loss": 0.8699,
+      "step": 7549
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2574337060724305,
+      "learning_rate": 5.513707797960823e-06,
+      "loss": 0.828,
+      "step": 7550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9193587952971773,
+      "learning_rate": 5.512676784750332e-06,
+      "loss": 0.829,
+      "step": 7551
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.453010240027757,
+      "learning_rate": 5.511645749508977e-06,
+      "loss": 0.9006,
+      "step": 7552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0017638155124202,
+      "learning_rate": 5.5106146922810664e-06,
+      "loss": 0.6838,
+      "step": 7553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.696094747251633,
+      "learning_rate": 5.509583613110904e-06,
+      "loss": 0.8967,
+      "step": 7554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7769062884102813,
+      "learning_rate": 5.508552512042799e-06,
+      "loss": 0.6654,
+      "step": 7555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0752521960270078,
+      "learning_rate": 5.50752138912106e-06,
+      "loss": 0.6812,
+      "step": 7556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9643802925946054,
+      "learning_rate": 5.506490244389997e-06,
+      "loss": 0.6199,
+      "step": 7557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.928544532576195,
+      "learning_rate": 5.50545907789392e-06,
+      "loss": 0.9321,
+      "step": 7558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.672706792033826,
+      "learning_rate": 5.504427889677141e-06,
+      "loss": 1.0044,
+      "step": 7559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.019706862546537,
+      "learning_rate": 5.50339667978397e-06,
+      "loss": 0.8745,
+      "step": 7560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3726640848767007,
+      "learning_rate": 5.502365448258724e-06,
+      "loss": 0.7177,
+      "step": 7561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.428032302926816,
+      "learning_rate": 5.5013341951457166e-06,
+      "loss": 0.919,
+      "step": 7562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.257804916803708,
+      "learning_rate": 5.500302920489262e-06,
+      "loss": 0.7545,
+      "step": 7563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4406860436496056,
+      "learning_rate": 5.499271624333676e-06,
+      "loss": 0.7316,
+      "step": 7564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2429493441382282,
+      "learning_rate": 5.498240306723277e-06,
+      "loss": 0.7531,
+      "step": 7565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.140311749658042,
+      "learning_rate": 5.497208967702382e-06,
+      "loss": 0.6079,
+      "step": 7566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.71429822642466,
+      "learning_rate": 5.496177607315312e-06,
+      "loss": 0.8273,
+      "step": 7567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2011782924312246,
+      "learning_rate": 5.495146225606383e-06,
+      "loss": 0.5861,
+      "step": 7568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.215117250311774,
+      "learning_rate": 5.494114822619918e-06,
+      "loss": 0.6935,
+      "step": 7569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9266250057659375,
+      "learning_rate": 5.493083398400239e-06,
+      "loss": 0.8499,
+      "step": 7570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6375987604068063,
+      "learning_rate": 5.492051952991669e-06,
+      "loss": 0.6024,
+      "step": 7571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9493423003402763,
+      "learning_rate": 5.4910204864385306e-06,
+      "loss": 0.9201,
+      "step": 7572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1786545317354022,
+      "learning_rate": 5.489988998785147e-06,
+      "loss": 0.776,
+      "step": 7573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3002858409365783,
+      "learning_rate": 5.488957490075846e-06,
+      "loss": 1.0125,
+      "step": 7574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4907138753672506,
+      "learning_rate": 5.487925960354953e-06,
+      "loss": 0.8707,
+      "step": 7575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.328219611202035,
+      "learning_rate": 5.486894409666794e-06,
+      "loss": 0.8607,
+      "step": 7576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0533231937104475,
+      "learning_rate": 5.485862838055699e-06,
+      "loss": 0.7305,
+      "step": 7577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.51369379790848,
+      "learning_rate": 5.4848312455659935e-06,
+      "loss": 0.8565,
+      "step": 7578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2185372665936303,
+      "learning_rate": 5.483799632242012e-06,
+      "loss": 0.8647,
+      "step": 7579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1870915454206084,
+      "learning_rate": 5.482767998128081e-06,
+      "loss": 0.7774,
+      "step": 7580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0694169797213795,
+      "learning_rate": 5.4817363432685355e-06,
+      "loss": 0.8879,
+      "step": 7581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7607081457141933,
+      "learning_rate": 5.4807046677077045e-06,
+      "loss": 0.7668,
+      "step": 7582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1693772820462573,
+      "learning_rate": 5.479672971489924e-06,
+      "loss": 0.9248,
+      "step": 7583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5196483290998626,
+      "learning_rate": 5.478641254659528e-06,
+      "loss": 0.7316,
+      "step": 7584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.131756012600528,
+      "learning_rate": 5.4776095172608535e-06,
+      "loss": 0.8343,
+      "step": 7585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0869733937306083,
+      "learning_rate": 5.476577759338232e-06,
+      "loss": 0.6229,
+      "step": 7586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5034107736005136,
+      "learning_rate": 5.475545980936003e-06,
+      "loss": 0.9352,
+      "step": 7587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8351722639562371,
+      "learning_rate": 5.474514182098504e-06,
+      "loss": 0.7636,
+      "step": 7588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0662432990712567,
+      "learning_rate": 5.473482362870073e-06,
+      "loss": 0.7802,
+      "step": 7589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.181396113765109,
+      "learning_rate": 5.472450523295052e-06,
+      "loss": 0.9015,
+      "step": 7590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3495115094515504,
+      "learning_rate": 5.471418663417779e-06,
+      "loss": 0.8997,
+      "step": 7591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7902970516169325,
+      "learning_rate": 5.470386783282597e-06,
+      "loss": 0.7421,
+      "step": 7592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9166025393097104,
+      "learning_rate": 5.469354882933845e-06,
+      "loss": 0.9458,
+      "step": 7593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0213983139033043,
+      "learning_rate": 5.468322962415871e-06,
+      "loss": 0.883,
+      "step": 7594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.115000606126187,
+      "learning_rate": 5.4672910217730155e-06,
+      "loss": 1.0327,
+      "step": 7595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.157667119623365,
+      "learning_rate": 5.4662590610496235e-06,
+      "loss": 0.6064,
+      "step": 7596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.03892809461472,
+      "learning_rate": 5.465227080290042e-06,
+      "loss": 0.974,
+      "step": 7597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4925272350973824,
+      "learning_rate": 5.464195079538615e-06,
+      "loss": 0.7722,
+      "step": 7598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.096829773089846,
+      "learning_rate": 5.463163058839694e-06,
+      "loss": 0.6682,
+      "step": 7599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1070681000098976,
+      "learning_rate": 5.462131018237623e-06,
+      "loss": 0.6108,
+      "step": 7600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8398717635396082,
+      "learning_rate": 5.461098957776755e-06,
+      "loss": 0.732,
+      "step": 7601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2792750522356435,
+      "learning_rate": 5.4600668775014355e-06,
+      "loss": 0.8178,
+      "step": 7602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1563346011097697,
+      "learning_rate": 5.459034777456018e-06,
+      "loss": 0.791,
+      "step": 7603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.217003986704711,
+      "learning_rate": 5.4580026576848565e-06,
+      "loss": 0.9067,
+      "step": 7604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.235862207185154,
+      "learning_rate": 5.4569705182322975e-06,
+      "loss": 0.9768,
+      "step": 7605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.654687653191058,
+      "learning_rate": 5.455938359142699e-06,
+      "loss": 0.8646,
+      "step": 7606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1445170823998334,
+      "learning_rate": 5.454906180460413e-06,
+      "loss": 0.6706,
+      "step": 7607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0398137036122046,
+      "learning_rate": 5.453873982229797e-06,
+      "loss": 0.7685,
+      "step": 7608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9133281250137464,
+      "learning_rate": 5.452841764495203e-06,
+      "loss": 0.9286,
+      "step": 7609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0886623989898405,
+      "learning_rate": 5.451809527300992e-06,
+      "loss": 0.7247,
+      "step": 7610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3272347502451707,
+      "learning_rate": 5.450777270691517e-06,
+      "loss": 0.8617,
+      "step": 7611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8666002419531191,
+      "learning_rate": 5.449744994711141e-06,
+      "loss": 0.8578,
+      "step": 7612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1933530563237373,
+      "learning_rate": 5.448712699404221e-06,
+      "loss": 0.9221,
+      "step": 7613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9942466534977017,
+      "learning_rate": 5.4476803848151146e-06,
+      "loss": 0.9064,
+      "step": 7614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2583683324416226,
+      "learning_rate": 5.446648050988187e-06,
+      "loss": 0.7667,
+      "step": 7615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.247179257490837,
+      "learning_rate": 5.445615697967797e-06,
+      "loss": 0.9974,
+      "step": 7616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4750694938604965,
+      "learning_rate": 5.444583325798308e-06,
+      "loss": 0.7125,
+      "step": 7617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0792251288806947,
+      "learning_rate": 5.443550934524085e-06,
+      "loss": 0.767,
+      "step": 7618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.149022605831248,
+      "learning_rate": 5.442518524189489e-06,
+      "loss": 0.6501,
+      "step": 7619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5326945296056214,
+      "learning_rate": 5.441486094838886e-06,
+      "loss": 0.8562,
+      "step": 7620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5871029771853795,
+      "learning_rate": 5.440453646516642e-06,
+      "loss": 0.7597,
+      "step": 7621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9620912705227127,
+      "learning_rate": 5.439421179267126e-06,
+      "loss": 0.7239,
+      "step": 7622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.529604774592548,
+      "learning_rate": 5.438388693134702e-06,
+      "loss": 0.6815,
+      "step": 7623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4997219921043228,
+      "learning_rate": 5.4373561881637405e-06,
+      "loss": 0.8413,
+      "step": 7624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.507903762081633,
+      "learning_rate": 5.436323664398607e-06,
+      "loss": 0.6975,
+      "step": 7625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9644129368376095,
+      "learning_rate": 5.435291121883675e-06,
+      "loss": 0.8469,
+      "step": 7626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9347852726555093,
+      "learning_rate": 5.434258560663316e-06,
+      "loss": 0.9693,
+      "step": 7627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2885604330955194,
+      "learning_rate": 5.433225980781898e-06,
+      "loss": 0.8127,
+      "step": 7628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.416215082832181,
+      "learning_rate": 5.432193382283794e-06,
+      "loss": 0.8187,
+      "step": 7629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1793449869362145,
+      "learning_rate": 5.431160765213379e-06,
+      "loss": 0.7098,
+      "step": 7630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.968508861598905,
+      "learning_rate": 5.430128129615025e-06,
+      "loss": 0.6371,
+      "step": 7631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.146362216009918,
+      "learning_rate": 5.429095475533108e-06,
+      "loss": 0.9967,
+      "step": 7632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0462081811344466,
+      "learning_rate": 5.428062803012003e-06,
+      "loss": 0.7865,
+      "step": 7633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0311365594218436,
+      "learning_rate": 5.4270301120960856e-06,
+      "loss": 0.858,
+      "step": 7634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3924921225255904,
+      "learning_rate": 5.425997402829732e-06,
+      "loss": 0.8809,
+      "step": 7635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1005377512725154,
+      "learning_rate": 5.424964675257324e-06,
+      "loss": 0.7302,
+      "step": 7636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3282073020814753,
+      "learning_rate": 5.423931929423235e-06,
+      "loss": 0.774,
+      "step": 7637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.978358909595548,
+      "learning_rate": 5.422899165371846e-06,
+      "loss": 0.7665,
+      "step": 7638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8566647783924632,
+      "learning_rate": 5.421866383147541e-06,
+      "loss": 0.7503,
+      "step": 7639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.560184672480332,
+      "learning_rate": 5.420833582794696e-06,
+      "loss": 0.8534,
+      "step": 7640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.177269853583522,
+      "learning_rate": 5.419800764357697e-06,
+      "loss": 0.9281,
+      "step": 7641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1882342880154375,
+      "learning_rate": 5.418767927880923e-06,
+      "loss": 0.6871,
+      "step": 7642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.111502132128684,
+      "learning_rate": 5.417735073408756e-06,
+      "loss": 0.6691,
+      "step": 7643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.31023649696484,
+      "learning_rate": 5.416702200985585e-06,
+      "loss": 0.7348,
+      "step": 7644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9830601830870898,
+      "learning_rate": 5.415669310655791e-06,
+      "loss": 0.7169,
+      "step": 7645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9468045817781123,
+      "learning_rate": 5.414636402463764e-06,
+      "loss": 0.9038,
+      "step": 7646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.527452501904816,
+      "learning_rate": 5.413603476453884e-06,
+      "loss": 0.9309,
+      "step": 7647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7236722085500897,
+      "learning_rate": 5.4125705326705425e-06,
+      "loss": 0.6247,
+      "step": 7648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9821914463288807,
+      "learning_rate": 5.411537571158127e-06,
+      "loss": 0.7835,
+      "step": 7649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6433347246461145,
+      "learning_rate": 5.410504591961025e-06,
+      "loss": 0.8872,
+      "step": 7650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2524384686178314,
+      "learning_rate": 5.409471595123628e-06,
+      "loss": 0.958,
+      "step": 7651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.156304645418658,
+      "learning_rate": 5.408438580690322e-06,
+      "loss": 0.8273,
+      "step": 7652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8196455492183663,
+      "learning_rate": 5.4074055487055025e-06,
+      "loss": 0.7854,
+      "step": 7653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.869567690242526,
+      "learning_rate": 5.406372499213557e-06,
+      "loss": 0.9517,
+      "step": 7654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6383038023550607,
+      "learning_rate": 5.4053394322588825e-06,
+      "loss": 1.2854,
+      "step": 7655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7478443023343475,
+      "learning_rate": 5.404306347885868e-06,
+      "loss": 0.9163,
+      "step": 7656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.064187465168344,
+      "learning_rate": 5.403273246138912e-06,
+      "loss": 0.9283,
+      "step": 7657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.30535853014505,
+      "learning_rate": 5.4022401270624036e-06,
+      "loss": 0.7965,
+      "step": 7658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.126152331951871,
+      "learning_rate": 5.401206990700741e-06,
+      "loss": 0.8506,
+      "step": 7659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8432394348384225,
+      "learning_rate": 5.400173837098324e-06,
+      "loss": 0.8386,
+      "step": 7660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.467154274111314,
+      "learning_rate": 5.399140666299543e-06,
+      "loss": 0.742,
+      "step": 7661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.331226688879289,
+      "learning_rate": 5.3981074783488e-06,
+      "loss": 0.8726,
+      "step": 7662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8786000907964215,
+      "learning_rate": 5.3970742732904905e-06,
+      "loss": 0.8483,
+      "step": 7663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.09763960510006,
+      "learning_rate": 5.396041051169016e-06,
+      "loss": 0.9979,
+      "step": 7664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.385392027500074,
+      "learning_rate": 5.395007812028775e-06,
+      "loss": 0.7629,
+      "step": 7665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1750539367814923,
+      "learning_rate": 5.39397455591417e-06,
+      "loss": 0.7065,
+      "step": 7666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3624527889151477,
+      "learning_rate": 5.3929412828696006e-06,
+      "loss": 0.8421,
+      "step": 7667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5523497698510997,
+      "learning_rate": 5.3919079929394665e-06,
+      "loss": 0.742,
+      "step": 7668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6539500468038724,
+      "learning_rate": 5.390874686168176e-06,
+      "loss": 0.7383,
+      "step": 7669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3169429285070158,
+      "learning_rate": 5.389841362600129e-06,
+      "loss": 0.8565,
+      "step": 7670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9562208392437606,
+      "learning_rate": 5.388808022279731e-06,
+      "loss": 0.8324,
+      "step": 7671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8320554333696069,
+      "learning_rate": 5.387774665251385e-06,
+      "loss": 0.8516,
+      "step": 7672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7742486108499085,
+      "learning_rate": 5.386741291559499e-06,
+      "loss": 0.9501,
+      "step": 7673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.289918037394939,
+      "learning_rate": 5.385707901248478e-06,
+      "loss": 0.894,
+      "step": 7674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4013194899812214,
+      "learning_rate": 5.3846744943627295e-06,
+      "loss": 0.823,
+      "step": 7675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8506823481003016,
+      "learning_rate": 5.38364107094666e-06,
+      "loss": 0.8953,
+      "step": 7676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3122209008178043,
+      "learning_rate": 5.382607631044682e-06,
+      "loss": 0.7809,
+      "step": 7677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.999933060106536,
+      "learning_rate": 5.381574174701201e-06,
+      "loss": 0.9295,
+      "step": 7678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1101812028551923,
+      "learning_rate": 5.380540701960627e-06,
+      "loss": 0.73,
+      "step": 7679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3442810816998803,
+      "learning_rate": 5.379507212867372e-06,
+      "loss": 0.862,
+      "step": 7680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1810041474570725,
+      "learning_rate": 5.378473707465847e-06,
+      "loss": 0.6307,
+      "step": 7681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4578092424007676,
+      "learning_rate": 5.377440185800463e-06,
+      "loss": 0.7843,
+      "step": 7682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2901174912314306,
+      "learning_rate": 5.376406647915635e-06,
+      "loss": 0.8614,
+      "step": 7683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3812497282299128,
+      "learning_rate": 5.375373093855774e-06,
+      "loss": 0.6618,
+      "step": 7684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1541842566138882,
+      "learning_rate": 5.374339523665295e-06,
+      "loss": 0.9132,
+      "step": 7685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1098988142898023,
+      "learning_rate": 5.373305937388613e-06,
+      "loss": 0.7924,
+      "step": 7686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1105486579071506,
+      "learning_rate": 5.372272335070145e-06,
+      "loss": 0.856,
+      "step": 7687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.277488691224032,
+      "learning_rate": 5.371238716754305e-06,
+      "loss": 0.8931,
+      "step": 7688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9225573797068913,
+      "learning_rate": 5.37020508248551e-06,
+      "loss": 0.9269,
+      "step": 7689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0499800952804024,
+      "learning_rate": 5.369171432308178e-06,
+      "loss": 0.7888,
+      "step": 7690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9440349180554728,
+      "learning_rate": 5.368137766266728e-06,
+      "loss": 0.781,
+      "step": 7691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2146404201199696,
+      "learning_rate": 5.367104084405578e-06,
+      "loss": 0.8457,
+      "step": 7692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.629958639015794,
+      "learning_rate": 5.3660703867691475e-06,
+      "loss": 1.0064,
+      "step": 7693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.98008396374203,
+      "learning_rate": 5.365036673401857e-06,
+      "loss": 0.8295,
+      "step": 7694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1241200790210244,
+      "learning_rate": 5.364002944348129e-06,
+      "loss": 0.7174,
+      "step": 7695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1104281945031205,
+      "learning_rate": 5.362969199652382e-06,
+      "loss": 0.806,
+      "step": 7696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7896307245806762,
+      "learning_rate": 5.361935439359042e-06,
+      "loss": 0.7171,
+      "step": 7697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.893867329768058,
+      "learning_rate": 5.36090166351253e-06,
+      "loss": 0.5862,
+      "step": 7698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9895947852451397,
+      "learning_rate": 5.359867872157267e-06,
+      "loss": 0.7392,
+      "step": 7699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8538833609473055,
+      "learning_rate": 5.358834065337684e-06,
+      "loss": 0.7311,
+      "step": 7700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0426782039491966,
+      "learning_rate": 5.357800243098199e-06,
+      "loss": 0.7146,
+      "step": 7701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.282800739777742,
+      "learning_rate": 5.3567664054832426e-06,
+      "loss": 0.7298,
+      "step": 7702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.733085292540357,
+      "learning_rate": 5.355732552537237e-06,
+      "loss": 0.8316,
+      "step": 7703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2197654972957794,
+      "learning_rate": 5.354698684304613e-06,
+      "loss": 0.8563,
+      "step": 7704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5974493526388205,
+      "learning_rate": 5.353664800829796e-06,
+      "loss": 0.7064,
+      "step": 7705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9609098681209336,
+      "learning_rate": 5.352630902157217e-06,
+      "loss": 0.7552,
+      "step": 7706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7512185145396684,
+      "learning_rate": 5.3515969883313e-06,
+      "loss": 0.7709,
+      "step": 7707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9049651040157658,
+      "learning_rate": 5.350563059396478e-06,
+      "loss": 0.7787,
+      "step": 7708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5338130658217712,
+      "learning_rate": 5.3495291153971806e-06,
+      "loss": 0.7676,
+      "step": 7709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1614811736038182,
+      "learning_rate": 5.348495156377839e-06,
+      "loss": 0.7158,
+      "step": 7710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.483534968349314,
+      "learning_rate": 5.347461182382886e-06,
+      "loss": 0.7789,
+      "step": 7711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4282918895659615,
+      "learning_rate": 5.346427193456749e-06,
+      "loss": 0.7411,
+      "step": 7712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7136495639259632,
+      "learning_rate": 5.345393189643866e-06,
+      "loss": 0.8056,
+      "step": 7713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6674451271323978,
+      "learning_rate": 5.344359170988668e-06,
+      "loss": 0.8256,
+      "step": 7714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7631171940952803,
+      "learning_rate": 5.3433251375355884e-06,
+      "loss": 0.8982,
+      "step": 7715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.791467582008035,
+      "learning_rate": 5.342291089329066e-06,
+      "loss": 0.8631,
+      "step": 7716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8245448169749936,
+      "learning_rate": 5.34125702641353e-06,
+      "loss": 0.8753,
+      "step": 7717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1255116836563195,
+      "learning_rate": 5.340222948833421e-06,
+      "loss": 0.7533,
+      "step": 7718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3493060404133117,
+      "learning_rate": 5.339188856633173e-06,
+      "loss": 0.8416,
+      "step": 7719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8975175130630033,
+      "learning_rate": 5.338154749857225e-06,
+      "loss": 0.836,
+      "step": 7720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4626255120040654,
+      "learning_rate": 5.337120628550016e-06,
+      "loss": 0.9336,
+      "step": 7721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.122697824224339,
+      "learning_rate": 5.3360864927559805e-06,
+      "loss": 0.7968,
+      "step": 7722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.852597677376704,
+      "learning_rate": 5.335052342519561e-06,
+      "loss": 0.8599,
+      "step": 7723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.355055151416337,
+      "learning_rate": 5.3340181778851954e-06,
+      "loss": 0.963,
+      "step": 7724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8532149594038723,
+      "learning_rate": 5.332983998897327e-06,
+      "loss": 0.7473,
+      "step": 7725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5701750316185143,
+      "learning_rate": 5.331949805600391e-06,
+      "loss": 0.7718,
+      "step": 7726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.051223631783717,
+      "learning_rate": 5.330915598038835e-06,
+      "loss": 0.651,
+      "step": 7727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1731709831984025,
+      "learning_rate": 5.329881376257098e-06,
+      "loss": 0.9161,
+      "step": 7728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0030465995662894,
+      "learning_rate": 5.328847140299624e-06,
+      "loss": 0.8346,
+      "step": 7729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2003783086510977,
+      "learning_rate": 5.327812890210856e-06,
+      "loss": 0.8884,
+      "step": 7730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1108896721848727,
+      "learning_rate": 5.326778626035238e-06,
+      "loss": 0.7336,
+      "step": 7731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.104208877986731,
+      "learning_rate": 5.3257443478172146e-06,
+      "loss": 0.6728,
+      "step": 7732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2073340525783594,
+      "learning_rate": 5.32471005560123e-06,
+      "loss": 0.8886,
+      "step": 7733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.274851200767253,
+      "learning_rate": 5.323675749431732e-06,
+      "loss": 0.8535,
+      "step": 7734
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.525060157179888,
+      "learning_rate": 5.322641429353167e-06,
+      "loss": 0.8798,
+      "step": 7735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.673910326958343,
+      "learning_rate": 5.321607095409981e-06,
+      "loss": 0.7885,
+      "step": 7736
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.159403079753894,
+      "learning_rate": 5.3205727476466206e-06,
+      "loss": 1.0125,
+      "step": 7737
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1226954346139135,
+      "learning_rate": 5.3195383861075365e-06,
+      "loss": 0.8492,
+      "step": 7738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2557423352259036,
+      "learning_rate": 5.318504010837175e-06,
+      "loss": 0.8117,
+      "step": 7739
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2810696767861884,
+      "learning_rate": 5.3174696218799894e-06,
+      "loss": 0.6123,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9709620801848375,
+      "learning_rate": 5.316435219280425e-06,
+      "loss": 1.0158,
+      "step": 7741
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2069115341114145,
+      "learning_rate": 5.315400803082934e-06,
+      "loss": 0.7456,
+      "step": 7742
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3713847143904796,
+      "learning_rate": 5.314366373331969e-06,
+      "loss": 0.8063,
+      "step": 7743
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.118515600167811,
+      "learning_rate": 5.313331930071981e-06,
+      "loss": 0.9044,
+      "step": 7744
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.547307481801456,
+      "learning_rate": 5.312297473347422e-06,
+      "loss": 0.828,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1813156256630433,
+      "learning_rate": 5.311263003202745e-06,
+      "loss": 0.8221,
+      "step": 7746
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.446239674153846,
+      "learning_rate": 5.310228519682403e-06,
+      "loss": 0.7901,
+      "step": 7747
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2014460287938107,
+      "learning_rate": 5.309194022830852e-06,
+      "loss": 0.9582,
+      "step": 7748
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.089679970361889,
+      "learning_rate": 5.308159512692544e-06,
+      "loss": 0.6006,
+      "step": 7749
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.291188945999785,
+      "learning_rate": 5.307124989311936e-06,
+      "loss": 0.7394,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3881454807530753,
+      "learning_rate": 5.306090452733482e-06,
+      "loss": 0.8397,
+      "step": 7751
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.412844149121066,
+      "learning_rate": 5.305055903001642e-06,
+      "loss": 0.7692,
+      "step": 7752
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.039258281378359,
+      "learning_rate": 5.304021340160868e-06,
+      "loss": 0.8771,
+      "step": 7753
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.586097202487261,
+      "learning_rate": 5.302986764255621e-06,
+      "loss": 0.7145,
+      "step": 7754
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0923761141145656,
+      "learning_rate": 5.301952175330357e-06,
+      "loss": 0.7857,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1202067427515043,
+      "learning_rate": 5.300917573429536e-06,
+      "loss": 0.6518,
+      "step": 7756
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9898068276801566,
+      "learning_rate": 5.299882958597617e-06,
+      "loss": 0.7374,
+      "step": 7757
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.372953032041102,
+      "learning_rate": 5.298848330879061e-06,
+      "loss": 1.0642,
+      "step": 7758
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.456847425673978,
+      "learning_rate": 5.297813690318325e-06,
+      "loss": 0.881,
+      "step": 7759
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.31535598407513,
+      "learning_rate": 5.29677903695987e-06,
+      "loss": 0.892,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8841385050844726,
+      "learning_rate": 5.295744370848162e-06,
+      "loss": 0.6087,
+      "step": 7761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1893391307890475,
+      "learning_rate": 5.294709692027658e-06,
+      "loss": 0.842,
+      "step": 7762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.921545904678426,
+      "learning_rate": 5.293675000542822e-06,
+      "loss": 0.7944,
+      "step": 7763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1752765714178346,
+      "learning_rate": 5.292640296438116e-06,
+      "loss": 0.8936,
+      "step": 7764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.615070732722082,
+      "learning_rate": 5.291605579758007e-06,
+      "loss": 0.8789,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.540108250373315,
+      "learning_rate": 5.290570850546954e-06,
+      "loss": 0.8932,
+      "step": 7766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.17632511900437,
+      "learning_rate": 5.289536108849429e-06,
+      "loss": 0.6683,
+      "step": 7767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3602194900574043,
+      "learning_rate": 5.28850135470989e-06,
+      "loss": 0.7194,
+      "step": 7768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0164612604505567,
+      "learning_rate": 5.287466588172804e-06,
+      "loss": 0.5581,
+      "step": 7769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.888388463837683,
+      "learning_rate": 5.286431809282639e-06,
+      "loss": 0.795,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0052673464293798,
+      "learning_rate": 5.2853970180838624e-06,
+      "loss": 0.6109,
+      "step": 7771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2372395788800636,
+      "learning_rate": 5.2843622146209405e-06,
+      "loss": 0.8785,
+      "step": 7772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.150252838217724,
+      "learning_rate": 5.283327398938339e-06,
+      "loss": 0.826,
+      "step": 7773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0656700300605375,
+      "learning_rate": 5.2822925710805305e-06,
+      "loss": 0.7565,
+      "step": 7774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2060934053553103,
+      "learning_rate": 5.28125773109198e-06,
+      "loss": 0.8792,
+      "step": 7775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0437774163103284,
+      "learning_rate": 5.2802228790171596e-06,
+      "loss": 0.7985,
+      "step": 7776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1011515384681987,
+      "learning_rate": 5.279188014900537e-06,
+      "loss": 0.5922,
+      "step": 7777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1293233760020671,
+      "learning_rate": 5.278153138786583e-06,
+      "loss": 0.65,
+      "step": 7778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.264514018408156,
+      "learning_rate": 5.27711825071977e-06,
+      "loss": 0.8848,
+      "step": 7779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0919745242493697,
+      "learning_rate": 5.276083350744568e-06,
+      "loss": 0.7903,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4216570123365995,
+      "learning_rate": 5.275048438905452e-06,
+      "loss": 0.8838,
+      "step": 7781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9752022158984712,
+      "learning_rate": 5.27401351524689e-06,
+      "loss": 0.8391,
+      "step": 7782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.389915945424939,
+      "learning_rate": 5.272978579813357e-06,
+      "loss": 0.7122,
+      "step": 7783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1943629455764073,
+      "learning_rate": 5.2719436326493255e-06,
+      "loss": 0.6868,
+      "step": 7784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1967168542746613,
+      "learning_rate": 5.270908673799272e-06,
+      "loss": 0.6494,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8044551370001365,
+      "learning_rate": 5.26987370330767e-06,
+      "loss": 0.727,
+      "step": 7786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4935534562734296,
+      "learning_rate": 5.2688387212189915e-06,
+      "loss": 0.9725,
+      "step": 7787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3523838603598484,
+      "learning_rate": 5.267803727577717e-06,
+      "loss": 0.9533,
+      "step": 7788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1430421907659314,
+      "learning_rate": 5.266768722428318e-06,
+      "loss": 1.0912,
+      "step": 7789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8875768683595953,
+      "learning_rate": 5.2657337058152735e-06,
+      "loss": 0.6367,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8996229735038486,
+      "learning_rate": 5.26469867778306e-06,
+      "loss": 0.8483,
+      "step": 7791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2203913065994443,
+      "learning_rate": 5.263663638376155e-06,
+      "loss": 0.6658,
+      "step": 7792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9774324708681037,
+      "learning_rate": 5.262628587639036e-06,
+      "loss": 0.6391,
+      "step": 7793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6665523263706556,
+      "learning_rate": 5.261593525616181e-06,
+      "loss": 0.9497,
+      "step": 7794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1150007054438005,
+      "learning_rate": 5.2605584523520715e-06,
+      "loss": 0.7647,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1615726875192367,
+      "learning_rate": 5.259523367891184e-06,
+      "loss": 0.7802,
+      "step": 7796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.095179140485325,
+      "learning_rate": 5.258488272277998e-06,
+      "loss": 0.6653,
+      "step": 7797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.182693068416853,
+      "learning_rate": 5.257453165556996e-06,
+      "loss": 0.8533,
+      "step": 7798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9476350622409218,
+      "learning_rate": 5.256418047772659e-06,
+      "loss": 0.8322,
+      "step": 7799
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9440387616834882,
+      "learning_rate": 5.255382918969467e-06,
+      "loss": 0.8065,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.497739103817519,
+      "learning_rate": 5.2543477791919025e-06,
+      "loss": 0.8428,
+      "step": 7801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.269515396746968,
+      "learning_rate": 5.253312628484447e-06,
+      "loss": 0.7841,
+      "step": 7802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9466225301204463,
+      "learning_rate": 5.252277466891583e-06,
+      "loss": 0.7323,
+      "step": 7803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.245802720233631,
+      "learning_rate": 5.251242294457796e-06,
+      "loss": 0.9474,
+      "step": 7804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1674863283705883,
+      "learning_rate": 5.2502071112275675e-06,
+      "loss": 0.9396,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.014687200946636,
+      "learning_rate": 5.249171917245382e-06,
+      "loss": 0.8543,
+      "step": 7806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5678310384045107,
+      "learning_rate": 5.248136712555724e-06,
+      "loss": 0.8904,
+      "step": 7807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.035075709793111,
+      "learning_rate": 5.247101497203079e-06,
+      "loss": 0.696,
+      "step": 7808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0468102469490073,
+      "learning_rate": 5.2460662712319335e-06,
+      "loss": 0.742,
+      "step": 7809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2450088787062894,
+      "learning_rate": 5.24503103468677e-06,
+      "loss": 0.7956,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0154244152211334,
+      "learning_rate": 5.243995787612078e-06,
+      "loss": 0.818,
+      "step": 7811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.976374355269559,
+      "learning_rate": 5.242960530052344e-06,
+      "loss": 0.6974,
+      "step": 7812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1876427739076325,
+      "learning_rate": 5.241925262052055e-06,
+      "loss": 0.825,
+      "step": 7813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0191659806809032,
+      "learning_rate": 5.240889983655701e-06,
+      "loss": 0.8925,
+      "step": 7814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9423331793924807,
+      "learning_rate": 5.239854694907764e-06,
+      "loss": 0.892,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2074503030077435,
+      "learning_rate": 5.238819395852738e-06,
+      "loss": 0.8245,
+      "step": 7816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2409934879138946,
+      "learning_rate": 5.2377840865351115e-06,
+      "loss": 0.635,
+      "step": 7817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2185529225017215,
+      "learning_rate": 5.236748766999372e-06,
+      "loss": 0.7961,
+      "step": 7818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.840666063960712,
+      "learning_rate": 5.235713437290012e-06,
+      "loss": 0.8118,
+      "step": 7819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.601473302396734,
+      "learning_rate": 5.234678097451519e-06,
+      "loss": 0.8043,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9141765959402093,
+      "learning_rate": 5.233642747528387e-06,
+      "loss": 0.7787,
+      "step": 7821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.778272028427539,
+      "learning_rate": 5.232607387565104e-06,
+      "loss": 0.7019,
+      "step": 7822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7183600765939473,
+      "learning_rate": 5.231572017606166e-06,
+      "loss": 0.7828,
+      "step": 7823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.220255484232171,
+      "learning_rate": 5.230536637696062e-06,
+      "loss": 0.9741,
+      "step": 7824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4179871056792077,
+      "learning_rate": 5.229501247879284e-06,
+      "loss": 0.7601,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.179847614136511,
+      "learning_rate": 5.228465848200327e-06,
+      "loss": 0.8532,
+      "step": 7826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0399405920558542,
+      "learning_rate": 5.227430438703683e-06,
+      "loss": 0.6172,
+      "step": 7827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9828776017009709,
+      "learning_rate": 5.226395019433847e-06,
+      "loss": 0.7222,
+      "step": 7828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5486221152396777,
+      "learning_rate": 5.225359590435312e-06,
+      "loss": 0.7125,
+      "step": 7829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.507667004835452,
+      "learning_rate": 5.224324151752575e-06,
+      "loss": 1.0774,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9813761362771138,
+      "learning_rate": 5.223288703430128e-06,
+      "loss": 0.8592,
+      "step": 7831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.815386982347808,
+      "learning_rate": 5.222253245512471e-06,
+      "loss": 0.8479,
+      "step": 7832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.02473366622345,
+      "learning_rate": 5.221217778044096e-06,
+      "loss": 0.7892,
+      "step": 7833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1559156273482063,
+      "learning_rate": 5.220182301069499e-06,
+      "loss": 0.5537,
+      "step": 7834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.015380946783386,
+      "learning_rate": 5.219146814633178e-06,
+      "loss": 0.8807,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7518856679915322,
+      "learning_rate": 5.218111318779631e-06,
+      "loss": 0.9557,
+      "step": 7836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.152145369403063,
+      "learning_rate": 5.2170758135533554e-06,
+      "loss": 0.9635,
+      "step": 7837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6445480621147808,
+      "learning_rate": 5.2160402989988475e-06,
+      "loss": 0.8551,
+      "step": 7838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0913669711625023,
+      "learning_rate": 5.215004775160608e-06,
+      "loss": 0.8843,
+      "step": 7839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9425730450815155,
+      "learning_rate": 5.2139692420831325e-06,
+      "loss": 0.861,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2565320136722566,
+      "learning_rate": 5.212933699810924e-06,
+      "loss": 0.8452,
+      "step": 7841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.861543110114202,
+      "learning_rate": 5.211898148388481e-06,
+      "loss": 0.8722,
+      "step": 7842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4603030751542576,
+      "learning_rate": 5.2108625878603e-06,
+      "loss": 0.8728,
+      "step": 7843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.86155890002565,
+      "learning_rate": 5.209827018270886e-06,
+      "loss": 0.895,
+      "step": 7844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8582599411009058,
+      "learning_rate": 5.208791439664736e-06,
+      "loss": 0.6926,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2215009156603864,
+      "learning_rate": 5.207755852086355e-06,
+      "loss": 0.9633,
+      "step": 7846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.092426798129594,
+      "learning_rate": 5.206720255580241e-06,
+      "loss": 0.8153,
+      "step": 7847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2030054030698123,
+      "learning_rate": 5.2056846501908975e-06,
+      "loss": 0.7768,
+      "step": 7848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9026062934327075,
+      "learning_rate": 5.204649035962825e-06,
+      "loss": 0.912,
+      "step": 7849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8109542148315265,
+      "learning_rate": 5.2036134129405305e-06,
+      "loss": 0.8636,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.335259301383186,
+      "learning_rate": 5.202577781168513e-06,
+      "loss": 0.8862,
+      "step": 7851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.084967381057004,
+      "learning_rate": 5.201542140691277e-06,
+      "loss": 0.6561,
+      "step": 7852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0865972252487777,
+      "learning_rate": 5.200506491553326e-06,
+      "loss": 0.9039,
+      "step": 7853
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.318705392997824,
+      "learning_rate": 5.199470833799164e-06,
+      "loss": 0.7381,
+      "step": 7854
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3253816746892264,
+      "learning_rate": 5.198435167473297e-06,
+      "loss": 0.8256,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3538670804696293,
+      "learning_rate": 5.1973994926202296e-06,
+      "loss": 0.9596,
+      "step": 7856
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9799008927287762,
+      "learning_rate": 5.196363809284466e-06,
+      "loss": 0.7941,
+      "step": 7857
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2868407877925847,
+      "learning_rate": 5.19532811751051e-06,
+      "loss": 0.8357,
+      "step": 7858
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2687730738267813,
+      "learning_rate": 5.1942924173428725e-06,
+      "loss": 0.7286,
+      "step": 7859
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1544027671233734,
+      "learning_rate": 5.193256708826055e-06,
+      "loss": 1.0365,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4842798662396883,
+      "learning_rate": 5.192220992004569e-06,
+      "loss": 0.7968,
+      "step": 7861
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.225592401445469,
+      "learning_rate": 5.191185266922917e-06,
+      "loss": 0.7298,
+      "step": 7862
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.7873766337889005,
+      "learning_rate": 5.190149533625607e-06,
+      "loss": 0.712,
+      "step": 7863
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2151952173067637,
+      "learning_rate": 5.18911379215715e-06,
+      "loss": 0.8217,
+      "step": 7864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1211958255213057,
+      "learning_rate": 5.188078042562052e-06,
+      "loss": 0.9084,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5604943202321206,
+      "learning_rate": 5.187042284884821e-06,
+      "loss": 0.8988,
+      "step": 7866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9677022278502152,
+      "learning_rate": 5.186006519169967e-06,
+      "loss": 0.8027,
+      "step": 7867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.103401083002649,
+      "learning_rate": 5.184970745461998e-06,
+      "loss": 0.7885,
+      "step": 7868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7621026408421143,
+      "learning_rate": 5.1839349638054245e-06,
+      "loss": 0.7818,
+      "step": 7869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7139993189571743,
+      "learning_rate": 5.182899174244757e-06,
+      "loss": 0.9496,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2215773463035322,
+      "learning_rate": 5.181863376824504e-06,
+      "loss": 0.7461,
+      "step": 7871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.277886476958112,
+      "learning_rate": 5.1808275715891756e-06,
+      "loss": 0.8253,
+      "step": 7872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.176555693347368,
+      "learning_rate": 5.1797917585832846e-06,
+      "loss": 0.7742,
+      "step": 7873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9888330000661345,
+      "learning_rate": 5.178755937851341e-06,
+      "loss": 0.8168,
+      "step": 7874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4976340414077414,
+      "learning_rate": 5.177720109437857e-06,
+      "loss": 0.7067,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.813653239546763,
+      "learning_rate": 5.176684273387344e-06,
+      "loss": 0.8905,
+      "step": 7876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8674955091113068,
+      "learning_rate": 5.1756484297443135e-06,
+      "loss": 0.8227,
+      "step": 7877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0676838594071785,
+      "learning_rate": 5.1746125785532795e-06,
+      "loss": 0.9576,
+      "step": 7878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1571572064396827,
+      "learning_rate": 5.173576719858755e-06,
+      "loss": 0.7373,
+      "step": 7879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1047715931709265,
+      "learning_rate": 5.172540853705251e-06,
+      "loss": 0.7951,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.75605090231729,
+      "learning_rate": 5.171504980137282e-06,
+      "loss": 0.9258,
+      "step": 7881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9786699908331116,
+      "learning_rate": 5.170469099199363e-06,
+      "loss": 0.6746,
+      "step": 7882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.3854255808660305,
+      "learning_rate": 5.169433210936007e-06,
+      "loss": 0.8249,
+      "step": 7883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.720313064768659,
+      "learning_rate": 5.168397315391729e-06,
+      "loss": 0.5868,
+      "step": 7884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.194361562989058,
+      "learning_rate": 5.167361412611042e-06,
+      "loss": 0.7833,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8839844620248725,
+      "learning_rate": 5.1663255026384625e-06,
+      "loss": 0.9087,
+      "step": 7886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.803846650660823,
+      "learning_rate": 5.165289585518507e-06,
+      "loss": 0.6491,
+      "step": 7887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.426234141610359,
+      "learning_rate": 5.164253661295689e-06,
+      "loss": 0.737,
+      "step": 7888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.047702975109114,
+      "learning_rate": 5.1632177300145255e-06,
+      "loss": 0.8642,
+      "step": 7889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0272350358843463,
+      "learning_rate": 5.162181791719532e-06,
+      "loss": 0.8849,
+      "step": 7890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.874140012138592,
+      "learning_rate": 5.161145846455227e-06,
+      "loss": 0.8266,
+      "step": 7891
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2205763957870013,
+      "learning_rate": 5.160109894266125e-06,
+      "loss": 0.806,
+      "step": 7892
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1095581005146333,
+      "learning_rate": 5.159073935196746e-06,
+      "loss": 0.6069,
+      "step": 7893
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.918862902171845,
+      "learning_rate": 5.1580379692916025e-06,
+      "loss": 0.839,
+      "step": 7894
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0375757982793177,
+      "learning_rate": 5.157001996595218e-06,
+      "loss": 0.6984,
+      "step": 7895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3014720613371904,
+      "learning_rate": 5.155966017152108e-06,
+      "loss": 0.8171,
+      "step": 7896
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2092730042539004,
+      "learning_rate": 5.1549300310067896e-06,
+      "loss": 0.7667,
+      "step": 7897
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0211344112113787,
+      "learning_rate": 5.153894038203785e-06,
+      "loss": 0.7908,
+      "step": 7898
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.244422877338181,
+      "learning_rate": 5.152858038787608e-06,
+      "loss": 0.7266,
+      "step": 7899
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4286190672976007,
+      "learning_rate": 5.151822032802782e-06,
+      "loss": 0.7084,
+      "step": 7900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.071424160194338,
+      "learning_rate": 5.150786020293824e-06,
+      "loss": 0.9282,
+      "step": 7901
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.275179190392393,
+      "learning_rate": 5.149750001305257e-06,
+      "loss": 0.5756,
+      "step": 7902
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9740420534846939,
+      "learning_rate": 5.148713975881598e-06,
+      "loss": 0.8064,
+      "step": 7903
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2760774948230904,
+      "learning_rate": 5.147677944067368e-06,
+      "loss": 0.8158,
+      "step": 7904
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0327806524952803,
+      "learning_rate": 5.146641905907088e-06,
+      "loss": 0.6538,
+      "step": 7905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8906099746285494,
+      "learning_rate": 5.145605861445279e-06,
+      "loss": 0.8476,
+      "step": 7906
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5574984505146796,
+      "learning_rate": 5.144569810726462e-06,
+      "loss": 0.803,
+      "step": 7907
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.358569607898119,
+      "learning_rate": 5.143533753795157e-06,
+      "loss": 0.8976,
+      "step": 7908
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0436827245826574,
+      "learning_rate": 5.142497690695888e-06,
+      "loss": 0.6934,
+      "step": 7909
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6303303074524047,
+      "learning_rate": 5.141461621473175e-06,
+      "loss": 0.7886,
+      "step": 7910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1768472515201025,
+      "learning_rate": 5.140425546171542e-06,
+      "loss": 0.8429,
+      "step": 7911
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4310415161566197,
+      "learning_rate": 5.13938946483551e-06,
+      "loss": 0.8214,
+      "step": 7912
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4730457820558085,
+      "learning_rate": 5.1383533775096025e-06,
+      "loss": 0.8873,
+      "step": 7913
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.621488681413786,
+      "learning_rate": 5.137317284238344e-06,
+      "loss": 0.6865,
+      "step": 7914
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.091197288453661,
+      "learning_rate": 5.136281185066253e-06,
+      "loss": 0.8687,
+      "step": 7915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5343194948236594,
+      "learning_rate": 5.135245080037857e-06,
+      "loss": 0.9205,
+      "step": 7916
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.412410595114181,
+      "learning_rate": 5.1342089691976794e-06,
+      "loss": 0.8949,
+      "step": 7917
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3461362988736822,
+      "learning_rate": 5.1331728525902435e-06,
+      "loss": 0.956,
+      "step": 7918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.482301825922856,
+      "learning_rate": 5.1321367302600726e-06,
+      "loss": 0.8428,
+      "step": 7919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.012151425370043,
+      "learning_rate": 5.131100602251692e-06,
+      "loss": 0.6778,
+      "step": 7920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9473509583756765,
+      "learning_rate": 5.130064468609627e-06,
+      "loss": 0.5623,
+      "step": 7921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.969311931930358,
+      "learning_rate": 5.129028329378403e-06,
+      "loss": 0.8533,
+      "step": 7922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1709956473701135,
+      "learning_rate": 5.127992184602542e-06,
+      "loss": 0.8873,
+      "step": 7923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.909484627768436,
+      "learning_rate": 5.126956034326573e-06,
+      "loss": 1.0548,
+      "step": 7924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.911551761259335,
+      "learning_rate": 5.125919878595022e-06,
+      "loss": 0.774,
+      "step": 7925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.880824106403495,
+      "learning_rate": 5.124883717452412e-06,
+      "loss": 0.979,
+      "step": 7926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8732899372950778,
+      "learning_rate": 5.123847550943272e-06,
+      "loss": 0.7978,
+      "step": 7927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4594005497001414,
+      "learning_rate": 5.122811379112124e-06,
+      "loss": 0.9025,
+      "step": 7928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.649188477439036,
+      "learning_rate": 5.121775202003499e-06,
+      "loss": 0.8931,
+      "step": 7929
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0474266520426683,
+      "learning_rate": 5.120739019661923e-06,
+      "loss": 0.6567,
+      "step": 7930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0513098291432463,
+      "learning_rate": 5.119702832131922e-06,
+      "loss": 1.1859,
+      "step": 7931
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5777893696585465,
+      "learning_rate": 5.118666639458023e-06,
+      "loss": 0.7585,
+      "step": 7932
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.692402540639146,
+      "learning_rate": 5.117630441684755e-06,
+      "loss": 0.8616,
+      "step": 7933
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.039488699634753,
+      "learning_rate": 5.116594238856645e-06,
+      "loss": 0.78,
+      "step": 7934
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3184782802856363,
+      "learning_rate": 5.115558031018221e-06,
+      "loss": 0.8376,
+      "step": 7935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0507191757647316,
+      "learning_rate": 5.114521818214011e-06,
+      "loss": 0.8692,
+      "step": 7936
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7469773483207456,
+      "learning_rate": 5.113485600488541e-06,
+      "loss": 0.7584,
+      "step": 7937
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6866278245985504,
+      "learning_rate": 5.112449377886345e-06,
+      "loss": 0.9065,
+      "step": 7938
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.608219779514243,
+      "learning_rate": 5.111413150451948e-06,
+      "loss": 0.8933,
+      "step": 7939
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3366155022199266,
+      "learning_rate": 5.11037691822988e-06,
+      "loss": 0.8705,
+      "step": 7940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9461968051877467,
+      "learning_rate": 5.10934068126467e-06,
+      "loss": 0.6842,
+      "step": 7941
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8947331274019468,
+      "learning_rate": 5.108304439600847e-06,
+      "loss": 0.7659,
+      "step": 7942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2362999609860663,
+      "learning_rate": 5.107268193282941e-06,
+      "loss": 0.6234,
+      "step": 7943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4000743136226115,
+      "learning_rate": 5.1062319423554815e-06,
+      "loss": 0.9037,
+      "step": 7944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.220531018546661,
+      "learning_rate": 5.105195686863e-06,
+      "loss": 0.8258,
+      "step": 7945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9894806352076129,
+      "learning_rate": 5.104159426850024e-06,
+      "loss": 0.663,
+      "step": 7946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1287937169685156,
+      "learning_rate": 5.103123162361087e-06,
+      "loss": 0.7272,
+      "step": 7947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.28171111318743,
+      "learning_rate": 5.102086893440717e-06,
+      "loss": 0.7652,
+      "step": 7948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9572684997271281,
+      "learning_rate": 5.101050620133447e-06,
+      "loss": 0.7347,
+      "step": 7949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2674996170684723,
+      "learning_rate": 5.100014342483804e-06,
+      "loss": 0.7011,
+      "step": 7950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9713034458271153,
+      "learning_rate": 5.098978060536325e-06,
+      "loss": 0.8122,
+      "step": 7951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9502916159948115,
+      "learning_rate": 5.097941774335538e-06,
+      "loss": 0.6886,
+      "step": 7952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1959935309977343,
+      "learning_rate": 5.096905483925972e-06,
+      "loss": 0.6501,
+      "step": 7953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2215232718471625,
+      "learning_rate": 5.095869189352166e-06,
+      "loss": 0.9677,
+      "step": 7954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.218622970680372,
+      "learning_rate": 5.094832890658644e-06,
+      "loss": 0.7153,
+      "step": 7955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.459898298109667,
+      "learning_rate": 5.0937965878899435e-06,
+      "loss": 0.6864,
+      "step": 7956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2435239613451738,
+      "learning_rate": 5.092760281090592e-06,
+      "loss": 0.6604,
+      "step": 7957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1722945254681383,
+      "learning_rate": 5.091723970305128e-06,
+      "loss": 0.7699,
+      "step": 7958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2149327065503015,
+      "learning_rate": 5.090687655578078e-06,
+      "loss": 0.8661,
+      "step": 7959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.787630668950538,
+      "learning_rate": 5.089651336953979e-06,
+      "loss": 0.9804,
+      "step": 7960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3413123274417234,
+      "learning_rate": 5.088615014477362e-06,
+      "loss": 0.8256,
+      "step": 7961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.639580092958613,
+      "learning_rate": 5.087578688192759e-06,
+      "loss": 0.7033,
+      "step": 7962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9386081806213231,
+      "learning_rate": 5.086542358144708e-06,
+      "loss": 0.5873,
+      "step": 7963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.137681786729451,
+      "learning_rate": 5.0855060243777366e-06,
+      "loss": 0.7075,
+      "step": 7964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1032911306344029,
+      "learning_rate": 5.084469686936382e-06,
+      "loss": 0.7779,
+      "step": 7965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8325247892527785,
+      "learning_rate": 5.083433345865175e-06,
+      "loss": 0.8545,
+      "step": 7966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.040551286659769,
+      "learning_rate": 5.082397001208653e-06,
+      "loss": 0.9607,
+      "step": 7967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4238864818314148,
+      "learning_rate": 5.081360653011349e-06,
+      "loss": 0.8454,
+      "step": 7968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.940323641327688,
+      "learning_rate": 5.080324301317795e-06,
+      "loss": 1.0599,
+      "step": 7969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.890948848204059,
+      "learning_rate": 5.079287946172529e-06,
+      "loss": 0.7311,
+      "step": 7970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6848928080677155,
+      "learning_rate": 5.07825158762008e-06,
+      "loss": 0.829,
+      "step": 7971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7161173491149038,
+      "learning_rate": 5.077215225704988e-06,
+      "loss": 0.9367,
+      "step": 7972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0398916531704612,
+      "learning_rate": 5.076178860471787e-06,
+      "loss": 0.7033,
+      "step": 7973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9992520153513118,
+      "learning_rate": 5.0751424919650085e-06,
+      "loss": 0.7316,
+      "step": 7974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.00692067582576,
+      "learning_rate": 5.07410612022919e-06,
+      "loss": 0.829,
+      "step": 7975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.159114520729551,
+      "learning_rate": 5.073069745308868e-06,
+      "loss": 0.601,
+      "step": 7976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0340252297014991,
+      "learning_rate": 5.072033367248574e-06,
+      "loss": 0.6829,
+      "step": 7977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.976944433279959,
+      "learning_rate": 5.0709969860928485e-06,
+      "loss": 0.8472,
+      "step": 7978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.508861158665068,
+      "learning_rate": 5.069960601886224e-06,
+      "loss": 1.0189,
+      "step": 7979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8792906834749574,
+      "learning_rate": 5.068924214673234e-06,
+      "loss": 0.811,
+      "step": 7980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2529626704174004,
+      "learning_rate": 5.067887824498419e-06,
+      "loss": 0.7968,
+      "step": 7981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1411925792975735,
+      "learning_rate": 5.0668514314063135e-06,
+      "loss": 0.7042,
+      "step": 7982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9497460963731992,
+      "learning_rate": 5.065815035441452e-06,
+      "loss": 0.8053,
+      "step": 7983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2920383067751,
+      "learning_rate": 5.064778636648371e-06,
+      "loss": 0.657,
+      "step": 7984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7569405078815636,
+      "learning_rate": 5.063742235071609e-06,
+      "loss": 0.9732,
+      "step": 7985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7605241644400502,
+      "learning_rate": 5.062705830755702e-06,
+      "loss": 0.6743,
+      "step": 7986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.986599793723074,
+      "learning_rate": 5.061669423745185e-06,
+      "loss": 0.8387,
+      "step": 7987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.559609751290006,
+      "learning_rate": 5.060633014084595e-06,
+      "loss": 0.9826,
+      "step": 7988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4719029171594102,
+      "learning_rate": 5.05959660181847e-06,
+      "loss": 0.8525,
+      "step": 7989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.210584420167013,
+      "learning_rate": 5.058560186991346e-06,
+      "loss": 0.8196,
+      "step": 7990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2536248398862826,
+      "learning_rate": 5.0575237696477596e-06,
+      "loss": 0.7956,
+      "step": 7991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4931790849738382,
+      "learning_rate": 5.05648734983225e-06,
+      "loss": 0.8027,
+      "step": 7992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.216214411974278,
+      "learning_rate": 5.055450927589352e-06,
+      "loss": 0.7785,
+      "step": 7993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.276713045086562,
+      "learning_rate": 5.054414502963605e-06,
+      "loss": 0.8573,
+      "step": 7994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9046094251550365,
+      "learning_rate": 5.053378075999545e-06,
+      "loss": 0.7165,
+      "step": 7995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.654416643215205,
+      "learning_rate": 5.052341646741711e-06,
+      "loss": 0.9175,
+      "step": 7996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.095582248112477,
+      "learning_rate": 5.05130521523464e-06,
+      "loss": 0.7126,
+      "step": 7997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.163770656593186,
+      "learning_rate": 5.050268781522868e-06,
+      "loss": 0.8951,
+      "step": 7998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1808124588895885,
+      "learning_rate": 5.049232345650936e-06,
+      "loss": 0.7215,
+      "step": 7999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.334535860107792,
+      "learning_rate": 5.04819590766338e-06,
+      "loss": 0.7139,
+      "step": 8000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0487060229988145,
+      "learning_rate": 5.0471594676047385e-06,
+      "loss": 0.8133,
+      "step": 8001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8556144662045235,
+      "learning_rate": 5.0461230255195495e-06,
+      "loss": 0.5734,
+      "step": 8002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8064953339912864,
+      "learning_rate": 5.045086581452352e-06,
+      "loss": 0.7876,
+      "step": 8003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.492939441191677,
+      "learning_rate": 5.044050135447682e-06,
+      "loss": 0.7287,
+      "step": 8004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9867908602319388,
+      "learning_rate": 5.043013687550083e-06,
+      "loss": 0.8449,
+      "step": 8005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0907571511278245,
+      "learning_rate": 5.041977237804088e-06,
+      "loss": 0.6619,
+      "step": 8006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1031830233129585,
+      "learning_rate": 5.040940786254236e-06,
+      "loss": 0.6953,
+      "step": 8007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.915670334819667,
+      "learning_rate": 5.039904332945069e-06,
+      "loss": 0.832,
+      "step": 8008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.289821737182655,
+      "learning_rate": 5.038867877921124e-06,
+      "loss": 0.7041,
+      "step": 8009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8246313090720387,
+      "learning_rate": 5.037831421226939e-06,
+      "loss": 0.7516,
+      "step": 8010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.433372322670569,
+      "learning_rate": 5.036794962907054e-06,
+      "loss": 0.763,
+      "step": 8011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.73284158006818,
+      "learning_rate": 5.035758503006007e-06,
+      "loss": 0.763,
+      "step": 8012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0165473874793176,
+      "learning_rate": 5.034722041568337e-06,
+      "loss": 1.0418,
+      "step": 8013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8253798916141153,
+      "learning_rate": 5.033685578638586e-06,
+      "loss": 0.9552,
+      "step": 8014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1727949289429245,
+      "learning_rate": 5.03264911426129e-06,
+      "loss": 0.9097,
+      "step": 8015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.247476246967929,
+      "learning_rate": 5.031612648480987e-06,
+      "loss": 0.8263,
+      "step": 8016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.865853372989161,
+      "learning_rate": 5.03057618134222e-06,
+      "loss": 0.9129,
+      "step": 8017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1460793845309625,
+      "learning_rate": 5.029539712889524e-06,
+      "loss": 0.6859,
+      "step": 8018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9305072950410622,
+      "learning_rate": 5.028503243167443e-06,
+      "loss": 0.8786,
+      "step": 8019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.027267743331633,
+      "learning_rate": 5.027466772220513e-06,
+      "loss": 0.884,
+      "step": 8020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3831746698745246,
+      "learning_rate": 5.026430300093275e-06,
+      "loss": 0.8679,
+      "step": 8021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9156795432329288,
+      "learning_rate": 5.025393826830267e-06,
+      "loss": 0.9729,
+      "step": 8022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0894776213196993,
+      "learning_rate": 5.024357352476032e-06,
+      "loss": 0.9571,
+      "step": 8023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8728931966663112,
+      "learning_rate": 5.023320877075107e-06,
+      "loss": 0.8285,
+      "step": 8024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.049003412141986,
+      "learning_rate": 5.02228440067203e-06,
+      "loss": 0.968,
+      "step": 8025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5498639717462286,
+      "learning_rate": 5.0212479233113445e-06,
+      "loss": 0.7428,
+      "step": 8026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.82575781908162,
+      "learning_rate": 5.020211445037587e-06,
+      "loss": 0.8383,
+      "step": 8027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.408692274038712,
+      "learning_rate": 5.019174965895299e-06,
+      "loss": 0.773,
+      "step": 8028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0345839210133303,
+      "learning_rate": 5.0181384859290215e-06,
+      "loss": 0.6991,
+      "step": 8029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6872503404715076,
+      "learning_rate": 5.017102005183292e-06,
+      "loss": 0.9925,
+      "step": 8030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1350058986352076,
+      "learning_rate": 5.01606552370265e-06,
+      "loss": 0.8104,
+      "step": 8031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.514074695935536,
+      "learning_rate": 5.015029041531639e-06,
+      "loss": 0.8182,
+      "step": 8032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.841710255447695,
+      "learning_rate": 5.013992558714795e-06,
+      "loss": 0.8865,
+      "step": 8033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5712704865209304,
+      "learning_rate": 5.01295607529666e-06,
+      "loss": 0.7394,
+      "step": 8034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9069668720295057,
+      "learning_rate": 5.011919591321773e-06,
+      "loss": 0.8602,
+      "step": 8035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1838672995972446,
+      "learning_rate": 5.010883106834676e-06,
+      "loss": 0.8957,
+      "step": 8036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0289173716985474,
+      "learning_rate": 5.009846621879906e-06,
+      "loss": 0.6226,
+      "step": 8037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7115432760031775,
+      "learning_rate": 5.008810136502006e-06,
+      "loss": 0.5981,
+      "step": 8038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3482016151606224,
+      "learning_rate": 5.007773650745514e-06,
+      "loss": 0.7799,
+      "step": 8039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1468593689021915,
+      "learning_rate": 5.00673716465497e-06,
+      "loss": 0.808,
+      "step": 8040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8985344885086648,
+      "learning_rate": 5.005700678274918e-06,
+      "loss": 0.7213,
+      "step": 8041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1831514631565456,
+      "learning_rate": 5.004664191649893e-06,
+      "loss": 0.6389,
+      "step": 8042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2512908412914192,
+      "learning_rate": 5.003627704824438e-06,
+      "loss": 0.7265,
+      "step": 8043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.347973952324506,
+      "learning_rate": 5.0025912178430925e-06,
+      "loss": 0.92,
+      "step": 8044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.083893785504998,
+      "learning_rate": 5.001554730750395e-06,
+      "loss": 0.8344,
+      "step": 8045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1102181765777606,
+      "learning_rate": 5.0005182435908885e-06,
+      "loss": 0.721,
+      "step": 8046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7215692074018185,
+      "learning_rate": 4.999481756409112e-06,
+      "loss": 0.8229,
+      "step": 8047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.52036489938483,
+      "learning_rate": 4.998445269249606e-06,
+      "loss": 1.0338,
+      "step": 8048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.182249140412354,
+      "learning_rate": 4.997408782156909e-06,
+      "loss": 0.7068,
+      "step": 8049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0861922304384506,
+      "learning_rate": 4.996372295175563e-06,
+      "loss": 0.8706,
+      "step": 8050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1768319936443068,
+      "learning_rate": 4.995335808350109e-06,
+      "loss": 0.6667,
+      "step": 8051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1980664550636084,
+      "learning_rate": 4.994299321725085e-06,
+      "loss": 0.9394,
+      "step": 8052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.025905088113314,
+      "learning_rate": 4.99326283534503e-06,
+      "loss": 0.9525,
+      "step": 8053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.362330760899771,
+      "learning_rate": 4.9922263492544885e-06,
+      "loss": 0.813,
+      "step": 8054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2002215025464094,
+      "learning_rate": 4.991189863497996e-06,
+      "loss": 0.9319,
+      "step": 8055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.11760599537595,
+      "learning_rate": 4.990153378120096e-06,
+      "loss": 0.6824,
+      "step": 8056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8758394261223703,
+      "learning_rate": 4.989116893165325e-06,
+      "loss": 0.5129,
+      "step": 8057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.578098287697007,
+      "learning_rate": 4.988080408678229e-06,
+      "loss": 0.7115,
+      "step": 8058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1723143210291258,
+      "learning_rate": 4.987043924703342e-06,
+      "loss": 0.7467,
+      "step": 8059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0014454223912115,
+      "learning_rate": 4.986007441285206e-06,
+      "loss": 0.636,
+      "step": 8060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1607051345401636,
+      "learning_rate": 4.984970958468364e-06,
+      "loss": 0.6175,
+      "step": 8061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5855460986852132,
+      "learning_rate": 4.983934476297351e-06,
+      "loss": 0.8193,
+      "step": 8062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.167218026174501,
+      "learning_rate": 4.9828979948167095e-06,
+      "loss": 0.797,
+      "step": 8063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0839958839264103,
+      "learning_rate": 4.981861514070979e-06,
+      "loss": 0.6747,
+      "step": 8064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5763064102428603,
+      "learning_rate": 4.980825034104702e-06,
+      "loss": 0.8369,
+      "step": 8065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.368466280109985,
+      "learning_rate": 4.979788554962415e-06,
+      "loss": 0.828,
+      "step": 8066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8361996134074006,
+      "learning_rate": 4.978752076688657e-06,
+      "loss": 0.7135,
+      "step": 8067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9192216478043576,
+      "learning_rate": 4.97771559932797e-06,
+      "loss": 0.8846,
+      "step": 8068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0991917613597575,
+      "learning_rate": 4.976679122924896e-06,
+      "loss": 0.7051,
+      "step": 8069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.22998626922305,
+      "learning_rate": 4.97564264752397e-06,
+      "loss": 0.8592,
+      "step": 8070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.974646820160656,
+      "learning_rate": 4.974606173169733e-06,
+      "loss": 0.5972,
+      "step": 8071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.916455508087424,
+      "learning_rate": 4.973569699906727e-06,
+      "loss": 0.8371,
+      "step": 8072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0962963448721075,
+      "learning_rate": 4.972533227779489e-06,
+      "loss": 0.8551,
+      "step": 8073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.815667082312754,
+      "learning_rate": 4.971496756832557e-06,
+      "loss": 0.7686,
+      "step": 8074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5588362774787488,
+      "learning_rate": 4.9704602871104765e-06,
+      "loss": 0.7363,
+      "step": 8075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.240163296082111,
+      "learning_rate": 4.969423818657783e-06,
+      "loss": 0.9445,
+      "step": 8076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5855074181938766,
+      "learning_rate": 4.9683873515190145e-06,
+      "loss": 0.933,
+      "step": 8077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0857729561273,
+      "learning_rate": 4.9673508857387115e-06,
+      "loss": 0.7444,
+      "step": 8078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2612131737904915,
+      "learning_rate": 4.966314421361416e-06,
+      "loss": 0.9554,
+      "step": 8079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.001982090784672,
+      "learning_rate": 4.965277958431664e-06,
+      "loss": 0.812,
+      "step": 8080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0455119863464497,
+      "learning_rate": 4.964241496993994e-06,
+      "loss": 0.7132,
+      "step": 8081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1452949605499465,
+      "learning_rate": 4.963205037092947e-06,
+      "loss": 0.8334,
+      "step": 8082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4008506793651905,
+      "learning_rate": 4.962168578773063e-06,
+      "loss": 0.8694,
+      "step": 8083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.659769147007191,
+      "learning_rate": 4.9611321220788775e-06,
+      "loss": 0.8198,
+      "step": 8084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4804851996818136,
+      "learning_rate": 4.9600956670549324e-06,
+      "loss": 0.7144,
+      "step": 8085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.245622863141485,
+      "learning_rate": 4.959059213745764e-06,
+      "loss": 0.7417,
+      "step": 8086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3541118074243745,
+      "learning_rate": 4.9580227621959145e-06,
+      "loss": 0.7836,
+      "step": 8087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9868724253974088,
+      "learning_rate": 4.956986312449919e-06,
+      "loss": 0.8484,
+      "step": 8088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3384662563657599,
+      "learning_rate": 4.955949864552318e-06,
+      "loss": 0.7935,
+      "step": 8089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.87384590973909,
+      "learning_rate": 4.954913418547651e-06,
+      "loss": 0.815,
+      "step": 8090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0091115840733895,
+      "learning_rate": 4.953876974480452e-06,
+      "loss": 0.8181,
+      "step": 8091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2783636462796504,
+      "learning_rate": 4.952840532395262e-06,
+      "loss": 0.9208,
+      "step": 8092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.003889561667185,
+      "learning_rate": 4.951804092336621e-06,
+      "loss": 0.8585,
+      "step": 8093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2358714465090008,
+      "learning_rate": 4.950767654349067e-06,
+      "loss": 0.696,
+      "step": 8094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.316525219036258,
+      "learning_rate": 4.949731218477133e-06,
+      "loss": 0.7721,
+      "step": 8095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.759727488915186,
+      "learning_rate": 4.9486947847653615e-06,
+      "loss": 0.9336,
+      "step": 8096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1979985554928376,
+      "learning_rate": 4.947658353258291e-06,
+      "loss": 0.6981,
+      "step": 8097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.1035420074124005,
+      "learning_rate": 4.946621924000456e-06,
+      "loss": 0.8394,
+      "step": 8098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1108498137817544,
+      "learning_rate": 4.945585497036396e-06,
+      "loss": 0.7349,
+      "step": 8099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.700710718507287,
+      "learning_rate": 4.944549072410648e-06,
+      "loss": 0.8829,
+      "step": 8100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.069064448025905,
+      "learning_rate": 4.9435126501677525e-06,
+      "loss": 0.8108,
+      "step": 8101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.448888950683515,
+      "learning_rate": 4.942476230352241e-06,
+      "loss": 0.7947,
+      "step": 8102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.332395044737338,
+      "learning_rate": 4.941439813008656e-06,
+      "loss": 0.767,
+      "step": 8103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2136349318172606,
+      "learning_rate": 4.940403398181531e-06,
+      "loss": 0.8609,
+      "step": 8104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.895651050753387,
+      "learning_rate": 4.939366985915408e-06,
+      "loss": 0.8577,
+      "step": 8105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6468521965504475,
+      "learning_rate": 4.938330576254817e-06,
+      "loss": 0.7242,
+      "step": 8106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.768270061985266,
+      "learning_rate": 4.9372941692443e-06,
+      "loss": 0.8191,
+      "step": 8107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.130216603047419,
+      "learning_rate": 4.936257764928392e-06,
+      "loss": 0.7497,
+      "step": 8108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2711220589988494,
+      "learning_rate": 4.935221363351631e-06,
+      "loss": 0.8637,
+      "step": 8109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.348589472674517,
+      "learning_rate": 4.934184964558549e-06,
+      "loss": 0.7459,
+      "step": 8110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.792347572516863,
+      "learning_rate": 4.933148568593687e-06,
+      "loss": 0.8737,
+      "step": 8111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8547833763653219,
+      "learning_rate": 4.932112175501583e-06,
+      "loss": 0.7964,
+      "step": 8112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.327767586681265,
+      "learning_rate": 4.931075785326767e-06,
+      "loss": 0.9278,
+      "step": 8113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.089022943444434,
+      "learning_rate": 4.930039398113779e-06,
+      "loss": 0.9137,
+      "step": 8114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9380001516481749,
+      "learning_rate": 4.929003013907152e-06,
+      "loss": 1.0268,
+      "step": 8115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.543351561749796,
+      "learning_rate": 4.927966632751427e-06,
+      "loss": 1.0177,
+      "step": 8116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1349680713428687,
+      "learning_rate": 4.926930254691134e-06,
+      "loss": 0.8207,
+      "step": 8117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.904443766835419,
+      "learning_rate": 4.92589387977081e-06,
+      "loss": 0.8676,
+      "step": 8118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2456995470154166,
+      "learning_rate": 4.924857508034994e-06,
+      "loss": 0.8903,
+      "step": 8119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.222358924074969,
+      "learning_rate": 4.9238211395282156e-06,
+      "loss": 0.8386,
+      "step": 8120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0223672314938423,
+      "learning_rate": 4.922784774295013e-06,
+      "loss": 0.8347,
+      "step": 8121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1495033510008712,
+      "learning_rate": 4.92174841237992e-06,
+      "loss": 0.6671,
+      "step": 8122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1318461694932875,
+      "learning_rate": 4.920712053827475e-06,
+      "loss": 0.715,
+      "step": 8123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.117252088500184,
+      "learning_rate": 4.919675698682206e-06,
+      "loss": 0.7496,
+      "step": 8124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0127582964910462,
+      "learning_rate": 4.918639346988652e-06,
+      "loss": 0.5792,
+      "step": 8125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.327409726225018,
+      "learning_rate": 4.917602998791348e-06,
+      "loss": 0.912,
+      "step": 8126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.460550021738644,
+      "learning_rate": 4.9165666541348265e-06,
+      "loss": 0.9534,
+      "step": 8127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9180218259462825,
+      "learning_rate": 4.91553031306362e-06,
+      "loss": 0.8131,
+      "step": 8128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0148170380284984,
+      "learning_rate": 4.914493975622263e-06,
+      "loss": 0.867,
+      "step": 8129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4941611799932124,
+      "learning_rate": 4.9134576418552945e-06,
+      "loss": 0.7295,
+      "step": 8130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6765063804088622,
+      "learning_rate": 4.9124213118072415e-06,
+      "loss": 0.6114,
+      "step": 8131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1113805514409054,
+      "learning_rate": 4.911384985522639e-06,
+      "loss": 0.7124,
+      "step": 8132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.745514656217285,
+      "learning_rate": 4.910348663046021e-06,
+      "loss": 0.7616,
+      "step": 8133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.181261292120689,
+      "learning_rate": 4.909312344421923e-06,
+      "loss": 0.8753,
+      "step": 8134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0716993256346825,
+      "learning_rate": 4.908276029694873e-06,
+      "loss": 0.7474,
+      "step": 8135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9884104209404663,
+      "learning_rate": 4.907239718909408e-06,
+      "loss": 1.0161,
+      "step": 8136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3335869752989318,
+      "learning_rate": 4.90620341211006e-06,
+      "loss": 0.8426,
+      "step": 8137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3512647282198627,
+      "learning_rate": 4.905167109341357e-06,
+      "loss": 0.8128,
+      "step": 8138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8852982920317847,
+      "learning_rate": 4.904130810647836e-06,
+      "loss": 0.7741,
+      "step": 8139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.451275413226546,
+      "learning_rate": 4.903094516074027e-06,
+      "loss": 0.6767,
+      "step": 8140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.964979938549868,
+      "learning_rate": 4.902058225664465e-06,
+      "loss": 0.9042,
+      "step": 8141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8275978580284122,
+      "learning_rate": 4.901021939463676e-06,
+      "loss": 0.8781,
+      "step": 8142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3975017377582617,
+      "learning_rate": 4.899985657516195e-06,
+      "loss": 0.8599,
+      "step": 8143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6977207994351966,
+      "learning_rate": 4.898949379866556e-06,
+      "loss": 0.7962,
+      "step": 8144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4629263995452915,
+      "learning_rate": 4.897913106559285e-06,
+      "loss": 0.7577,
+      "step": 8145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7620673332209633,
+      "learning_rate": 4.896876837638915e-06,
+      "loss": 0.9286,
+      "step": 8146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.659870934751098,
+      "learning_rate": 4.895840573149977e-06,
+      "loss": 0.8273,
+      "step": 8147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9450946218908285,
+      "learning_rate": 4.8948043131370025e-06,
+      "loss": 0.8646,
+      "step": 8148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.910171476068765,
+      "learning_rate": 4.893768057644519e-06,
+      "loss": 0.7756,
+      "step": 8149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4696738771885753,
+      "learning_rate": 4.89273180671706e-06,
+      "loss": 0.9385,
+      "step": 8150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.830177949650449,
+      "learning_rate": 4.891695560399154e-06,
+      "loss": 0.8219,
+      "step": 8151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2723095469574055,
+      "learning_rate": 4.890659318735333e-06,
+      "loss": 0.8623,
+      "step": 8152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.006079450670445,
+      "learning_rate": 4.8896230817701214e-06,
+      "loss": 0.8002,
+      "step": 8153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6832873982555265,
+      "learning_rate": 4.888586849548053e-06,
+      "loss": 0.5674,
+      "step": 8154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6178377315634527,
+      "learning_rate": 4.887550622113657e-06,
+      "loss": 0.9711,
+      "step": 8155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5559779046145197,
+      "learning_rate": 4.88651439951146e-06,
+      "loss": 0.7662,
+      "step": 8156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9925088147655488,
+      "learning_rate": 4.885478181785991e-06,
+      "loss": 1.0779,
+      "step": 8157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9865228169603488,
+      "learning_rate": 4.884441968981781e-06,
+      "loss": 0.6303,
+      "step": 8158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8754046360249936,
+      "learning_rate": 4.883405761143357e-06,
+      "loss": 0.7502,
+      "step": 8159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5781191387160503,
+      "learning_rate": 4.882369558315246e-06,
+      "loss": 0.7054,
+      "step": 8160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0725517471913992,
+      "learning_rate": 4.8813333605419775e-06,
+      "loss": 0.9197,
+      "step": 8161
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.524594210073642,
+      "learning_rate": 4.88029716786808e-06,
+      "loss": 0.753,
+      "step": 8162
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9522955451773243,
+      "learning_rate": 4.879260980338078e-06,
+      "loss": 0.643,
+      "step": 8163
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3130378574575747,
+      "learning_rate": 4.878224797996502e-06,
+      "loss": 1.0124,
+      "step": 8164
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3968306994446578,
+      "learning_rate": 4.877188620887876e-06,
+      "loss": 0.8754,
+      "step": 8165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1356504417592133,
+      "learning_rate": 4.8761524490567315e-06,
+      "loss": 0.7938,
+      "step": 8166
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1321381776800867,
+      "learning_rate": 4.875116282547589e-06,
+      "loss": 0.6681,
+      "step": 8167
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3882024066610423,
+      "learning_rate": 4.874080121404979e-06,
+      "loss": 0.7377,
+      "step": 8168
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0832424312913345,
+      "learning_rate": 4.873043965673427e-06,
+      "loss": 0.9012,
+      "step": 8169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9015757749249371,
+      "learning_rate": 4.872007815397459e-06,
+      "loss": 0.9222,
+      "step": 8170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7275534024052837,
+      "learning_rate": 4.870971670621599e-06,
+      "loss": 0.9201,
+      "step": 8171
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.657490525586087,
+      "learning_rate": 4.869935531390374e-06,
+      "loss": 0.7655,
+      "step": 8172
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2402549575345705,
+      "learning_rate": 4.86889939774831e-06,
+      "loss": 0.7884,
+      "step": 8173
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2496611159400333,
+      "learning_rate": 4.86786326973993e-06,
+      "loss": 0.6722,
+      "step": 8174
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.079607938095052,
+      "learning_rate": 4.866827147409758e-06,
+      "loss": 0.7469,
+      "step": 8175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.229654335848921,
+      "learning_rate": 4.8657910308023205e-06,
+      "loss": 0.7351,
+      "step": 8176
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1828227795620765,
+      "learning_rate": 4.864754919962144e-06,
+      "loss": 0.9133,
+      "step": 8177
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2003786707872848,
+      "learning_rate": 4.863718814933749e-06,
+      "loss": 0.6221,
+      "step": 8178
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.394429626039037,
+      "learning_rate": 4.862682715761658e-06,
+      "loss": 1.0171,
+      "step": 8179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1437427281029393,
+      "learning_rate": 4.861646622490399e-06,
+      "loss": 0.7377,
+      "step": 8180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8869242919246956,
+      "learning_rate": 4.860610535164491e-06,
+      "loss": 0.7146,
+      "step": 8181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7119929347679843,
+      "learning_rate": 4.85957445382846e-06,
+      "loss": 0.8803,
+      "step": 8182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.948641053131319,
+      "learning_rate": 4.858538378526825e-06,
+      "loss": 0.7463,
+      "step": 8183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.449783365776469,
+      "learning_rate": 4.857502309304114e-06,
+      "loss": 0.8921,
+      "step": 8184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0193528855622394,
+      "learning_rate": 4.856466246204844e-06,
+      "loss": 0.6283,
+      "step": 8185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8746304004596879,
+      "learning_rate": 4.8554301892735385e-06,
+      "loss": 0.7383,
+      "step": 8186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8741120707982788,
+      "learning_rate": 4.854394138554722e-06,
+      "loss": 0.7724,
+      "step": 8187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.151078420117743,
+      "learning_rate": 4.853358094092913e-06,
+      "loss": 0.9419,
+      "step": 8188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0326501621207589,
+      "learning_rate": 4.852322055932633e-06,
+      "loss": 0.5583,
+      "step": 8189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.355902413036743,
+      "learning_rate": 4.851286024118402e-06,
+      "loss": 0.7167,
+      "step": 8190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8480986310373977,
+      "learning_rate": 4.850249998694745e-06,
+      "loss": 0.7489,
+      "step": 8191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.558939463125196,
+      "learning_rate": 4.849213979706177e-06,
+      "loss": 0.8075,
+      "step": 8192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9384886501010774,
+      "learning_rate": 4.848177967197219e-06,
+      "loss": 0.7373,
+      "step": 8193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.293841190231117,
+      "learning_rate": 4.8471419612123925e-06,
+      "loss": 0.9815,
+      "step": 8194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0673624700720954,
+      "learning_rate": 4.846105961796218e-06,
+      "loss": 0.6905,
+      "step": 8195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7908173328389265,
+      "learning_rate": 4.845069968993212e-06,
+      "loss": 0.836,
+      "step": 8196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.52933480662095,
+      "learning_rate": 4.844033982847893e-06,
+      "loss": 0.8741,
+      "step": 8197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.611198267453255,
+      "learning_rate": 4.842998003404784e-06,
+      "loss": 0.7953,
+      "step": 8198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.916451395995736,
+      "learning_rate": 4.841962030708398e-06,
+      "loss": 0.885,
+      "step": 8199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.117413691126587,
+      "learning_rate": 4.840926064803256e-06,
+      "loss": 0.8668,
+      "step": 8200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9147138092692024,
+      "learning_rate": 4.839890105733876e-06,
+      "loss": 0.9183,
+      "step": 8201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3873384185297115,
+      "learning_rate": 4.8388541535447756e-06,
+      "loss": 0.8369,
+      "step": 8202
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9058466301225896,
+      "learning_rate": 4.837818208280469e-06,
+      "loss": 0.8385,
+      "step": 8203
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.465423421438705,
+      "learning_rate": 4.836782269985475e-06,
+      "loss": 0.8919,
+      "step": 8204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.155553825446175,
+      "learning_rate": 4.8357463387043124e-06,
+      "loss": 0.8585,
+      "step": 8205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4399841130808975,
+      "learning_rate": 4.834710414481496e-06,
+      "loss": 0.8081,
+      "step": 8206
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9888716224383438,
+      "learning_rate": 4.833674497361539e-06,
+      "loss": 0.8846,
+      "step": 8207
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.62197791520574,
+      "learning_rate": 4.832638587388959e-06,
+      "loss": 0.6361,
+      "step": 8208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9397419588139064,
+      "learning_rate": 4.831602684608274e-06,
+      "loss": 0.9389,
+      "step": 8209
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.446090940188656,
+      "learning_rate": 4.830566789063995e-06,
+      "loss": 0.9562,
+      "step": 8210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1714718751168816,
+      "learning_rate": 4.829530900800638e-06,
+      "loss": 0.8339,
+      "step": 8211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.266524144721732,
+      "learning_rate": 4.8284950198627186e-06,
+      "loss": 0.8295,
+      "step": 8212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.036106920189726,
+      "learning_rate": 4.827459146294751e-06,
+      "loss": 1.0783,
+      "step": 8213
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1525995635632826,
+      "learning_rate": 4.826423280141247e-06,
+      "loss": 0.6906,
+      "step": 8214
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.159461020778012,
+      "learning_rate": 4.825387421446721e-06,
+      "loss": 0.7534,
+      "step": 8215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.373108562147627,
+      "learning_rate": 4.824351570255689e-06,
+      "loss": 0.6461,
+      "step": 8216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.431394504074747,
+      "learning_rate": 4.823315726612659e-06,
+      "loss": 0.8519,
+      "step": 8217
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4085116527934518,
+      "learning_rate": 4.8222798905621445e-06,
+      "loss": 0.7278,
+      "step": 8218
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.286041444381717,
+      "learning_rate": 4.82124406214866e-06,
+      "loss": 0.7358,
+      "step": 8219
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9213445154472792,
+      "learning_rate": 4.820208241416719e-06,
+      "loss": 0.9245,
+      "step": 8220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2594612814732775,
+      "learning_rate": 4.819172428410826e-06,
+      "loss": 0.7292,
+      "step": 8221
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8945710593213583,
+      "learning_rate": 4.818136623175498e-06,
+      "loss": 0.8806,
+      "step": 8222
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.86072288860185,
+      "learning_rate": 4.817100825755244e-06,
+      "loss": 0.8164,
+      "step": 8223
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.060095757299578,
+      "learning_rate": 4.816065036194576e-06,
+      "loss": 0.7297,
+      "step": 8224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.202862856451875,
+      "learning_rate": 4.815029254538003e-06,
+      "loss": 0.8984,
+      "step": 8225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.269153536744708,
+      "learning_rate": 4.813993480830034e-06,
+      "loss": 1.1697,
+      "step": 8226
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.293412191616694,
+      "learning_rate": 4.812957715115181e-06,
+      "loss": 0.9449,
+      "step": 8227
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.101562258106534,
+      "learning_rate": 4.8119219574379495e-06,
+      "loss": 0.6889,
+      "step": 8228
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4922787729459945,
+      "learning_rate": 4.810886207842852e-06,
+      "loss": 0.7915,
+      "step": 8229
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9945475888445319,
+      "learning_rate": 4.809850466374393e-06,
+      "loss": 0.6986,
+      "step": 8230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.218000838786888,
+      "learning_rate": 4.8088147330770855e-06,
+      "loss": 0.8972,
+      "step": 8231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4952099811468385,
+      "learning_rate": 4.807779007995434e-06,
+      "loss": 0.8342,
+      "step": 8232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.031420544790295,
+      "learning_rate": 4.8067432911739455e-06,
+      "loss": 0.8533,
+      "step": 8233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1344766330550886,
+      "learning_rate": 4.80570758265713e-06,
+      "loss": 0.8454,
+      "step": 8234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7134887887061732,
+      "learning_rate": 4.804671882489491e-06,
+      "loss": 0.8924,
+      "step": 8235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.803471509894886,
+      "learning_rate": 4.803636190715536e-06,
+      "loss": 0.8086,
+      "step": 8236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3135366113822964,
+      "learning_rate": 4.802600507379771e-06,
+      "loss": 0.8525,
+      "step": 8237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9214279002349812,
+      "learning_rate": 4.801564832526704e-06,
+      "loss": 0.842,
+      "step": 8238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3490856867035035,
+      "learning_rate": 4.800529166200837e-06,
+      "loss": 0.6785,
+      "step": 8239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.087262239635389,
+      "learning_rate": 4.7994935084466745e-06,
+      "loss": 0.7916,
+      "step": 8240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1768313473073144,
+      "learning_rate": 4.798457859308724e-06,
+      "loss": 0.7436,
+      "step": 8241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.311709117331815,
+      "learning_rate": 4.797422218831488e-06,
+      "loss": 0.8437,
+      "step": 8242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21295380213105,
+      "learning_rate": 4.796386587059471e-06,
+      "loss": 0.6292,
+      "step": 8243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1652845568107773,
+      "learning_rate": 4.795350964037174e-06,
+      "loss": 0.9306,
+      "step": 8244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4576325754364667,
+      "learning_rate": 4.794315349809105e-06,
+      "loss": 0.7577,
+      "step": 8245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.938896124667751,
+      "learning_rate": 4.7932797444197604e-06,
+      "loss": 0.9367,
+      "step": 8246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.607096749628023,
+      "learning_rate": 4.792244147913647e-06,
+      "loss": 0.8748,
+      "step": 8247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8155656185854223,
+      "learning_rate": 4.791208560335264e-06,
+      "loss": 0.6973,
+      "step": 8248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8925805734732215,
+      "learning_rate": 4.790172981729116e-06,
+      "loss": 0.6438,
+      "step": 8249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9675178903032036,
+      "learning_rate": 4.789137412139701e-06,
+      "loss": 0.9484,
+      "step": 8250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.062347467558229,
+      "learning_rate": 4.7881018516115205e-06,
+      "loss": 0.9057,
+      "step": 8251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3483412913335124,
+      "learning_rate": 4.787066300189077e-06,
+      "loss": 0.6543,
+      "step": 8252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8928246269136817,
+      "learning_rate": 4.786030757916868e-06,
+      "loss": 0.7885,
+      "step": 8253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0578398437251195,
+      "learning_rate": 4.784995224839394e-06,
+      "loss": 0.9154,
+      "step": 8254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0722228760081816,
+      "learning_rate": 4.783959701001153e-06,
+      "loss": 0.7059,
+      "step": 8255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.162771498881213,
+      "learning_rate": 4.782924186446647e-06,
+      "loss": 0.8676,
+      "step": 8256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.357061279408213,
+      "learning_rate": 4.7818886812203715e-06,
+      "loss": 0.7504,
+      "step": 8257
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.495253358522335,
+      "learning_rate": 4.780853185366823e-06,
+      "loss": 0.6232,
+      "step": 8258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.891023656651448,
+      "learning_rate": 4.779817698930502e-06,
+      "loss": 0.8068,
+      "step": 8259
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.483432807802025,
+      "learning_rate": 4.778782221955907e-06,
+      "loss": 0.9067,
+      "step": 8260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7684977458978,
+      "learning_rate": 4.77774675448753e-06,
+      "loss": 0.7532,
+      "step": 8261
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.146280776631712,
+      "learning_rate": 4.776711296569872e-06,
+      "loss": 0.8349,
+      "step": 8262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.398654019369303,
+      "learning_rate": 4.775675848247427e-06,
+      "loss": 0.7315,
+      "step": 8263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5080276799064634,
+      "learning_rate": 4.774640409564688e-06,
+      "loss": 0.7735,
+      "step": 8264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9056824833179906,
+      "learning_rate": 4.773604980566154e-06,
+      "loss": 0.8695,
+      "step": 8265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2808549090465613,
+      "learning_rate": 4.772569561296318e-06,
+      "loss": 0.8982,
+      "step": 8266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8941670547239295,
+      "learning_rate": 4.771534151799676e-06,
+      "loss": 0.8771,
+      "step": 8267
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.55292615363853,
+      "learning_rate": 4.770498752120718e-06,
+      "loss": 0.7195,
+      "step": 8268
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8830521672481093,
+      "learning_rate": 4.76946336230394e-06,
+      "loss": 0.9131,
+      "step": 8269
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9882077298301783,
+      "learning_rate": 4.768427982393836e-06,
+      "loss": 0.8706,
+      "step": 8270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9793138216821151,
+      "learning_rate": 4.767392612434897e-06,
+      "loss": 0.8035,
+      "step": 8271
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8760749055586312,
+      "learning_rate": 4.766357252471615e-06,
+      "loss": 0.7246,
+      "step": 8272
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3086943963271205,
+      "learning_rate": 4.765321902548482e-06,
+      "loss": 0.8566,
+      "step": 8273
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.261169001430196,
+      "learning_rate": 4.76428656270999e-06,
+      "loss": 0.6144,
+      "step": 8274
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.073995247657764,
+      "learning_rate": 4.7632512330006296e-06,
+      "loss": 0.8229,
+      "step": 8275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0841679602313345,
+      "learning_rate": 4.76221591346489e-06,
+      "loss": 0.5434,
+      "step": 8276
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1480202947274725,
+      "learning_rate": 4.761180604147262e-06,
+      "loss": 0.7183,
+      "step": 8277
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.92252556103466,
+      "learning_rate": 4.760145305092238e-06,
+      "loss": 0.949,
+      "step": 8278
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.883100852065464,
+      "learning_rate": 4.759110016344302e-06,
+      "loss": 0.797,
+      "step": 8279
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3056082095817614,
+      "learning_rate": 4.7580747379479455e-06,
+      "loss": 0.7701,
+      "step": 8280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.636791260621539,
+      "learning_rate": 4.757039469947658e-06,
+      "loss": 0.9217,
+      "step": 8281
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1271675422611542,
+      "learning_rate": 4.756004212387923e-06,
+      "loss": 0.7719,
+      "step": 8282
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9088837050282097,
+      "learning_rate": 4.7549689653132304e-06,
+      "loss": 0.5935,
+      "step": 8283
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0512271354870224,
+      "learning_rate": 4.753933728768069e-06,
+      "loss": 0.9948,
+      "step": 8284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.284153469498935,
+      "learning_rate": 4.752898502796922e-06,
+      "loss": 0.915,
+      "step": 8285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2876675626389051,
+      "learning_rate": 4.751863287444278e-06,
+      "loss": 0.6948,
+      "step": 8286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2015351562327516,
+      "learning_rate": 4.750828082754619e-06,
+      "loss": 0.6803,
+      "step": 8287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.311367972965784,
+      "learning_rate": 4.7497928887724325e-06,
+      "loss": 0.9468,
+      "step": 8288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2848864894183305,
+      "learning_rate": 4.748757705542205e-06,
+      "loss": 0.8953,
+      "step": 8289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.77476502740408,
+      "learning_rate": 4.7477225331084174e-06,
+      "loss": 0.9232,
+      "step": 8290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1136500442095505,
+      "learning_rate": 4.746687371515554e-06,
+      "loss": 0.8572,
+      "step": 8291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.365742944706805,
+      "learning_rate": 4.7456522208081e-06,
+      "loss": 0.8835,
+      "step": 8292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7643898161909757,
+      "learning_rate": 4.7446170810305346e-06,
+      "loss": 0.8907,
+      "step": 8293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8062657760573717,
+      "learning_rate": 4.743581952227342e-06,
+      "loss": 0.7728,
+      "step": 8294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7990354336766354,
+      "learning_rate": 4.7425468344430035e-06,
+      "loss": 0.8177,
+      "step": 8295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.378863135449792,
+      "learning_rate": 4.741511727722003e-06,
+      "loss": 1.0001,
+      "step": 8296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0586476937930263,
+      "learning_rate": 4.740476632108818e-06,
+      "loss": 0.8001,
+      "step": 8297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.870309330172199,
+      "learning_rate": 4.73944154764793e-06,
+      "loss": 0.8695,
+      "step": 8298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.764241285563021,
+      "learning_rate": 4.73840647438382e-06,
+      "loss": 0.7414,
+      "step": 8299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.831037675010587,
+      "learning_rate": 4.737371412360966e-06,
+      "loss": 0.8665,
+      "step": 8300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1998149371552962,
+      "learning_rate": 4.7363363616238465e-06,
+      "loss": 0.6865,
+      "step": 8301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.928089169511513,
+      "learning_rate": 4.73530132221694e-06,
+      "loss": 0.7649,
+      "step": 8302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5265274841272296,
+      "learning_rate": 4.734266294184728e-06,
+      "loss": 0.7108,
+      "step": 8303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2060522474152076,
+      "learning_rate": 4.733231277571683e-06,
+      "loss": 0.6211,
+      "step": 8304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.135474769715715,
+      "learning_rate": 4.732196272422285e-06,
+      "loss": 0.7147,
+      "step": 8305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0663279156828023,
+      "learning_rate": 4.731161278781008e-06,
+      "loss": 0.8385,
+      "step": 8306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8519970978495426,
+      "learning_rate": 4.730126296692332e-06,
+      "loss": 0.7006,
+      "step": 8307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.633355303470362,
+      "learning_rate": 4.72909132620073e-06,
+      "loss": 0.6469,
+      "step": 8308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1327093748050316,
+      "learning_rate": 4.7280563673506745e-06,
+      "loss": 0.9291,
+      "step": 8309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7919138568199013,
+      "learning_rate": 4.727021420186646e-06,
+      "loss": 0.7927,
+      "step": 8310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.020138060875692,
+      "learning_rate": 4.725986484753112e-06,
+      "loss": 0.7309,
+      "step": 8311
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.474137961399376,
+      "learning_rate": 4.72495156109455e-06,
+      "loss": 0.8522,
+      "step": 8312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.278193627739544,
+      "learning_rate": 4.723916649255432e-06,
+      "loss": 0.8258,
+      "step": 8313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9716043370844023,
+      "learning_rate": 4.722881749280232e-06,
+      "loss": 0.8289,
+      "step": 8314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0569554051757195,
+      "learning_rate": 4.7218468612134175e-06,
+      "loss": 0.6047,
+      "step": 8315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8497026035717665,
+      "learning_rate": 4.720811985099464e-06,
+      "loss": 0.8209,
+      "step": 8316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7732925672022264,
+      "learning_rate": 4.719777120982843e-06,
+      "loss": 0.6505,
+      "step": 8317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.637070674070581,
+      "learning_rate": 4.718742268908022e-06,
+      "loss": 0.742,
+      "step": 8318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.564238120338512,
+      "learning_rate": 4.717707428919471e-06,
+      "loss": 0.8393,
+      "step": 8319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2644730863223836,
+      "learning_rate": 4.716672601061661e-06,
+      "loss": 0.7748,
+      "step": 8320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.971474949224426,
+      "learning_rate": 4.715637785379062e-06,
+      "loss": 0.8847,
+      "step": 8321
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.422230432184576,
+      "learning_rate": 4.714602981916139e-06,
+      "loss": 0.8841,
+      "step": 8322
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.669667288633144,
+      "learning_rate": 4.713568190717362e-06,
+      "loss": 0.7702,
+      "step": 8323
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0820446580042902,
+      "learning_rate": 4.712533411827197e-06,
+      "loss": 0.6388,
+      "step": 8324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.492416575973486,
+      "learning_rate": 4.711498645290113e-06,
+      "loss": 0.8907,
+      "step": 8325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4083090171696973,
+      "learning_rate": 4.710463891150573e-06,
+      "loss": 0.8861,
+      "step": 8326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.912138988607449,
+      "learning_rate": 4.709429149453046e-06,
+      "loss": 0.9003,
+      "step": 8327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.135752726311647,
+      "learning_rate": 4.708394420241996e-06,
+      "loss": 0.6597,
+      "step": 8328
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0854960397899416,
+      "learning_rate": 4.707359703561885e-06,
+      "loss": 0.6802,
+      "step": 8329
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.09380291457219,
+      "learning_rate": 4.70632499945718e-06,
+      "loss": 0.8658,
+      "step": 8330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.707583144430604,
+      "learning_rate": 4.705290307972344e-06,
+      "loss": 0.7008,
+      "step": 8331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1127500179906022,
+      "learning_rate": 4.7042556291518415e-06,
+      "loss": 0.6575,
+      "step": 8332
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2921458824869605,
+      "learning_rate": 4.703220963040131e-06,
+      "loss": 0.786,
+      "step": 8333
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5856518663635017,
+      "learning_rate": 4.702186309681677e-06,
+      "loss": 0.7481,
+      "step": 8334
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0195873402033113,
+      "learning_rate": 4.701151669120942e-06,
+      "loss": 0.9594,
+      "step": 8335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.065184156956872,
+      "learning_rate": 4.700117041402384e-06,
+      "loss": 0.8382,
+      "step": 8336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.129582257071434,
+      "learning_rate": 4.699082426570465e-06,
+      "loss": 0.6486,
+      "step": 8337
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1995022989440667,
+      "learning_rate": 4.6980478246696435e-06,
+      "loss": 0.698,
+      "step": 8338
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1841462328109613,
+      "learning_rate": 4.697013235744382e-06,
+      "loss": 0.8501,
+      "step": 8339
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.185897043318179,
+      "learning_rate": 4.695978659839133e-06,
+      "loss": 0.9185,
+      "step": 8340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0880513269320224,
+      "learning_rate": 4.694944096998361e-06,
+      "loss": 0.6553,
+      "step": 8341
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.041827353919736,
+      "learning_rate": 4.693909547266518e-06,
+      "loss": 0.8636,
+      "step": 8342
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.43386511802473,
+      "learning_rate": 4.692875010688066e-06,
+      "loss": 0.7407,
+      "step": 8343
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8509888903903617,
+      "learning_rate": 4.6918404873074574e-06,
+      "loss": 0.8082,
+      "step": 8344
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9777813336829475,
+      "learning_rate": 4.69080597716915e-06,
+      "loss": 0.6798,
+      "step": 8345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0517902255211675,
+      "learning_rate": 4.6897714803175995e-06,
+      "loss": 0.791,
+      "step": 8346
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9714133918355587,
+      "learning_rate": 4.688736996797257e-06,
+      "loss": 0.7373,
+      "step": 8347
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.018721767526977,
+      "learning_rate": 4.687702526652579e-06,
+      "loss": 0.7758,
+      "step": 8348
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8877491454213642,
+      "learning_rate": 4.68666806992802e-06,
+      "loss": 0.7747,
+      "step": 8349
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9040408649904101,
+      "learning_rate": 4.685633626668032e-06,
+      "loss": 0.844,
+      "step": 8350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9791885774159379,
+      "learning_rate": 4.684599196917067e-06,
+      "loss": 0.6722,
+      "step": 8351
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4784774637308624,
+      "learning_rate": 4.683564780719576e-06,
+      "loss": 0.8134,
+      "step": 8352
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9237658154890382,
+      "learning_rate": 4.682530378120014e-06,
+      "loss": 0.9131,
+      "step": 8353
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2234564107435872,
+      "learning_rate": 4.681495989162826e-06,
+      "loss": 0.7522,
+      "step": 8354
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.032842381498665,
+      "learning_rate": 4.680461613892465e-06,
+      "loss": 0.8683,
+      "step": 8355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109508289193972,
+      "learning_rate": 4.679427252353379e-06,
+      "loss": 0.8249,
+      "step": 8356
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0540926094982965,
+      "learning_rate": 4.678392904590021e-06,
+      "loss": 0.8217,
+      "step": 8357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0853871923228477,
+      "learning_rate": 4.677358570646834e-06,
+      "loss": 0.9322,
+      "step": 8358
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8671152386477434,
+      "learning_rate": 4.676324250568269e-06,
+      "loss": 0.7009,
+      "step": 8359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9722617440461195,
+      "learning_rate": 4.6752899443987694e-06,
+      "loss": 0.6831,
+      "step": 8360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.494700198422774,
+      "learning_rate": 4.674255652182788e-06,
+      "loss": 0.7427,
+      "step": 8361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.094685371589439,
+      "learning_rate": 4.673221373964764e-06,
+      "loss": 0.9107,
+      "step": 8362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.774102501338246,
+      "learning_rate": 4.672187109789144e-06,
+      "loss": 0.7881,
+      "step": 8363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1398936830773614,
+      "learning_rate": 4.671152859700377e-06,
+      "loss": 0.5967,
+      "step": 8364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.67200863268378,
+      "learning_rate": 4.670118623742904e-06,
+      "loss": 0.7379,
+      "step": 8365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0879928443738374,
+      "learning_rate": 4.669084401961166e-06,
+      "loss": 0.9509,
+      "step": 8366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1056096296955513,
+      "learning_rate": 4.668050194399609e-06,
+      "loss": 0.6522,
+      "step": 8367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.549722607099087,
+      "learning_rate": 4.667016001102675e-06,
+      "loss": 0.9446,
+      "step": 8368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.175626722520248,
+      "learning_rate": 4.665981822114805e-06,
+      "loss": 0.9252,
+      "step": 8369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4601219411582944,
+      "learning_rate": 4.66494765748044e-06,
+      "loss": 0.8174,
+      "step": 8370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1388630529819963,
+      "learning_rate": 4.6639135072440195e-06,
+      "loss": 0.8098,
+      "step": 8371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.766930305944748,
+      "learning_rate": 4.662879371449987e-06,
+      "loss": 1.053,
+      "step": 8372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1244106584340967,
+      "learning_rate": 4.6618452501427755e-06,
+      "loss": 0.6599,
+      "step": 8373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1856977371679762,
+      "learning_rate": 4.660811143366828e-06,
+      "loss": 0.6255,
+      "step": 8374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9449389741581633,
+      "learning_rate": 4.6597770511665814e-06,
+      "loss": 0.7786,
+      "step": 8375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9188055023697084,
+      "learning_rate": 4.658742973586471e-06,
+      "loss": 0.7961,
+      "step": 8376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4094427777953202,
+      "learning_rate": 4.657708910670936e-06,
+      "loss": 0.6591,
+      "step": 8377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2687506434575275,
+      "learning_rate": 4.656674862464412e-06,
+      "loss": 0.8748,
+      "step": 8378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.702212957985695,
+      "learning_rate": 4.655640829011335e-06,
+      "loss": 0.7451,
+      "step": 8379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5205135442524274,
+      "learning_rate": 4.654606810356135e-06,
+      "loss": 0.7642,
+      "step": 8380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.087707400847041,
+      "learning_rate": 4.653572806543251e-06,
+      "loss": 0.7189,
+      "step": 8381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.719004979364866,
+      "learning_rate": 4.652538817617117e-06,
+      "loss": 0.7807,
+      "step": 8382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1550981624169026,
+      "learning_rate": 4.651504843622163e-06,
+      "loss": 0.8463,
+      "step": 8383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0918945719910416,
+      "learning_rate": 4.65047088460282e-06,
+      "loss": 0.8601,
+      "step": 8384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.023484872154769,
+      "learning_rate": 4.6494369406035225e-06,
+      "loss": 0.902,
+      "step": 8385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5302110579867954,
+      "learning_rate": 4.6484030116687014e-06,
+      "loss": 0.8366,
+      "step": 8386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.064476551595988,
+      "learning_rate": 4.647369097842785e-06,
+      "loss": 0.5967,
+      "step": 8387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.81630015089988,
+      "learning_rate": 4.646335199170205e-06,
+      "loss": 0.8028,
+      "step": 8388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4313433174277126,
+      "learning_rate": 4.645301315695387e-06,
+      "loss": 0.7687,
+      "step": 8389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.526937700869058,
+      "learning_rate": 4.6442674474627645e-06,
+      "loss": 0.7427,
+      "step": 8390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2220881462400424,
+      "learning_rate": 4.643233594516759e-06,
+      "loss": 0.6418,
+      "step": 8391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7268904088945494,
+      "learning_rate": 4.642199756901802e-06,
+      "loss": 0.8941,
+      "step": 8392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3282299251830767,
+      "learning_rate": 4.64116593466232e-06,
+      "loss": 0.8103,
+      "step": 8393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.809514989275365,
+      "learning_rate": 4.6401321278427334e-06,
+      "loss": 0.7815,
+      "step": 8394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7306821499627496,
+      "learning_rate": 4.639098336487472e-06,
+      "loss": 0.7417,
+      "step": 8395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4189012788420734,
+      "learning_rate": 4.638064560640959e-06,
+      "loss": 0.9216,
+      "step": 8396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.618552728805161,
+      "learning_rate": 4.63703080034762e-06,
+      "loss": 0.706,
+      "step": 8397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.266655247169185,
+      "learning_rate": 4.635997055651873e-06,
+      "loss": 0.9715,
+      "step": 8398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7987540199998984,
+      "learning_rate": 4.634963326598143e-06,
+      "loss": 0.671,
+      "step": 8399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.441027025588627,
+      "learning_rate": 4.633929613230855e-06,
+      "loss": 0.7205,
+      "step": 8400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.023592409219347,
+      "learning_rate": 4.632895915594424e-06,
+      "loss": 0.9489,
+      "step": 8401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6468938450154464,
+      "learning_rate": 4.631862233733274e-06,
+      "loss": 0.7671,
+      "step": 8402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.156915404360699,
+      "learning_rate": 4.630828567691823e-06,
+      "loss": 0.6814,
+      "step": 8403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1208424478743244,
+      "learning_rate": 4.629794917514492e-06,
+      "loss": 0.8431,
+      "step": 8404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8957410338215288,
+      "learning_rate": 4.628761283245697e-06,
+      "loss": 0.5952,
+      "step": 8405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3608331809109675,
+      "learning_rate": 4.627727664929856e-06,
+      "loss": 0.728,
+      "step": 8406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9844288723025532,
+      "learning_rate": 4.626694062611387e-06,
+      "loss": 0.9156,
+      "step": 8407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.917833279670697,
+      "learning_rate": 4.6256604763347066e-06,
+      "loss": 0.8357,
+      "step": 8408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8537095690047474,
+      "learning_rate": 4.624626906144227e-06,
+      "loss": 0.9432,
+      "step": 8409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9924350017055528,
+      "learning_rate": 4.623593352084367e-06,
+      "loss": 0.9328,
+      "step": 8410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9958342407730614,
+      "learning_rate": 4.622559814199538e-06,
+      "loss": 0.9721,
+      "step": 8411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.879186183971373,
+      "learning_rate": 4.621526292534156e-06,
+      "loss": 0.8091,
+      "step": 8412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.270992437043114,
+      "learning_rate": 4.6204927871326295e-06,
+      "loss": 0.9312,
+      "step": 8413
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.540554497588833,
+      "learning_rate": 4.619459298039373e-06,
+      "loss": 0.8106,
+      "step": 8414
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0936419246546158,
+      "learning_rate": 4.6184258252988016e-06,
+      "loss": 0.6269,
+      "step": 8415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.05905254823702,
+      "learning_rate": 4.6173923689553205e-06,
+      "loss": 0.7818,
+      "step": 8416
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1816926674561468,
+      "learning_rate": 4.616358929053339e-06,
+      "loss": 0.6777,
+      "step": 8417
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1580277545458375,
+      "learning_rate": 4.615325505637272e-06,
+      "loss": 0.8282,
+      "step": 8418
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3044013382858575,
+      "learning_rate": 4.614292098751524e-06,
+      "loss": 0.783,
+      "step": 8419
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2827517968640842,
+      "learning_rate": 4.613258708440503e-06,
+      "loss": 0.769,
+      "step": 8420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1582130641458384,
+      "learning_rate": 4.612225334748616e-06,
+      "loss": 0.7013,
+      "step": 8421
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4483127467715593,
+      "learning_rate": 4.611191977720272e-06,
+      "loss": 1.0252,
+      "step": 8422
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.503536363984517,
+      "learning_rate": 4.610158637399872e-06,
+      "loss": 0.7896,
+      "step": 8423
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8747498641337406,
+      "learning_rate": 4.609125313831826e-06,
+      "loss": 0.6275,
+      "step": 8424
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8338447169317853,
+      "learning_rate": 4.608092007060533e-06,
+      "loss": 0.6787,
+      "step": 8425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.401736257598924,
+      "learning_rate": 4.607058717130403e-06,
+      "loss": 0.8498,
+      "step": 8426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.466219182885034,
+      "learning_rate": 4.6060254440858315e-06,
+      "loss": 0.714,
+      "step": 8427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0780426522635644,
+      "learning_rate": 4.6049921879712254e-06,
+      "loss": 0.933,
+      "step": 8428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5727820136899386,
+      "learning_rate": 4.603958948830985e-06,
+      "loss": 0.8722,
+      "step": 8429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.786030705310982,
+      "learning_rate": 4.602925726709512e-06,
+      "loss": 0.8191,
+      "step": 8430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.246608962739744,
+      "learning_rate": 4.6018925216512025e-06,
+      "loss": 0.5645,
+      "step": 8431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3062753222203907,
+      "learning_rate": 4.600859333700457e-06,
+      "loss": 0.9943,
+      "step": 8432
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7847095942788385,
+      "learning_rate": 4.599826162901679e-06,
+      "loss": 0.7375,
+      "step": 8433
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.414658035501849,
+      "learning_rate": 4.5987930092992596e-06,
+      "loss": 0.8423,
+      "step": 8434
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9421159689703313,
+      "learning_rate": 4.597759872937597e-06,
+      "loss": 0.8066,
+      "step": 8435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.827807871643433,
+      "learning_rate": 4.5967267538610915e-06,
+      "loss": 0.9222,
+      "step": 8436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.205357425397472,
+      "learning_rate": 4.595693652114133e-06,
+      "loss": 0.8016,
+      "step": 8437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6215688656675833,
+      "learning_rate": 4.594660567741118e-06,
+      "loss": 0.7143,
+      "step": 8438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9622088479712985,
+      "learning_rate": 4.593627500786444e-06,
+      "loss": 1.0542,
+      "step": 8439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.074403759243639,
+      "learning_rate": 4.592594451294501e-06,
+      "loss": 0.9854,
+      "step": 8440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6750759313410444,
+      "learning_rate": 4.59156141930968e-06,
+      "loss": 0.5992,
+      "step": 8441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0201062865640496,
+      "learning_rate": 4.590528404876374e-06,
+      "loss": 0.6865,
+      "step": 8442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0754825021206806,
+      "learning_rate": 4.5894954080389755e-06,
+      "loss": 0.8488,
+      "step": 8443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6515254962283032,
+      "learning_rate": 4.588462428841875e-06,
+      "loss": 0.8697,
+      "step": 8444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7308247809173778,
+      "learning_rate": 4.587429467329458e-06,
+      "loss": 0.8207,
+      "step": 8445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0221278506752554,
+      "learning_rate": 4.586396523546116e-06,
+      "loss": 0.8408,
+      "step": 8446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5544191131680125,
+      "learning_rate": 4.585363597536239e-06,
+      "loss": 0.7802,
+      "step": 8447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1912651256377709,
+      "learning_rate": 4.584330689344211e-06,
+      "loss": 0.7034,
+      "step": 8448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8996559961841952,
+      "learning_rate": 4.5832977990144165e-06,
+      "loss": 0.7397,
+      "step": 8449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1992082936961803,
+      "learning_rate": 4.582264926591244e-06,
+      "loss": 0.7045,
+      "step": 8450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0637919809287704,
+      "learning_rate": 4.581232072119081e-06,
+      "loss": 0.7811,
+      "step": 8451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.149712194062328,
+      "learning_rate": 4.580199235642306e-06,
+      "loss": 0.8824,
+      "step": 8452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9510324601267108,
+      "learning_rate": 4.5791664172053044e-06,
+      "loss": 0.7387,
+      "step": 8453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4710167700368193,
+      "learning_rate": 4.578133616852462e-06,
+      "loss": 0.8224,
+      "step": 8454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9467932254264089,
+      "learning_rate": 4.577100834628155e-06,
+      "loss": 0.8553,
+      "step": 8455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8188371760990447,
+      "learning_rate": 4.5760680705767665e-06,
+      "loss": 0.8098,
+      "step": 8456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.688948548550079,
+      "learning_rate": 4.5750353247426785e-06,
+      "loss": 0.6644,
+      "step": 8457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28395375793518,
+      "learning_rate": 4.5740025971702695e-06,
+      "loss": 0.7013,
+      "step": 8458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2616105547973886,
+      "learning_rate": 4.572969887903916e-06,
+      "loss": 0.858,
+      "step": 8459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1964554987161717,
+      "learning_rate": 4.571937196987998e-06,
+      "loss": 0.733,
+      "step": 8460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8472279778356804,
+      "learning_rate": 4.570904524466893e-06,
+      "loss": 0.7382,
+      "step": 8461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9372782777304194,
+      "learning_rate": 4.5698718703849755e-06,
+      "loss": 0.8643,
+      "step": 8462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1042136542852454,
+      "learning_rate": 4.5688392347866226e-06,
+      "loss": 0.6602,
+      "step": 8463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3719579983109074,
+      "learning_rate": 4.5678066177162065e-06,
+      "loss": 0.8239,
+      "step": 8464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2677479448548894,
+      "learning_rate": 4.566774019218104e-06,
+      "loss": 0.7578,
+      "step": 8465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3143680526129877,
+      "learning_rate": 4.565741439336686e-06,
+      "loss": 0.7386,
+      "step": 8466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3319328498211593,
+      "learning_rate": 4.5647088781163255e-06,
+      "loss": 0.8636,
+      "step": 8467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.452694450361191,
+      "learning_rate": 4.563676335601393e-06,
+      "loss": 0.8306,
+      "step": 8468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0841413488337053,
+      "learning_rate": 4.562643811836263e-06,
+      "loss": 0.953,
+      "step": 8469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3390795113587948,
+      "learning_rate": 4.561611306865299e-06,
+      "loss": 0.7502,
+      "step": 8470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.446642704251839,
+      "learning_rate": 4.560578820732876e-06,
+      "loss": 0.8531,
+      "step": 8471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.569874462453896,
+      "learning_rate": 4.559546353483359e-06,
+      "loss": 0.9945,
+      "step": 8472
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9178880984243267,
+      "learning_rate": 4.558513905161116e-06,
+      "loss": 0.9198,
+      "step": 8473
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4368549760591063,
+      "learning_rate": 4.557481475810512e-06,
+      "loss": 0.8682,
+      "step": 8474
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.787248605892682,
+      "learning_rate": 4.556449065475917e-06,
+      "loss": 0.8293,
+      "step": 8475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0607069520105574,
+      "learning_rate": 4.555416674201693e-06,
+      "loss": 0.6391,
+      "step": 8476
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.198000814128326,
+      "learning_rate": 4.554384302032204e-06,
+      "loss": 0.7228,
+      "step": 8477
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7448240498751326,
+      "learning_rate": 4.553351949011814e-06,
+      "loss": 0.6987,
+      "step": 8478
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.430507311720347,
+      "learning_rate": 4.5523196151848846e-06,
+      "loss": 0.9178,
+      "step": 8479
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2730996814933058,
+      "learning_rate": 4.551287300595781e-06,
+      "loss": 0.6991,
+      "step": 8480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.720110844723674,
+      "learning_rate": 4.550255005288861e-06,
+      "loss": 0.8711,
+      "step": 8481
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1953920432159464,
+      "learning_rate": 4.549222729308483e-06,
+      "loss": 0.8712,
+      "step": 8482
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.266013358537719,
+      "learning_rate": 4.548190472699011e-06,
+      "loss": 0.8799,
+      "step": 8483
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5667895114206623,
+      "learning_rate": 4.547158235504797e-06,
+      "loss": 0.8286,
+      "step": 8484
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.229241287294286,
+      "learning_rate": 4.546126017770205e-06,
+      "loss": 0.946,
+      "step": 8485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7890213760250187,
+      "learning_rate": 4.5450938195395875e-06,
+      "loss": 0.8605,
+      "step": 8486
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2533857696827098,
+      "learning_rate": 4.544061640857303e-06,
+      "loss": 0.6968,
+      "step": 8487
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2735468984708374,
+      "learning_rate": 4.543029481767703e-06,
+      "loss": 0.9501,
+      "step": 8488
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2966852619187494,
+      "learning_rate": 4.541997342315145e-06,
+      "loss": 0.9373,
+      "step": 8489
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.249570477471089,
+      "learning_rate": 4.540965222543981e-06,
+      "loss": 0.7583,
+      "step": 8490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9960326019284036,
+      "learning_rate": 4.539933122498566e-06,
+      "loss": 0.8249,
+      "step": 8491
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.638289787437442,
+      "learning_rate": 4.5389010422232474e-06,
+      "loss": 0.8029,
+      "step": 8492
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7039035902999418,
+      "learning_rate": 4.5378689817623765e-06,
+      "loss": 0.8701,
+      "step": 8493
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9161565007856394,
+      "learning_rate": 4.536836941160308e-06,
+      "loss": 0.9124,
+      "step": 8494
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8800817700748629,
+      "learning_rate": 4.535804920461386e-06,
+      "loss": 0.7034,
+      "step": 8495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.139505896644392,
+      "learning_rate": 4.53477291970996e-06,
+      "loss": 0.8016,
+      "step": 8496
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.394023886241422,
+      "learning_rate": 4.5337409389503764e-06,
+      "loss": 0.9468,
+      "step": 8497
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9904216760146384,
+      "learning_rate": 4.532708978226987e-06,
+      "loss": 0.6225,
+      "step": 8498
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1408839333516623,
+      "learning_rate": 4.5316770375841315e-06,
+      "loss": 0.621,
+      "step": 8499
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5022666050771036,
+      "learning_rate": 4.530645117066155e-06,
+      "loss": 0.7882,
+      "step": 8500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0223649374873713,
+      "learning_rate": 4.529613216717406e-06,
+      "loss": 0.949,
+      "step": 8501
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9606831727896288,
+      "learning_rate": 4.528581336582223e-06,
+      "loss": 0.6972,
+      "step": 8502
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.172069489337756,
+      "learning_rate": 4.527549476704949e-06,
+      "loss": 0.7398,
+      "step": 8503
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6720759257496858,
+      "learning_rate": 4.526517637129927e-06,
+      "loss": 0.9188,
+      "step": 8504
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6505617894009328,
+      "learning_rate": 4.525485817901499e-06,
+      "loss": 0.5963,
+      "step": 8505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.919459990346172,
+      "learning_rate": 4.524454019063999e-06,
+      "loss": 0.6952,
+      "step": 8506
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.031672939266718,
+      "learning_rate": 4.52342224066177e-06,
+      "loss": 0.6983,
+      "step": 8507
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6359731960542896,
+      "learning_rate": 4.522390482739148e-06,
+      "loss": 0.8205,
+      "step": 8508
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1809600735867343,
+      "learning_rate": 4.5213587453404736e-06,
+      "loss": 0.7291,
+      "step": 8509
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7218341394858867,
+      "learning_rate": 4.520327028510076e-06,
+      "loss": 0.774,
+      "step": 8510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.185896729843122,
+      "learning_rate": 4.5192953322922955e-06,
+      "loss": 0.8405,
+      "step": 8511
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2078342688058696,
+      "learning_rate": 4.518263656731468e-06,
+      "loss": 0.6138,
+      "step": 8512
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1936584476209293,
+      "learning_rate": 4.5172320018719205e-06,
+      "loss": 0.7161,
+      "step": 8513
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.363442932138885,
+      "learning_rate": 4.5162003677579905e-06,
+      "loss": 0.8371,
+      "step": 8514
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1508370735386533,
+      "learning_rate": 4.5151687544340065e-06,
+      "loss": 0.6049,
+      "step": 8515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9012353599930238,
+      "learning_rate": 4.514137161944304e-06,
+      "loss": 0.8875,
+      "step": 8516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6942622157876905,
+      "learning_rate": 4.513105590333207e-06,
+      "loss": 0.8074,
+      "step": 8517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171758622145124,
+      "learning_rate": 4.512074039645049e-06,
+      "loss": 0.9179,
+      "step": 8518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.075702497813586,
+      "learning_rate": 4.511042509924157e-06,
+      "loss": 0.8754,
+      "step": 8519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9570662630248647,
+      "learning_rate": 4.5100110012148546e-06,
+      "loss": 0.7961,
+      "step": 8520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9928903439577963,
+      "learning_rate": 4.508979513561471e-06,
+      "loss": 0.8694,
+      "step": 8521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.11345886150789,
+      "learning_rate": 4.507948047008332e-06,
+      "loss": 0.6845,
+      "step": 8522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.335393114063844,
+      "learning_rate": 4.506916601599763e-06,
+      "loss": 0.768,
+      "step": 8523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1218750160899686,
+      "learning_rate": 4.505885177380083e-06,
+      "loss": 0.8226,
+      "step": 8524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.297150249428625,
+      "learning_rate": 4.504853774393618e-06,
+      "loss": 0.7545,
+      "step": 8525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140765152904224,
+      "learning_rate": 4.5038223926846905e-06,
+      "loss": 0.7002,
+      "step": 8526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.364334632760159,
+      "learning_rate": 4.5027910322976186e-06,
+      "loss": 0.8208,
+      "step": 8527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4215448627223264,
+      "learning_rate": 4.501759693276724e-06,
+      "loss": 0.6116,
+      "step": 8528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.891317442220451,
+      "learning_rate": 4.5007283756663245e-06,
+      "loss": 0.7332,
+      "step": 8529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.045080759168804,
+      "learning_rate": 4.49969707951074e-06,
+      "loss": 0.7347,
+      "step": 8530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.247047693991767,
+      "learning_rate": 4.498665804854285e-06,
+      "loss": 0.7365,
+      "step": 8531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1510441625635823,
+      "learning_rate": 4.497634551741277e-06,
+      "loss": 0.6775,
+      "step": 8532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.419936710030729,
+      "learning_rate": 4.49660332021603e-06,
+      "loss": 0.8051,
+      "step": 8533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.178070027949427,
+      "learning_rate": 4.495572110322862e-06,
+      "loss": 0.6375,
+      "step": 8534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2228614546748007,
+      "learning_rate": 4.494540922106082e-06,
+      "loss": 0.7197,
+      "step": 8535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8113887529080253,
+      "learning_rate": 4.4935097556100045e-06,
+      "loss": 0.7761,
+      "step": 8536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9822858597945228,
+      "learning_rate": 4.492478610878942e-06,
+      "loss": 0.7793,
+      "step": 8537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.119753295037239,
+      "learning_rate": 4.491447487957203e-06,
+      "loss": 0.7221,
+      "step": 8538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9720017173664541,
+      "learning_rate": 4.490416386889097e-06,
+      "loss": 0.8547,
+      "step": 8539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2837508476777433,
+      "learning_rate": 4.489385307718934e-06,
+      "loss": 0.7842,
+      "step": 8540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.097423377176818,
+      "learning_rate": 4.488354250491024e-06,
+      "loss": 0.5538,
+      "step": 8541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.124180332570562,
+      "learning_rate": 4.48732321524967e-06,
+      "loss": 0.7024,
+      "step": 8542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1644842321697757,
+      "learning_rate": 4.486292202039178e-06,
+      "loss": 0.9216,
+      "step": 8543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.475604044306503,
+      "learning_rate": 4.485261210903854e-06,
+      "loss": 0.8637,
+      "step": 8544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3280726273204884,
+      "learning_rate": 4.484230241888005e-06,
+      "loss": 0.841,
+      "step": 8545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3199055104456314,
+      "learning_rate": 4.48319929503593e-06,
+      "loss": 0.8402,
+      "step": 8546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0809042301523504,
+      "learning_rate": 4.482168370391931e-06,
+      "loss": 0.89,
+      "step": 8547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0742544870701718,
+      "learning_rate": 4.481137468000312e-06,
+      "loss": 0.5657,
+      "step": 8548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.82955310795131,
+      "learning_rate": 4.48010658790537e-06,
+      "loss": 0.9,
+      "step": 8549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.380962395787537,
+      "learning_rate": 4.479075730151406e-06,
+      "loss": 0.7312,
+      "step": 8550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2897638042938313,
+      "learning_rate": 4.478044894782718e-06,
+      "loss": 0.8458,
+      "step": 8551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.501518664738562,
+      "learning_rate": 4.477014081843605e-06,
+      "loss": 0.8695,
+      "step": 8552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.319150097924406,
+      "learning_rate": 4.47598329137836e-06,
+      "loss": 0.8077,
+      "step": 8553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6312079000228756,
+      "learning_rate": 4.47495252343128e-06,
+      "loss": 0.8596,
+      "step": 8554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9701460983930883,
+      "learning_rate": 4.473921778046661e-06,
+      "loss": 0.8787,
+      "step": 8555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.033197412115866,
+      "learning_rate": 4.4728910552687935e-06,
+      "loss": 0.7707,
+      "step": 8556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1106994449673167,
+      "learning_rate": 4.471860355141971e-06,
+      "loss": 0.7135,
+      "step": 8557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.048144564655911,
+      "learning_rate": 4.470829677710485e-06,
+      "loss": 0.6139,
+      "step": 8558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1149854904973386,
+      "learning_rate": 4.469799023018628e-06,
+      "loss": 0.6949,
+      "step": 8559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.545715755704066,
+      "learning_rate": 4.468768391110687e-06,
+      "loss": 0.8345,
+      "step": 8560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.655728983356372,
+      "learning_rate": 4.467737782030951e-06,
+      "loss": 0.7623,
+      "step": 8561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4915994614353707,
+      "learning_rate": 4.466707195823707e-06,
+      "loss": 0.9622,
+      "step": 8562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2699182166720384,
+      "learning_rate": 4.465676632533245e-06,
+      "loss": 0.7364,
+      "step": 8563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0807544376879816,
+      "learning_rate": 4.464646092203846e-06,
+      "loss": 0.7344,
+      "step": 8564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2539012162868226,
+      "learning_rate": 4.463615574879798e-06,
+      "loss": 0.7127,
+      "step": 8565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.04931445557265,
+      "learning_rate": 4.462585080605384e-06,
+      "loss": 0.6962,
+      "step": 8566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.012642572074873,
+      "learning_rate": 4.461554609424884e-06,
+      "loss": 0.6922,
+      "step": 8567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.076452397827735,
+      "learning_rate": 4.460524161382582e-06,
+      "loss": 0.7191,
+      "step": 8568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.477953480832426,
+      "learning_rate": 4.459493736522759e-06,
+      "loss": 0.7085,
+      "step": 8569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6440021453879847,
+      "learning_rate": 4.4584633348896945e-06,
+      "loss": 0.7905,
+      "step": 8570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.41895776566258,
+      "learning_rate": 4.457432956527665e-06,
+      "loss": 0.8796,
+      "step": 8571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.551352737165428,
+      "learning_rate": 4.456402601480949e-06,
+      "loss": 0.821,
+      "step": 8572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.156262621008799,
+      "learning_rate": 4.455372269793826e-06,
+      "loss": 0.754,
+      "step": 8573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0305644913353174,
+      "learning_rate": 4.4543419615105685e-06,
+      "loss": 0.7285,
+      "step": 8574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8201248434524238,
+      "learning_rate": 4.453311676675453e-06,
+      "loss": 0.8754,
+      "step": 8575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9076727825717015,
+      "learning_rate": 4.452281415332751e-06,
+      "loss": 1.0174,
+      "step": 8576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.870031846483881,
+      "learning_rate": 4.451251177526738e-06,
+      "loss": 0.8666,
+      "step": 8577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8283542714357526,
+      "learning_rate": 4.450220963301683e-06,
+      "loss": 0.9695,
+      "step": 8578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.284591576267694,
+      "learning_rate": 4.449190772701857e-06,
+      "loss": 0.7351,
+      "step": 8579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2584557693085743,
+      "learning_rate": 4.44816060577153e-06,
+      "loss": 0.7889,
+      "step": 8580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0864138450683538,
+      "learning_rate": 4.447130462554974e-06,
+      "loss": 0.731,
+      "step": 8581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1016460062617965,
+      "learning_rate": 4.44610034309645e-06,
+      "loss": 0.7061,
+      "step": 8582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8611532910125026,
+      "learning_rate": 4.4450702474402295e-06,
+      "loss": 0.9035,
+      "step": 8583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.192388852192989,
+      "learning_rate": 4.444040175630577e-06,
+      "loss": 0.9229,
+      "step": 8584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2938417474306334,
+      "learning_rate": 4.443010127711755e-06,
+      "loss": 0.8988,
+      "step": 8585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.607520843434854,
+      "learning_rate": 4.441980103728027e-06,
+      "loss": 0.7802,
+      "step": 8586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2621544666151236,
+      "learning_rate": 4.440950103723658e-06,
+      "loss": 0.7793,
+      "step": 8587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0696825699436183,
+      "learning_rate": 4.439920127742909e-06,
+      "loss": 0.7911,
+      "step": 8588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0351100149331196,
+      "learning_rate": 4.438890175830039e-06,
+      "loss": 0.8146,
+      "step": 8589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2509526538357787,
+      "learning_rate": 4.437860248029307e-06,
+      "loss": 1.019,
+      "step": 8590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.247840897816768,
+      "learning_rate": 4.4368303443849735e-06,
+      "loss": 0.6916,
+      "step": 8591
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0217131214186264,
+      "learning_rate": 4.435800464941292e-06,
+      "loss": 0.8519,
+      "step": 8592
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.429713056539012,
+      "learning_rate": 4.434770609742523e-06,
+      "loss": 0.6666,
+      "step": 8593
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1449854205213077,
+      "learning_rate": 4.433740778832919e-06,
+      "loss": 0.9821,
+      "step": 8594
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0116873729822515,
+      "learning_rate": 4.432710972256737e-06,
+      "loss": 0.7476,
+      "step": 8595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.168193228366865,
+      "learning_rate": 4.431681190058224e-06,
+      "loss": 0.827,
+      "step": 8596
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9322530906181172,
+      "learning_rate": 4.430651432281639e-06,
+      "loss": 0.7314,
+      "step": 8597
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0812636243416869,
+      "learning_rate": 4.429621698971228e-06,
+      "loss": 0.5557,
+      "step": 8598
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9723135046168707,
+      "learning_rate": 4.428591990171246e-06,
+      "loss": 0.8799,
+      "step": 8599
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2530959560460575,
+      "learning_rate": 4.4275623059259355e-06,
+      "loss": 0.7734,
+      "step": 8600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0367186049162416,
+      "learning_rate": 4.426532646279548e-06,
+      "loss": 0.6215,
+      "step": 8601
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1354775573366556,
+      "learning_rate": 4.425503011276332e-06,
+      "loss": 0.6204,
+      "step": 8602
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3759889633441107,
+      "learning_rate": 4.42447340096053e-06,
+      "loss": 0.6558,
+      "step": 8603
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9810468464090403,
+      "learning_rate": 4.423443815376387e-06,
+      "loss": 0.9834,
+      "step": 8604
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0555354330463709,
+      "learning_rate": 4.422414254568147e-06,
+      "loss": 0.5501,
+      "step": 8605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6191001882704645,
+      "learning_rate": 4.421384718580055e-06,
+      "loss": 0.7614,
+      "step": 8606
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.998457742676599,
+      "learning_rate": 4.420355207456349e-06,
+      "loss": 0.9601,
+      "step": 8607
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1620183515212084,
+      "learning_rate": 4.4193257212412695e-06,
+      "loss": 0.9226,
+      "step": 8608
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.271775461119913,
+      "learning_rate": 4.41829625997906e-06,
+      "loss": 0.6721,
+      "step": 8609
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03992726680075,
+      "learning_rate": 4.417266823713953e-06,
+      "loss": 0.8789,
+      "step": 8610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3079673385616712,
+      "learning_rate": 4.416237412490189e-06,
+      "loss": 0.8389,
+      "step": 8611
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.461562518806499,
+      "learning_rate": 4.415208026352003e-06,
+      "loss": 0.8362,
+      "step": 8612
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9125485032663105,
+      "learning_rate": 4.414178665343633e-06,
+      "loss": 0.6875,
+      "step": 8613
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.06915068199687,
+      "learning_rate": 4.413149329509307e-06,
+      "loss": 0.8419,
+      "step": 8614
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7716047856609323,
+      "learning_rate": 4.412120018893263e-06,
+      "loss": 0.9242,
+      "step": 8615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1713076482562217,
+      "learning_rate": 4.411090733539731e-06,
+      "loss": 0.6047,
+      "step": 8616
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0477944517620026,
+      "learning_rate": 4.410061473492943e-06,
+      "loss": 0.6626,
+      "step": 8617
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.172613593827419,
+      "learning_rate": 4.409032238797125e-06,
+      "loss": 0.9271,
+      "step": 8618
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.577486586568226,
+      "learning_rate": 4.4080030294965085e-06,
+      "loss": 0.8328,
+      "step": 8619
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.399718074768047,
+      "learning_rate": 4.406973845635322e-06,
+      "loss": 0.6667,
+      "step": 8620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.150099382404684,
+      "learning_rate": 4.405944687257789e-06,
+      "loss": 0.7928,
+      "step": 8621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9648673878109555,
+      "learning_rate": 4.4049155544081355e-06,
+      "loss": 0.5437,
+      "step": 8622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3619206859621804,
+      "learning_rate": 4.4038864471305845e-06,
+      "loss": 0.9048,
+      "step": 8623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.085307925428524,
+      "learning_rate": 4.402857365469364e-06,
+      "loss": 0.7921,
+      "step": 8624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.541350166521125,
+      "learning_rate": 4.40182830946869e-06,
+      "loss": 0.9058,
+      "step": 8625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0232026686259073,
+      "learning_rate": 4.400799279172786e-06,
+      "loss": 0.5717,
+      "step": 8626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.137662209286804,
+      "learning_rate": 4.399770274625871e-06,
+      "loss": 0.7189,
+      "step": 8627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9720769875569053,
+      "learning_rate": 4.3987412958721664e-06,
+      "loss": 0.7509,
+      "step": 8628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7206814201959393,
+      "learning_rate": 4.397712342955885e-06,
+      "loss": 0.9861,
+      "step": 8629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.076668852854691,
+      "learning_rate": 4.396683415921247e-06,
+      "loss": 0.7817,
+      "step": 8630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.022052402672946,
+      "learning_rate": 4.3956545148124665e-06,
+      "loss": 0.8815,
+      "step": 8631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3752924624824945,
+      "learning_rate": 4.394625639673756e-06,
+      "loss": 1.0877,
+      "step": 8632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5134435743391204,
+      "learning_rate": 4.39359679054933e-06,
+      "loss": 1.0862,
+      "step": 8633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0867470871999676,
+      "learning_rate": 4.392567967483401e-06,
+      "loss": 0.8666,
+      "step": 8634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5232767385190127,
+      "learning_rate": 4.3915391705201805e-06,
+      "loss": 0.8205,
+      "step": 8635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.118753745581955,
+      "learning_rate": 4.390510399703875e-06,
+      "loss": 0.783,
+      "step": 8636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.147436103010902,
+      "learning_rate": 4.389481655078695e-06,
+      "loss": 0.8393,
+      "step": 8637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9788745480310628,
+      "learning_rate": 4.38845293668885e-06,
+      "loss": 0.8424,
+      "step": 8638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.060536917053715,
+      "learning_rate": 4.387424244578543e-06,
+      "loss": 0.7859,
+      "step": 8639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.416419069957067,
+      "learning_rate": 4.386395578791981e-06,
+      "loss": 0.8069,
+      "step": 8640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9566579659217327,
+      "learning_rate": 4.385366939373365e-06,
+      "loss": 0.6693,
+      "step": 8641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2271422556077436,
+      "learning_rate": 4.384338326366905e-06,
+      "loss": 0.7483,
+      "step": 8642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1374779329270668,
+      "learning_rate": 4.383309739816795e-06,
+      "loss": 0.8189,
+      "step": 8643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.041907974525692,
+      "learning_rate": 4.38228117976724e-06,
+      "loss": 0.7278,
+      "step": 8644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0531053184292531,
+      "learning_rate": 4.381252646262437e-06,
+      "loss": 0.6691,
+      "step": 8645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1259436993028276,
+      "learning_rate": 4.3802241393465885e-06,
+      "loss": 0.7649,
+      "step": 8646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5660931004250855,
+      "learning_rate": 4.3791956590638866e-06,
+      "loss": 0.8672,
+      "step": 8647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.272593255782078,
+      "learning_rate": 4.378167205458531e-06,
+      "loss": 0.8072,
+      "step": 8648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.125009927314158,
+      "learning_rate": 4.377138778574716e-06,
+      "loss": 0.8741,
+      "step": 8649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4585520511550394,
+      "learning_rate": 4.376110378456634e-06,
+      "loss": 0.8601,
+      "step": 8650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2295245525257723,
+      "learning_rate": 4.375082005148479e-06,
+      "loss": 1.0372,
+      "step": 8651
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9694209657953918,
+      "learning_rate": 4.37405365869444e-06,
+      "loss": 0.8015,
+      "step": 8652
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0799324627219646,
+      "learning_rate": 4.373025339138713e-06,
+      "loss": 0.6662,
+      "step": 8653
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.120099064035063,
+      "learning_rate": 4.371997046525481e-06,
+      "loss": 0.7671,
+      "step": 8654
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8095869784656762,
+      "learning_rate": 4.370968780898935e-06,
+      "loss": 0.8349,
+      "step": 8655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.499539483395979,
+      "learning_rate": 4.3699405423032625e-06,
+      "loss": 0.6049,
+      "step": 8656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.364039487200678,
+      "learning_rate": 4.368912330782647e-06,
+      "loss": 0.8195,
+      "step": 8657
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.056044389861105,
+      "learning_rate": 4.367884146381274e-06,
+      "loss": 0.758,
+      "step": 8658
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8921749250730215,
+      "learning_rate": 4.366855989143326e-06,
+      "loss": 0.7907,
+      "step": 8659
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3842394331957375,
+      "learning_rate": 4.365827859112989e-06,
+      "loss": 0.9124,
+      "step": 8660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.644379157763721,
+      "learning_rate": 4.364799756334439e-06,
+      "loss": 0.8781,
+      "step": 8661
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.725404143616811,
+      "learning_rate": 4.3637716808518596e-06,
+      "loss": 0.8207,
+      "step": 8662
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9070158033859128,
+      "learning_rate": 4.362743632709426e-06,
+      "loss": 0.822,
+      "step": 8663
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0767475373450273,
+      "learning_rate": 4.3617156119513206e-06,
+      "loss": 0.8398,
+      "step": 8664
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8272160056459097,
+      "learning_rate": 4.360687618621715e-06,
+      "loss": 0.8913,
+      "step": 8665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8735588320936385,
+      "learning_rate": 4.359659652764786e-06,
+      "loss": 0.6462,
+      "step": 8666
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9763798617233062,
+      "learning_rate": 4.35863171442471e-06,
+      "loss": 0.6949,
+      "step": 8667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.719808994103021,
+      "learning_rate": 4.357603803645657e-06,
+      "loss": 0.8358,
+      "step": 8668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.107522457869771,
+      "learning_rate": 4.356575920471796e-06,
+      "loss": 0.9076,
+      "step": 8669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4831233142085742,
+      "learning_rate": 4.355548064947303e-06,
+      "loss": 0.7902,
+      "step": 8670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3052292831761205,
+      "learning_rate": 4.354520237116347e-06,
+      "loss": 0.8166,
+      "step": 8671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0731343817455659,
+      "learning_rate": 4.353492437023092e-06,
+      "loss": 0.5873,
+      "step": 8672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7504237617981975,
+      "learning_rate": 4.352464664711706e-06,
+      "loss": 0.8397,
+      "step": 8673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.094451530710377,
+      "learning_rate": 4.351436920226357e-06,
+      "loss": 0.851,
+      "step": 8674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3151554431393624,
+      "learning_rate": 4.350409203611207e-06,
+      "loss": 0.6607,
+      "step": 8675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1515726481058628,
+      "learning_rate": 4.34938151491042e-06,
+      "loss": 0.6935,
+      "step": 8676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.513962483845536,
+      "learning_rate": 4.348353854168158e-06,
+      "loss": 0.7909,
+      "step": 8677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2708847007213433,
+      "learning_rate": 4.347326221428585e-06,
+      "loss": 0.8056,
+      "step": 8678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.103376780775161,
+      "learning_rate": 4.346298616735855e-06,
+      "loss": 0.938,
+      "step": 8679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.823351499318559,
+      "learning_rate": 4.345271040134129e-06,
+      "loss": 0.9085,
+      "step": 8680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.501092530649925,
+      "learning_rate": 4.344243491667566e-06,
+      "loss": 0.6745,
+      "step": 8681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0935685572046014,
+      "learning_rate": 4.343215971380323e-06,
+      "loss": 0.734,
+      "step": 8682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1182977802120457,
+      "learning_rate": 4.342188479316548e-06,
+      "loss": 0.843,
+      "step": 8683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.416492540114239,
+      "learning_rate": 4.3411610155204e-06,
+      "loss": 0.6101,
+      "step": 8684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7458249186299337,
+      "learning_rate": 4.340133580036033e-06,
+      "loss": 0.8607,
+      "step": 8685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0408020669483795,
+      "learning_rate": 4.339106172907594e-06,
+      "loss": 0.6989,
+      "step": 8686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1906087017988765,
+      "learning_rate": 4.338078794179234e-06,
+      "loss": 0.6763,
+      "step": 8687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8413652028128014,
+      "learning_rate": 4.337051443895102e-06,
+      "loss": 0.7568,
+      "step": 8688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.291687004569664,
+      "learning_rate": 4.336024122099348e-06,
+      "loss": 0.8776,
+      "step": 8689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.933362228152818,
+      "learning_rate": 4.334996828836115e-06,
+      "loss": 0.9538,
+      "step": 8690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.264997584812418,
+      "learning_rate": 4.333969564149549e-06,
+      "loss": 0.7624,
+      "step": 8691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.917283220674988,
+      "learning_rate": 4.332942328083796e-06,
+      "loss": 0.7814,
+      "step": 8692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1916403414242267,
+      "learning_rate": 4.3319151206829955e-06,
+      "loss": 0.7742,
+      "step": 8693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.198388505757368,
+      "learning_rate": 4.330887941991288e-06,
+      "loss": 0.9013,
+      "step": 8694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.959093086302977,
+      "learning_rate": 4.329860792052819e-06,
+      "loss": 0.7215,
+      "step": 8695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0988120155797625,
+      "learning_rate": 4.3288336709117246e-06,
+      "loss": 0.8111,
+      "step": 8696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.45958716875849,
+      "learning_rate": 4.32780657861214e-06,
+      "loss": 0.7888,
+      "step": 8697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9830013411871632,
+      "learning_rate": 4.326779515198203e-06,
+      "loss": 1.0427,
+      "step": 8698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.574491953520629,
+      "learning_rate": 4.325752480714052e-06,
+      "loss": 0.8047,
+      "step": 8699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1135353358220432,
+      "learning_rate": 4.324725475203818e-06,
+      "loss": 0.7453,
+      "step": 8700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1600548692795947,
+      "learning_rate": 4.323698498711634e-06,
+      "loss": 0.8846,
+      "step": 8701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.440446472879541,
+      "learning_rate": 4.32267155128163e-06,
+      "loss": 0.7866,
+      "step": 8702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2121906478402193,
+      "learning_rate": 4.321644632957941e-06,
+      "loss": 0.711,
+      "step": 8703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.255645658134479,
+      "learning_rate": 4.320617743784691e-06,
+      "loss": 0.8662,
+      "step": 8704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.056562828831023,
+      "learning_rate": 4.31959088380601e-06,
+      "loss": 0.5913,
+      "step": 8705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2723376718485944,
+      "learning_rate": 4.318564053066023e-06,
+      "loss": 0.7265,
+      "step": 8706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.979510423176679,
+      "learning_rate": 4.317537251608859e-06,
+      "loss": 0.7153,
+      "step": 8707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9175847020915162,
+      "learning_rate": 4.316510479478636e-06,
+      "loss": 0.7973,
+      "step": 8708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.193175265092701,
+      "learning_rate": 4.315483736719482e-06,
+      "loss": 0.823,
+      "step": 8709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3533922673561474,
+      "learning_rate": 4.314457023375517e-06,
+      "loss": 0.7455,
+      "step": 8710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1057339351840723,
+      "learning_rate": 4.313430339490859e-06,
+      "loss": 0.8789,
+      "step": 8711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0051339083989679,
+      "learning_rate": 4.312403685109627e-06,
+      "loss": 0.6599,
+      "step": 8712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8619782448231543,
+      "learning_rate": 4.311377060275942e-06,
+      "loss": 0.8672,
+      "step": 8713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0258731161102785,
+      "learning_rate": 4.310350465033919e-06,
+      "loss": 0.8797,
+      "step": 8714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.331033946708023,
+      "learning_rate": 4.309323899427671e-06,
+      "loss": 0.9406,
+      "step": 8715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5020151009109206,
+      "learning_rate": 4.308297363501314e-06,
+      "loss": 0.9877,
+      "step": 8716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9366620256261995,
+      "learning_rate": 4.3072708572989585e-06,
+      "loss": 0.8419,
+      "step": 8717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0934716503085666,
+      "learning_rate": 4.306244380864719e-06,
+      "loss": 0.6844,
+      "step": 8718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9528575443162257,
+      "learning_rate": 4.305217934242703e-06,
+      "loss": 0.9701,
+      "step": 8719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.090726086762507,
+      "learning_rate": 4.304191517477019e-06,
+      "loss": 0.8068,
+      "step": 8720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.532283612144906,
+      "learning_rate": 4.303165130611776e-06,
+      "loss": 0.8276,
+      "step": 8721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1286374506850723,
+      "learning_rate": 4.302138773691079e-06,
+      "loss": 0.6066,
+      "step": 8722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8691960368776974,
+      "learning_rate": 4.301112446759033e-06,
+      "loss": 0.8109,
+      "step": 8723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328789758967382,
+      "learning_rate": 4.30008614985974e-06,
+      "loss": 0.9475,
+      "step": 8724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2058577444770107,
+      "learning_rate": 4.299059883037308e-06,
+      "loss": 0.6888,
+      "step": 8725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.305443131235101,
+      "learning_rate": 4.2980336463358296e-06,
+      "loss": 0.8668,
+      "step": 8726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9259512583980964,
+      "learning_rate": 4.29700743979941e-06,
+      "loss": 0.8365,
+      "step": 8727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1292090243885253,
+      "learning_rate": 4.295981263472148e-06,
+      "loss": 0.7845,
+      "step": 8728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5612255005543068,
+      "learning_rate": 4.294955117398139e-06,
+      "loss": 0.7335,
+      "step": 8729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3372264464347645,
+      "learning_rate": 4.293929001621477e-06,
+      "loss": 0.6307,
+      "step": 8730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9230971628752124,
+      "learning_rate": 4.2929029161862575e-06,
+      "loss": 0.6053,
+      "step": 8731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0226440449792438,
+      "learning_rate": 4.291876861136578e-06,
+      "loss": 0.7023,
+      "step": 8732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8455563785008855,
+      "learning_rate": 4.290850836516526e-06,
+      "loss": 0.6847,
+      "step": 8733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7721852414245607,
+      "learning_rate": 4.28982484237019e-06,
+      "loss": 0.7799,
+      "step": 8734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3111890217990823,
+      "learning_rate": 4.288798878741664e-06,
+      "loss": 0.7189,
+      "step": 8735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9263976870630837,
+      "learning_rate": 4.287772945675035e-06,
+      "loss": 0.8865,
+      "step": 8736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8230411446624193,
+      "learning_rate": 4.286747043214388e-06,
+      "loss": 0.8558,
+      "step": 8737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0260401915703725,
+      "learning_rate": 4.285721171403809e-06,
+      "loss": 0.8536,
+      "step": 8738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2820187217179966,
+      "learning_rate": 4.284695330287383e-06,
+      "loss": 0.694,
+      "step": 8739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4262284619971255,
+      "learning_rate": 4.283669519909189e-06,
+      "loss": 0.9291,
+      "step": 8740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.41404086357814,
+      "learning_rate": 4.282643740313312e-06,
+      "loss": 0.8291,
+      "step": 8741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.647822696498617,
+      "learning_rate": 4.281617991543832e-06,
+      "loss": 0.7344,
+      "step": 8742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.929562416777994,
+      "learning_rate": 4.280592273644829e-06,
+      "loss": 1.0308,
+      "step": 8743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9891171192929407,
+      "learning_rate": 4.279566586660375e-06,
+      "loss": 0.8952,
+      "step": 8744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.364526886729992,
+      "learning_rate": 4.278540930634549e-06,
+      "loss": 0.8798,
+      "step": 8745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2936443428014672,
+      "learning_rate": 4.277515305611427e-06,
+      "loss": 0.7936,
+      "step": 8746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.552399123486727,
+      "learning_rate": 4.276489711635083e-06,
+      "loss": 0.6872,
+      "step": 8747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1862718579439315,
+      "learning_rate": 4.275464148749585e-06,
+      "loss": 0.866,
+      "step": 8748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2223458267225547,
+      "learning_rate": 4.274438616999007e-06,
+      "loss": 0.6706,
+      "step": 8749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.118915828739957,
+      "learning_rate": 4.273413116427419e-06,
+      "loss": 0.8664,
+      "step": 8750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0923049883878364,
+      "learning_rate": 4.272387647078885e-06,
+      "loss": 0.8115,
+      "step": 8751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3217670280521956,
+      "learning_rate": 4.271362208997476e-06,
+      "loss": 0.854,
+      "step": 8752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.154302073593988,
+      "learning_rate": 4.270336802227255e-06,
+      "loss": 0.6911,
+      "step": 8753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8933219713078568,
+      "learning_rate": 4.269311426812287e-06,
+      "loss": 1.0884,
+      "step": 8754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3986198594280577,
+      "learning_rate": 4.268286082796634e-06,
+      "loss": 0.6758,
+      "step": 8755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0123683731599362,
+      "learning_rate": 4.267260770224358e-06,
+      "loss": 0.7276,
+      "step": 8756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.904791649706006,
+      "learning_rate": 4.26623548913952e-06,
+      "loss": 0.7674,
+      "step": 8757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.378944764786245,
+      "learning_rate": 4.2652102395861746e-06,
+      "loss": 0.7079,
+      "step": 8758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.266871419018355,
+      "learning_rate": 4.264185021608382e-06,
+      "loss": 0.735,
+      "step": 8759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0959753300921973,
+      "learning_rate": 4.263159835250199e-06,
+      "loss": 0.6691,
+      "step": 8760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9439567452390456,
+      "learning_rate": 4.26213468055568e-06,
+      "loss": 0.8596,
+      "step": 8761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.960945290945652,
+      "learning_rate": 4.2611095575688745e-06,
+      "loss": 0.7578,
+      "step": 8762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.445772037247374,
+      "learning_rate": 4.260084466333837e-06,
+      "loss": 0.664,
+      "step": 8763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.023779159966783,
+      "learning_rate": 4.259059406894619e-06,
+      "loss": 0.7964,
+      "step": 8764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2197153394279243,
+      "learning_rate": 4.258034379295268e-06,
+      "loss": 0.5997,
+      "step": 8765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2577005932213288,
+      "learning_rate": 4.257009383579833e-06,
+      "loss": 0.7551,
+      "step": 8766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1606016351047543,
+      "learning_rate": 4.255984419792358e-06,
+      "loss": 0.6667,
+      "step": 8767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234268123814571,
+      "learning_rate": 4.254959487976892e-06,
+      "loss": 0.8985,
+      "step": 8768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0822652705980262,
+      "learning_rate": 4.253934588177473e-06,
+      "loss": 0.8572,
+      "step": 8769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.079719044671894,
+      "learning_rate": 4.252909720438149e-06,
+      "loss": 0.7487,
+      "step": 8770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.491378838011629,
+      "learning_rate": 4.251884884802956e-06,
+      "loss": 0.9016,
+      "step": 8771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.813318919705775,
+      "learning_rate": 4.250860081315937e-06,
+      "loss": 0.8822,
+      "step": 8772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09285375534984,
+      "learning_rate": 4.2498353100211275e-06,
+      "loss": 0.9044,
+      "step": 8773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4497164771006568,
+      "learning_rate": 4.248810570962567e-06,
+      "loss": 0.724,
+      "step": 8774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9784831980293826,
+      "learning_rate": 4.247785864184289e-06,
+      "loss": 0.6151,
+      "step": 8775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.302630009701562,
+      "learning_rate": 4.246761189730327e-06,
+      "loss": 0.7881,
+      "step": 8776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4811894217288586,
+      "learning_rate": 4.245736547644714e-06,
+      "loss": 0.8815,
+      "step": 8777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9208229838237185,
+      "learning_rate": 4.2447119379714805e-06,
+      "loss": 0.9553,
+      "step": 8778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7638975864012023,
+      "learning_rate": 4.24368736075466e-06,
+      "loss": 0.9013,
+      "step": 8779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.571953055973867,
+      "learning_rate": 4.242662816038276e-06,
+      "loss": 0.8192,
+      "step": 8780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1654904006769238,
+      "learning_rate": 4.241638303866358e-06,
+      "loss": 0.7478,
+      "step": 8781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1919423161516147,
+      "learning_rate": 4.24061382428293e-06,
+      "loss": 0.8453,
+      "step": 8782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9709960566187742,
+      "learning_rate": 4.23958937733202e-06,
+      "loss": 0.7065,
+      "step": 8783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8710000615668225,
+      "learning_rate": 4.238564963057646e-06,
+      "loss": 0.7462,
+      "step": 8784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2265584219824206,
+      "learning_rate": 4.237540581503831e-06,
+      "loss": 0.7375,
+      "step": 8785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.554824689862581,
+      "learning_rate": 4.236516232714597e-06,
+      "loss": 0.8031,
+      "step": 8786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0265275042916153,
+      "learning_rate": 4.235491916733959e-06,
+      "loss": 0.8903,
+      "step": 8787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7466310205157016,
+      "learning_rate": 4.234467633605937e-06,
+      "loss": 0.9136,
+      "step": 8788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.425332120883534,
+      "learning_rate": 4.233443383374545e-06,
+      "loss": 0.7329,
+      "step": 8789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9898314596130866,
+      "learning_rate": 4.232419166083799e-06,
+      "loss": 0.8662,
+      "step": 8790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1418475836792013,
+      "learning_rate": 4.23139498177771e-06,
+      "loss": 0.8241,
+      "step": 8791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.191060651082542,
+      "learning_rate": 4.23037083050029e-06,
+      "loss": 0.8689,
+      "step": 8792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0183336993772745,
+      "learning_rate": 4.229346712295551e-06,
+      "loss": 0.5909,
+      "step": 8793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6971728348712243,
+      "learning_rate": 4.228322627207499e-06,
+      "loss": 0.8662,
+      "step": 8794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.200912840106264,
+      "learning_rate": 4.227298575280142e-06,
+      "loss": 0.6884,
+      "step": 8795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8887955452616967,
+      "learning_rate": 4.226274556557487e-06,
+      "loss": 0.765,
+      "step": 8796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7105327939527182,
+      "learning_rate": 4.225250571083538e-06,
+      "loss": 0.6068,
+      "step": 8797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2543724590866425,
+      "learning_rate": 4.2242266189022975e-06,
+      "loss": 0.9382,
+      "step": 8798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1915511877303464,
+      "learning_rate": 4.223202700057765e-06,
+      "loss": 0.8929,
+      "step": 8799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4439609320838054,
+      "learning_rate": 4.2221788145939425e-06,
+      "loss": 0.7251,
+      "step": 8800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.943320159225328,
+      "learning_rate": 4.221154962554831e-06,
+      "loss": 0.7777,
+      "step": 8801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0749426043219312,
+      "learning_rate": 4.220131143984424e-06,
+      "loss": 0.9076,
+      "step": 8802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.280238061051281,
+      "learning_rate": 4.219107358926718e-06,
+      "loss": 0.7867,
+      "step": 8803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3325866255058667,
+      "learning_rate": 4.21808360742571e-06,
+      "loss": 0.6824,
+      "step": 8804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6366564142364566,
+      "learning_rate": 4.217059889525389e-06,
+      "loss": 0.7212,
+      "step": 8805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.66715820372282,
+      "learning_rate": 4.216036205269748e-06,
+      "loss": 0.979,
+      "step": 8806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3075137155118175,
+      "learning_rate": 4.215012554702778e-06,
+      "loss": 0.7822,
+      "step": 8807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0737864151085397,
+      "learning_rate": 4.213988937868469e-06,
+      "loss": 0.6843,
+      "step": 8808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.045439695654747,
+      "learning_rate": 4.212965354810802e-06,
+      "loss": 0.8377,
+      "step": 8809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0028190934686694,
+      "learning_rate": 4.211941805573767e-06,
+      "loss": 0.8575,
+      "step": 8810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.223393409960287,
+      "learning_rate": 4.21091829020135e-06,
+      "loss": 0.5987,
+      "step": 8811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.220179719063524,
+      "learning_rate": 4.209894808737531e-06,
+      "loss": 0.772,
+      "step": 8812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0693511676886525,
+      "learning_rate": 4.20887136122629e-06,
+      "loss": 0.7667,
+      "step": 8813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.108289846130056,
+      "learning_rate": 4.207847947711609e-06,
+      "loss": 0.9124,
+      "step": 8814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2395398890929377,
+      "learning_rate": 4.206824568237468e-06,
+      "loss": 0.8378,
+      "step": 8815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.074779565297696,
+      "learning_rate": 4.205801222847839e-06,
+      "loss": 0.7017,
+      "step": 8816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4383431071238286,
+      "learning_rate": 4.204777911586702e-06,
+      "loss": 0.79,
+      "step": 8817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.837521648397956,
+      "learning_rate": 4.203754634498027e-06,
+      "loss": 0.6897,
+      "step": 8818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1702353210715186,
+      "learning_rate": 4.202731391625793e-06,
+      "loss": 0.8028,
+      "step": 8819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1560016391189145,
+      "learning_rate": 4.201708183013963e-06,
+      "loss": 0.6935,
+      "step": 8820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375201551914784,
+      "learning_rate": 4.200685008706511e-06,
+      "loss": 0.8895,
+      "step": 8821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6502639159430963,
+      "learning_rate": 4.199661868747406e-06,
+      "loss": 0.8556,
+      "step": 8822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.383090811579424,
+      "learning_rate": 4.198638763180611e-06,
+      "loss": 0.7027,
+      "step": 8823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4035825319926514,
+      "learning_rate": 4.1976156920500935e-06,
+      "loss": 0.7385,
+      "step": 8824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1464481790758914,
+      "learning_rate": 4.196592655399816e-06,
+      "loss": 0.8802,
+      "step": 8825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.06342427044473,
+      "learning_rate": 4.195569653273743e-06,
+      "loss": 1.0667,
+      "step": 8826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1748174862448337,
+      "learning_rate": 4.1945466857158336e-06,
+      "loss": 0.7956,
+      "step": 8827
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6807554546783017,
+      "learning_rate": 4.193523752770045e-06,
+      "loss": 0.8175,
+      "step": 8828
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.641535857899166,
+      "learning_rate": 4.19250085448034e-06,
+      "loss": 0.9912,
+      "step": 8829
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0995562147820475,
+      "learning_rate": 4.191477990890668e-06,
+      "loss": 0.8157,
+      "step": 8830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.145603752615203,
+      "learning_rate": 4.19045516204499e-06,
+      "loss": 0.8595,
+      "step": 8831
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5054996000193066,
+      "learning_rate": 4.189432367987255e-06,
+      "loss": 0.719,
+      "step": 8832
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.348188698046692,
+      "learning_rate": 4.188409608761419e-06,
+      "loss": 0.8195,
+      "step": 8833
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6269933100649205,
+      "learning_rate": 4.187386884411426e-06,
+      "loss": 0.7683,
+      "step": 8834
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2339256611134233,
+      "learning_rate": 4.186364194981231e-06,
+      "loss": 0.9163,
+      "step": 8835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.168446905197613,
+      "learning_rate": 4.1853415405147765e-06,
+      "loss": 0.9382,
+      "step": 8836
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1716370887700425,
+      "learning_rate": 4.184318921056013e-06,
+      "loss": 0.7951,
+      "step": 8837
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3212351496615775,
+      "learning_rate": 4.18329633664888e-06,
+      "loss": 0.9272,
+      "step": 8838
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.237091598158292,
+      "learning_rate": 4.182273787337323e-06,
+      "loss": 0.6354,
+      "step": 8839
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7134682982740777,
+      "learning_rate": 4.181251273165283e-06,
+      "loss": 0.689,
+      "step": 8840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.226223441427891,
+      "learning_rate": 4.1802287941767e-06,
+      "loss": 0.8224,
+      "step": 8841
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.928386268250444,
+      "learning_rate": 4.179206350415509e-06,
+      "loss": 0.7401,
+      "step": 8842
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0488982210113247,
+      "learning_rate": 4.17818394192565e-06,
+      "loss": 0.8828,
+      "step": 8843
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.227188786824046,
+      "learning_rate": 4.177161568751058e-06,
+      "loss": 0.7073,
+      "step": 8844
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9640573347268275,
+      "learning_rate": 4.176139230935666e-06,
+      "loss": 0.8545,
+      "step": 8845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.016963130059186,
+      "learning_rate": 4.175116928523405e-06,
+      "loss": 0.8034,
+      "step": 8846
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.279871113014191,
+      "learning_rate": 4.174094661558209e-06,
+      "loss": 0.7135,
+      "step": 8847
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6927251543173796,
+      "learning_rate": 4.173072430084002e-06,
+      "loss": 0.679,
+      "step": 8848
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0542777512961523,
+      "learning_rate": 4.172050234144716e-06,
+      "loss": 0.9327,
+      "step": 8849
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0752347907738087,
+      "learning_rate": 4.171028073784274e-06,
+      "loss": 0.8499,
+      "step": 8850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.240145086637045,
+      "learning_rate": 4.170005949046604e-06,
+      "loss": 0.9288,
+      "step": 8851
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0087497534355805,
+      "learning_rate": 4.168983859975625e-06,
+      "loss": 0.894,
+      "step": 8852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9338358219557905,
+      "learning_rate": 4.16796180661526e-06,
+      "loss": 0.7624,
+      "step": 8853
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.249176297669427,
+      "learning_rate": 4.16693978900943e-06,
+      "loss": 0.7996,
+      "step": 8854
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.545390799533865,
+      "learning_rate": 4.165917807202055e-06,
+      "loss": 0.7837,
+      "step": 8855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.098025216232705,
+      "learning_rate": 4.164895861237046e-06,
+      "loss": 0.8525,
+      "step": 8856
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.317882852220666,
+      "learning_rate": 4.1638739511583224e-06,
+      "loss": 0.9089,
+      "step": 8857
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.261801526302563,
+      "learning_rate": 4.1628520770097994e-06,
+      "loss": 1.0517,
+      "step": 8858
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9719673617578883,
+      "learning_rate": 4.161830238835386e-06,
+      "loss": 0.9398,
+      "step": 8859
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8167647156206783,
+      "learning_rate": 4.160808436678992e-06,
+      "loss": 0.8233,
+      "step": 8860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0463278266923717,
+      "learning_rate": 4.1597866705845295e-06,
+      "loss": 0.8244,
+      "step": 8861
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9064857346486124,
+      "learning_rate": 4.1587649405959065e-06,
+      "loss": 0.8169,
+      "step": 8862
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2778348740287624,
+      "learning_rate": 4.157743246757026e-06,
+      "loss": 0.6916,
+      "step": 8863
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1104251743798237,
+      "learning_rate": 4.156721589111794e-06,
+      "loss": 0.8871,
+      "step": 8864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2651198841034204,
+      "learning_rate": 4.155699967704113e-06,
+      "loss": 0.6674,
+      "step": 8865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5303951236730566,
+      "learning_rate": 4.154678382577887e-06,
+      "loss": 0.7412,
+      "step": 8866
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9612822685447213,
+      "learning_rate": 4.1536568337770114e-06,
+      "loss": 0.8668,
+      "step": 8867
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.326308502514908,
+      "learning_rate": 4.152635321345387e-06,
+      "loss": 1.0368,
+      "step": 8868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.583509523210767,
+      "learning_rate": 4.151613845326912e-06,
+      "loss": 0.7699,
+      "step": 8869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1544295322161573,
+      "learning_rate": 4.150592405765476e-06,
+      "loss": 0.826,
+      "step": 8870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.745209270750841,
+      "learning_rate": 4.1495710027049776e-06,
+      "loss": 1.0038,
+      "step": 8871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.042811012152795,
+      "learning_rate": 4.148549636189308e-06,
+      "loss": 0.769,
+      "step": 8872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.216301487867239,
+      "learning_rate": 4.147528306262357e-06,
+      "loss": 0.9933,
+      "step": 8873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.333824609652981,
+      "learning_rate": 4.146507012968013e-06,
+      "loss": 0.8185,
+      "step": 8874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9196488235955929,
+      "learning_rate": 4.145485756350162e-06,
+      "loss": 0.7191,
+      "step": 8875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.413122092355999,
+      "learning_rate": 4.144464536452693e-06,
+      "loss": 0.7763,
+      "step": 8876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9790133608030884,
+      "learning_rate": 4.1434433533194865e-06,
+      "loss": 0.7655,
+      "step": 8877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.357160618492115,
+      "learning_rate": 4.142422206994428e-06,
+      "loss": 0.8277,
+      "step": 8878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.942089565609363,
+      "learning_rate": 4.141401097521396e-06,
+      "loss": 0.7467,
+      "step": 8879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.207660509318157,
+      "learning_rate": 4.140380024944273e-06,
+      "loss": 0.726,
+      "step": 8880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8268647710146944,
+      "learning_rate": 4.139358989306933e-06,
+      "loss": 0.7716,
+      "step": 8881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6122233044325673,
+      "learning_rate": 4.138337990653255e-06,
+      "loss": 0.8368,
+      "step": 8882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1815948364664184,
+      "learning_rate": 4.137317029027111e-06,
+      "loss": 0.8392,
+      "step": 8883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6181290903352479,
+      "learning_rate": 4.136296104472378e-06,
+      "loss": 0.6719,
+      "step": 8884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.31160603552362,
+      "learning_rate": 4.1352752170329236e-06,
+      "loss": 0.7489,
+      "step": 8885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0419731960429988,
+      "learning_rate": 4.1342543667526195e-06,
+      "loss": 0.5922,
+      "step": 8886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.542030672678135,
+      "learning_rate": 4.133233553675334e-06,
+      "loss": 0.8717,
+      "step": 8887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1769777744456227,
+      "learning_rate": 4.1322127778449315e-06,
+      "loss": 1.0445,
+      "step": 8888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7591462292537023,
+      "learning_rate": 4.131192039305278e-06,
+      "loss": 0.8291,
+      "step": 8889
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6839556892813934,
+      "learning_rate": 4.1301713381002394e-06,
+      "loss": 0.9393,
+      "step": 8890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6711687748567159,
+      "learning_rate": 4.129150674273675e-06,
+      "loss": 0.6757,
+      "step": 8891
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.406369213658981,
+      "learning_rate": 4.128130047869446e-06,
+      "loss": 1.0016,
+      "step": 8892
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984760311961497,
+      "learning_rate": 4.12710945893141e-06,
+      "loss": 0.5972,
+      "step": 8893
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.070629017208962,
+      "learning_rate": 4.1260889075034254e-06,
+      "loss": 0.7822,
+      "step": 8894
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1402114434289503,
+      "learning_rate": 4.125068393629346e-06,
+      "loss": 0.6553,
+      "step": 8895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9183279340541508,
+      "learning_rate": 4.124047917353026e-06,
+      "loss": 0.8227,
+      "step": 8896
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.071224942998601,
+      "learning_rate": 4.123027478718318e-06,
+      "loss": 0.8767,
+      "step": 8897
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1213560863950962,
+      "learning_rate": 4.1220070777690736e-06,
+      "loss": 0.6589,
+      "step": 8898
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0569732731547115,
+      "learning_rate": 4.120986714549139e-06,
+      "loss": 0.7359,
+      "step": 8899
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.997073232999956,
+      "learning_rate": 4.119966389102363e-06,
+      "loss": 0.8311,
+      "step": 8900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3338384779569434,
+      "learning_rate": 4.118946101472591e-06,
+      "loss": 0.7794,
+      "step": 8901
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7037582591009324,
+      "learning_rate": 4.117925851703669e-06,
+      "loss": 0.7392,
+      "step": 8902
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6988806301860153,
+      "learning_rate": 4.116905639839436e-06,
+      "loss": 0.7557,
+      "step": 8903
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5927093370671592,
+      "learning_rate": 4.115885465923734e-06,
+      "loss": 0.8313,
+      "step": 8904
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.583318450575263,
+      "learning_rate": 4.114865330000405e-06,
+      "loss": 0.8844,
+      "step": 8905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.163221113437769,
+      "learning_rate": 4.113845232113282e-06,
+      "loss": 0.6878,
+      "step": 8906
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03154243728841,
+      "learning_rate": 4.112825172306203e-06,
+      "loss": 0.8277,
+      "step": 8907
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.056647874242309,
+      "learning_rate": 4.111805150623002e-06,
+      "loss": 0.8314,
+      "step": 8908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.170503521161471,
+      "learning_rate": 4.110785167107514e-06,
+      "loss": 0.7179,
+      "step": 8909
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5459020717936194,
+      "learning_rate": 4.1097652218035666e-06,
+      "loss": 0.649,
+      "step": 8910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0995409921928467,
+      "learning_rate": 4.108745314754989e-06,
+      "loss": 0.8828,
+      "step": 8911
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.033936682644736,
+      "learning_rate": 4.107725446005612e-06,
+      "loss": 0.7883,
+      "step": 8912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9281132277922701,
+      "learning_rate": 4.106705615599258e-06,
+      "loss": 0.7943,
+      "step": 8913
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3358500647144025,
+      "learning_rate": 4.1056858235797545e-06,
+      "loss": 0.8344,
+      "step": 8914
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7620248324556629,
+      "learning_rate": 4.104666069990921e-06,
+      "loss": 0.8437,
+      "step": 8915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8028717189353187,
+      "learning_rate": 4.103646354876584e-06,
+      "loss": 0.7964,
+      "step": 8916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.161331161695739,
+      "learning_rate": 4.102626678280556e-06,
+      "loss": 0.7068,
+      "step": 8917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8902608048004217,
+      "learning_rate": 4.101607040246659e-06,
+      "loss": 0.6447,
+      "step": 8918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.077330633390969,
+      "learning_rate": 4.100587440818709e-06,
+      "loss": 0.6875,
+      "step": 8919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8036268315533042,
+      "learning_rate": 4.099567880040521e-06,
+      "loss": 0.7312,
+      "step": 8920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.001151088585636,
+      "learning_rate": 4.098548357955904e-06,
+      "loss": 0.6659,
+      "step": 8921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.92299332020749,
+      "learning_rate": 4.0975288746086724e-06,
+      "loss": 0.8518,
+      "step": 8922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2220394366037155,
+      "learning_rate": 4.0965094300426364e-06,
+      "loss": 0.8056,
+      "step": 8923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1435070085216241,
+      "learning_rate": 4.0954900243016016e-06,
+      "loss": 0.6869,
+      "step": 8924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2490669165192827,
+      "learning_rate": 4.094470657429374e-06,
+      "loss": 0.8317,
+      "step": 8925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1351967013304813,
+      "learning_rate": 4.093451329469758e-06,
+      "loss": 0.6005,
+      "step": 8926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.387987361129966,
+      "learning_rate": 4.092432040466561e-06,
+      "loss": 0.7738,
+      "step": 8927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.604574788518258,
+      "learning_rate": 4.091412790463578e-06,
+      "loss": 0.8211,
+      "step": 8928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.058052216490458,
+      "learning_rate": 4.090393579504612e-06,
+      "loss": 0.6912,
+      "step": 8929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.082034239672739,
+      "learning_rate": 4.0893744076334606e-06,
+      "loss": 1.0221,
+      "step": 8930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0569659061115677,
+      "learning_rate": 4.088355274893916e-06,
+      "loss": 0.765,
+      "step": 8931
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0935706187772096,
+      "learning_rate": 4.087336181329777e-06,
+      "loss": 0.6507,
+      "step": 8932
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4311175553285786,
+      "learning_rate": 4.086317126984835e-06,
+      "loss": 0.6825,
+      "step": 8933
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1827129120335913,
+      "learning_rate": 4.085298111902882e-06,
+      "loss": 0.6941,
+      "step": 8934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0663816699132456,
+      "learning_rate": 4.084279136127704e-06,
+      "loss": 0.6065,
+      "step": 8935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9130702479058717,
+      "learning_rate": 4.083260199703092e-06,
+      "loss": 0.9414,
+      "step": 8936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2986291650143564,
+      "learning_rate": 4.082241302672831e-06,
+      "loss": 0.8264,
+      "step": 8937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9501300448064811,
+      "learning_rate": 4.081222445080707e-06,
+      "loss": 0.9538,
+      "step": 8938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.525122770816009,
+      "learning_rate": 4.080203626970498e-06,
+      "loss": 0.8282,
+      "step": 8939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5747650919366576,
+      "learning_rate": 4.079184848385988e-06,
+      "loss": 0.7652,
+      "step": 8940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0869515889882038,
+      "learning_rate": 4.078166109370957e-06,
+      "loss": 0.649,
+      "step": 8941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8761112829093092,
+      "learning_rate": 4.077147409969181e-06,
+      "loss": 0.6965,
+      "step": 8942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9244999535200518,
+      "learning_rate": 4.076128750224436e-06,
+      "loss": 0.8002,
+      "step": 8943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9267455806877645,
+      "learning_rate": 4.0751101301804945e-06,
+      "loss": 0.8507,
+      "step": 8944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4101016547433387,
+      "learning_rate": 4.074091549881134e-06,
+      "loss": 0.7845,
+      "step": 8945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9016556808349943,
+      "learning_rate": 4.0730730093701185e-06,
+      "loss": 0.791,
+      "step": 8946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3765021424888575,
+      "learning_rate": 4.072054508691222e-06,
+      "loss": 0.8286,
+      "step": 8947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.91532411842323,
+      "learning_rate": 4.07103604788821e-06,
+      "loss": 0.8801,
+      "step": 8948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.963870912379682,
+      "learning_rate": 4.070017627004847e-06,
+      "loss": 0.7864,
+      "step": 8949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.037640855358544,
+      "learning_rate": 4.068999246084898e-06,
+      "loss": 0.9329,
+      "step": 8950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.101208496954104,
+      "learning_rate": 4.067980905172125e-06,
+      "loss": 0.7047,
+      "step": 8951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4180239906473213,
+      "learning_rate": 4.066962604310288e-06,
+      "loss": 0.9818,
+      "step": 8952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.455453734202035,
+      "learning_rate": 4.065944343543146e-06,
+      "loss": 0.7563,
+      "step": 8953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6209685368537174,
+      "learning_rate": 4.0649261229144554e-06,
+      "loss": 0.7383,
+      "step": 8954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.728442031775685,
+      "learning_rate": 4.063907942467971e-06,
+      "loss": 0.9409,
+      "step": 8955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0274180641528075,
+      "learning_rate": 4.06288980224745e-06,
+      "loss": 0.8104,
+      "step": 8956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.30413205089729,
+      "learning_rate": 4.0618717022966395e-06,
+      "loss": 0.966,
+      "step": 8957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3672345506027743,
+      "learning_rate": 4.060853642659291e-06,
+      "loss": 0.7805,
+      "step": 8958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.914124395531821,
+      "learning_rate": 4.059835623379155e-06,
+      "loss": 0.8809,
+      "step": 8959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8498221569060116,
+      "learning_rate": 4.058817644499973e-06,
+      "loss": 0.6683,
+      "step": 8960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4725467689573724,
+      "learning_rate": 4.057799706065495e-06,
+      "loss": 0.5603,
+      "step": 8961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7144990512998568,
+      "learning_rate": 4.056781808119461e-06,
+      "loss": 0.7806,
+      "step": 8962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.208070988791041,
+      "learning_rate": 4.055763950705615e-06,
+      "loss": 0.7882,
+      "step": 8963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.645772862532264,
+      "learning_rate": 4.054746133867693e-06,
+      "loss": 0.9263,
+      "step": 8964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1158997441758727,
+      "learning_rate": 4.053728357649436e-06,
+      "loss": 0.8793,
+      "step": 8965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6458028711700003,
+      "learning_rate": 4.052710622094579e-06,
+      "loss": 0.6071,
+      "step": 8966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0750490259946544,
+      "learning_rate": 4.051692927246857e-06,
+      "loss": 0.8528,
+      "step": 8967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.524424652552847,
+      "learning_rate": 4.050675273149999e-06,
+      "loss": 0.8689,
+      "step": 8968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6306003923460524,
+      "learning_rate": 4.0496576598477396e-06,
+      "loss": 0.7387,
+      "step": 8969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140092612859472,
+      "learning_rate": 4.048640087383809e-06,
+      "loss": 0.9142,
+      "step": 8970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.001688928412969,
+      "learning_rate": 4.0476225558019324e-06,
+      "loss": 0.8972,
+      "step": 8971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1723724236720146,
+      "learning_rate": 4.046605065145834e-06,
+      "loss": 0.8682,
+      "step": 8972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.135674713697281,
+      "learning_rate": 4.04558761545924e-06,
+      "loss": 0.9536,
+      "step": 8973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6007501685265217,
+      "learning_rate": 4.044570206785874e-06,
+      "loss": 0.878,
+      "step": 8974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7831626503688438,
+      "learning_rate": 4.043552839169451e-06,
+      "loss": 0.7576,
+      "step": 8975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.086802857475686,
+      "learning_rate": 4.042535512653693e-06,
+      "loss": 1.0387,
+      "step": 8976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5202827283410416,
+      "learning_rate": 4.041518227282319e-06,
+      "loss": 0.7523,
+      "step": 8977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.810945534008567,
+      "learning_rate": 4.040500983099038e-06,
+      "loss": 0.8468,
+      "step": 8978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.500928822900256,
+      "learning_rate": 4.039483780147568e-06,
+      "loss": 0.8442,
+      "step": 8979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0395373880264216,
+      "learning_rate": 4.038466618471619e-06,
+      "loss": 0.6366,
+      "step": 8980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.088146633227487,
+      "learning_rate": 4.037449498114903e-06,
+      "loss": 0.7172,
+      "step": 8981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.668241976338205,
+      "learning_rate": 4.036432419121123e-06,
+      "loss": 0.7229,
+      "step": 8982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0670002719219398,
+      "learning_rate": 4.035415381533988e-06,
+      "loss": 0.7563,
+      "step": 8983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.135841695365821,
+      "learning_rate": 4.0343983853972045e-06,
+      "loss": 0.9137,
+      "step": 8984
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9278484743728048,
+      "learning_rate": 4.0333814307544726e-06,
+      "loss": 0.9158,
+      "step": 8985
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6603520032299923,
+      "learning_rate": 4.032364517649491e-06,
+      "loss": 0.7421,
+      "step": 8986
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0151961203054214,
+      "learning_rate": 4.0313476461259615e-06,
+      "loss": 0.6628,
+      "step": 8987
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.036574009153376,
+      "learning_rate": 4.0303308162275835e-06,
+      "loss": 0.8592,
+      "step": 8988
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3579588086473158,
+      "learning_rate": 4.029314027998049e-06,
+      "loss": 0.8103,
+      "step": 8989
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.207029870643519,
+      "learning_rate": 4.028297281481051e-06,
+      "loss": 0.7811,
+      "step": 8990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1979482849956344,
+      "learning_rate": 4.027280576720282e-06,
+      "loss": 0.8223,
+      "step": 8991
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.864440952982911,
+      "learning_rate": 4.026263913759437e-06,
+      "loss": 0.898,
+      "step": 8992
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96756212081733,
+      "learning_rate": 4.025247292642197e-06,
+      "loss": 0.8503,
+      "step": 8993
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9497436480864591,
+      "learning_rate": 4.024230713412253e-06,
+      "loss": 0.6396,
+      "step": 8994
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4158215611594036,
+      "learning_rate": 4.0232141761132894e-06,
+      "loss": 0.8601,
+      "step": 8995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0112889896420327,
+      "learning_rate": 4.022197680788986e-06,
+      "loss": 0.8126,
+      "step": 8996
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.881852859706649,
+      "learning_rate": 4.021181227483026e-06,
+      "loss": 0.6919,
+      "step": 8997
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8296561518199077,
+      "learning_rate": 4.02016481623909e-06,
+      "loss": 0.9994,
+      "step": 8998
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3464601397290115,
+      "learning_rate": 4.019148447100855e-06,
+      "loss": 0.7113,
+      "step": 8999
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1274938703684945,
+      "learning_rate": 4.018132120111992e-06,
+      "loss": 0.7011,
+      "step": 9000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2058846936238328,
+      "learning_rate": 4.0171158353161805e-06,
+      "loss": 1.099,
+      "step": 9001
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8453823620190974,
+      "learning_rate": 4.016099592757091e-06,
+      "loss": 0.9046,
+      "step": 9002
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.607678806750495,
+      "learning_rate": 4.015083392478393e-06,
+      "loss": 0.729,
+      "step": 9003
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.305619067212868,
+      "learning_rate": 4.014067234523756e-06,
+      "loss": 0.5387,
+      "step": 9004
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.616780061653314,
+      "learning_rate": 4.0130511189368435e-06,
+      "loss": 0.9015,
+      "step": 9005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5625416731121815,
+      "learning_rate": 4.012035045761326e-06,
+      "loss": 0.8305,
+      "step": 9006
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7116339161097167,
+      "learning_rate": 4.0110190150408594e-06,
+      "loss": 0.7399,
+      "step": 9007
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0533467340320188,
+      "learning_rate": 4.01000302681911e-06,
+      "loss": 0.6204,
+      "step": 9008
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9209788479714571,
+      "learning_rate": 4.008987081139734e-06,
+      "loss": 0.8387,
+      "step": 9009
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7296541603234363,
+      "learning_rate": 4.007971178046393e-06,
+      "loss": 0.8177,
+      "step": 9010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.559502139113864,
+      "learning_rate": 4.006955317582737e-06,
+      "loss": 0.7363,
+      "step": 9011
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5274950282256112,
+      "learning_rate": 4.005939499792424e-06,
+      "loss": 0.8382,
+      "step": 9012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0511278771658534,
+      "learning_rate": 4.004923724719106e-06,
+      "loss": 0.7844,
+      "step": 9013
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.210849635960579,
+      "learning_rate": 4.0039079924064285e-06,
+      "loss": 0.7634,
+      "step": 9014
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7053831162743018,
+      "learning_rate": 4.0028923028980435e-06,
+      "loss": 0.7662,
+      "step": 9015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4436716488074324,
+      "learning_rate": 4.0018766562375984e-06,
+      "loss": 0.7027,
+      "step": 9016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1899036557195086,
+      "learning_rate": 4.000861052468736e-06,
+      "loss": 0.73,
+      "step": 9017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.690105879923989,
+      "learning_rate": 3.9998454916350996e-06,
+      "loss": 0.7228,
+      "step": 9018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.780849102782743,
+      "learning_rate": 3.998829973780329e-06,
+      "loss": 0.8141,
+      "step": 9019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6957604111314504,
+      "learning_rate": 3.997814498948064e-06,
+      "loss": 0.9074,
+      "step": 9020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3208778655081255,
+      "learning_rate": 3.996799067181943e-06,
+      "loss": 0.8174,
+      "step": 9021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.369193541639771,
+      "learning_rate": 3.995783678525601e-06,
+      "loss": 0.8823,
+      "step": 9022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.086605694095968,
+      "learning_rate": 3.994768333022669e-06,
+      "loss": 0.8573,
+      "step": 9023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.39138318845999,
+      "learning_rate": 3.993753030716783e-06,
+      "loss": 0.794,
+      "step": 9024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.048153263593755,
+      "learning_rate": 3.992737771651567e-06,
+      "loss": 0.9515,
+      "step": 9025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.227545666216226,
+      "learning_rate": 3.991722555870655e-06,
+      "loss": 0.9114,
+      "step": 9026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1121738259372083,
+      "learning_rate": 3.990707383417668e-06,
+      "loss": 0.8213,
+      "step": 9027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9222043837354863,
+      "learning_rate": 3.989692254336235e-06,
+      "loss": 0.8857,
+      "step": 9028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.52249216307406,
+      "learning_rate": 3.988677168669974e-06,
+      "loss": 0.6686,
+      "step": 9029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0942961973427427,
+      "learning_rate": 3.987662126462507e-06,
+      "loss": 0.5995,
+      "step": 9030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1192433224725447,
+      "learning_rate": 3.9866471277574554e-06,
+      "loss": 0.9806,
+      "step": 9031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.065360611962693,
+      "learning_rate": 3.985632172598433e-06,
+      "loss": 0.9887,
+      "step": 9032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0765297271002279,
+      "learning_rate": 3.984617261029054e-06,
+      "loss": 0.6622,
+      "step": 9033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1648971806296853,
+      "learning_rate": 3.983602393092931e-06,
+      "loss": 0.8604,
+      "step": 9034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.452780104828427,
+      "learning_rate": 3.982587568833681e-06,
+      "loss": 0.966,
+      "step": 9035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2969285125148804,
+      "learning_rate": 3.981572788294907e-06,
+      "loss": 0.6803,
+      "step": 9036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.338532058091247,
+      "learning_rate": 3.980558051520218e-06,
+      "loss": 0.9456,
+      "step": 9037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7053098440886822,
+      "learning_rate": 3.97954335855322e-06,
+      "loss": 0.8541,
+      "step": 9038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.162262203000623,
+      "learning_rate": 3.978528709437518e-06,
+      "loss": 0.8519,
+      "step": 9039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0381892960939023,
+      "learning_rate": 3.977514104216711e-06,
+      "loss": 0.6538,
+      "step": 9040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5902688748904366,
+      "learning_rate": 3.9764995429344e-06,
+      "loss": 0.8151,
+      "step": 9041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.84708312066987,
+      "learning_rate": 3.975485025634185e-06,
+      "loss": 0.7223,
+      "step": 9042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.008799785262304,
+      "learning_rate": 3.9744705523596575e-06,
+      "loss": 0.5889,
+      "step": 9043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7377702355452964,
+      "learning_rate": 3.973456123154415e-06,
+      "loss": 0.6108,
+      "step": 9044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6307810075689164,
+      "learning_rate": 3.972441738062049e-06,
+      "loss": 0.8062,
+      "step": 9045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.020710914032143,
+      "learning_rate": 3.971427397126153e-06,
+      "loss": 0.801,
+      "step": 9046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8179508561987,
+      "learning_rate": 3.970413100390309e-06,
+      "loss": 0.8518,
+      "step": 9047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.405143651937342,
+      "learning_rate": 3.9693988478981076e-06,
+      "loss": 0.8693,
+      "step": 9048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.186273266418504,
+      "learning_rate": 3.9683846396931345e-06,
+      "loss": 0.6324,
+      "step": 9049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2008006655083703,
+      "learning_rate": 3.967370475818971e-06,
+      "loss": 0.9569,
+      "step": 9050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0604868806000753,
+      "learning_rate": 3.966356356319196e-06,
+      "loss": 0.5927,
+      "step": 9051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8255465827976156,
+      "learning_rate": 3.965342281237391e-06,
+      "loss": 0.8791,
+      "step": 9052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5985071509838447,
+      "learning_rate": 3.964328250617134e-06,
+      "loss": 0.9014,
+      "step": 9053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.458819287268004,
+      "learning_rate": 3.9633142645019965e-06,
+      "loss": 0.639,
+      "step": 9054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3059864134113988,
+      "learning_rate": 3.962300322935556e-06,
+      "loss": 0.7576,
+      "step": 9055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.730809234764289,
+      "learning_rate": 3.961286425961379e-06,
+      "loss": 0.5673,
+      "step": 9056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.84615399792289,
+      "learning_rate": 3.9602725736230405e-06,
+      "loss": 0.8674,
+      "step": 9057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.00587010380008,
+      "learning_rate": 3.959258765964104e-06,
+      "loss": 0.7648,
+      "step": 9058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0412808130950437,
+      "learning_rate": 3.958245003028136e-06,
+      "loss": 0.761,
+      "step": 9059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6967562298014474,
+      "learning_rate": 3.957231284858701e-06,
+      "loss": 0.9944,
+      "step": 9060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9276706318162657,
+      "learning_rate": 3.956217611499359e-06,
+      "loss": 0.6548,
+      "step": 9061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.383474232994146,
+      "learning_rate": 3.9552039829936705e-06,
+      "loss": 0.9435,
+      "step": 9062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9823092867809248,
+      "learning_rate": 3.954190399385195e-06,
+      "loss": 0.6914,
+      "step": 9063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2272689118618088,
+      "learning_rate": 3.953176860717488e-06,
+      "loss": 0.8518,
+      "step": 9064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0510932738406287,
+      "learning_rate": 3.9521633670341005e-06,
+      "loss": 0.6791,
+      "step": 9065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2768804217003455,
+      "learning_rate": 3.9511499183785875e-06,
+      "loss": 0.8352,
+      "step": 9066
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8986203176889282,
+      "learning_rate": 3.950136514794499e-06,
+      "loss": 0.7799,
+      "step": 9067
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6566892403165443,
+      "learning_rate": 3.9491231563253815e-06,
+      "loss": 0.7976,
+      "step": 9068
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7506130483380675,
+      "learning_rate": 3.948109843014784e-06,
+      "loss": 0.7234,
+      "step": 9069
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1710745391666344,
+      "learning_rate": 3.947096574906248e-06,
+      "loss": 0.8309,
+      "step": 9070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4241726976672981,
+      "learning_rate": 3.946083352043318e-06,
+      "loss": 0.7549,
+      "step": 9071
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9370246214817861,
+      "learning_rate": 3.9450701744695325e-06,
+      "loss": 0.6365,
+      "step": 9072
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.176422832247924,
+      "learning_rate": 3.944057042228432e-06,
+      "loss": 0.8491,
+      "step": 9073
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7259656621485475,
+      "learning_rate": 3.9430439553635504e-06,
+      "loss": 0.8502,
+      "step": 9074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.945928373985674,
+      "learning_rate": 3.942030913918427e-06,
+      "loss": 0.7778,
+      "step": 9075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1426863521613178,
+      "learning_rate": 3.9410179179365895e-06,
+      "loss": 0.7077,
+      "step": 9076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9432655796977942,
+      "learning_rate": 3.940004967461571e-06,
+      "loss": 0.6738,
+      "step": 9077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9138058200097026,
+      "learning_rate": 3.9389920625368996e-06,
+      "loss": 0.9735,
+      "step": 9078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3824994229999805,
+      "learning_rate": 3.937979203206103e-06,
+      "loss": 0.8313,
+      "step": 9079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.540411566578323,
+      "learning_rate": 3.936966389512703e-06,
+      "loss": 0.7261,
+      "step": 9080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7816385892630948,
+      "learning_rate": 3.935953621500226e-06,
+      "loss": 0.8276,
+      "step": 9081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2455162555195027,
+      "learning_rate": 3.934940899212193e-06,
+      "loss": 0.6501,
+      "step": 9082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.129397207048865,
+      "learning_rate": 3.9339282226921215e-06,
+      "loss": 0.7854,
+      "step": 9083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0238406130830526,
+      "learning_rate": 3.932915591983526e-06,
+      "loss": 0.7456,
+      "step": 9084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.249368309236517,
+      "learning_rate": 3.931903007129927e-06,
+      "loss": 0.6631,
+      "step": 9085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.190814436319678,
+      "learning_rate": 3.930890468174833e-06,
+      "loss": 0.6979,
+      "step": 9086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.146092794615014,
+      "learning_rate": 3.929877975161757e-06,
+      "loss": 0.8499,
+      "step": 9087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.826975904446776,
+      "learning_rate": 3.928865528134206e-06,
+      "loss": 0.8696,
+      "step": 9088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.853925850537654,
+      "learning_rate": 3.927853127135692e-06,
+      "loss": 0.8065,
+      "step": 9089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3056842041922083,
+      "learning_rate": 3.926840772209714e-06,
+      "loss": 0.9749,
+      "step": 9090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.420982202997337,
+      "learning_rate": 3.92582846339978e-06,
+      "loss": 0.9312,
+      "step": 9091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6921327634805408,
+      "learning_rate": 3.924816200749387e-06,
+      "loss": 0.8298,
+      "step": 9092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9146902898111182,
+      "learning_rate": 3.92380398430204e-06,
+      "loss": 0.8175,
+      "step": 9093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6310802095874775,
+      "learning_rate": 3.92279181410123e-06,
+      "loss": 0.7387,
+      "step": 9094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2700033284279555,
+      "learning_rate": 3.921779690190454e-06,
+      "loss": 0.7127,
+      "step": 9095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.074542256286348,
+      "learning_rate": 3.920767612613209e-06,
+      "loss": 0.7372,
+      "step": 9096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.238556067406386,
+      "learning_rate": 3.919755581412982e-06,
+      "loss": 0.862,
+      "step": 9097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2746214260601088,
+      "learning_rate": 3.918743596633262e-06,
+      "loss": 0.926,
+      "step": 9098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.738403658537067,
+      "learning_rate": 3.917731658317538e-06,
+      "loss": 0.7713,
+      "step": 9099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.909247288936746,
+      "learning_rate": 3.916719766509297e-06,
+      "loss": 0.7755,
+      "step": 9100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.615717231405539,
+      "learning_rate": 3.91570792125202e-06,
+      "loss": 0.7825,
+      "step": 9101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.457351043261078,
+      "learning_rate": 3.914696122589187e-06,
+      "loss": 0.7227,
+      "step": 9102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.245308869924724,
+      "learning_rate": 3.91368437056428e-06,
+      "loss": 0.7385,
+      "step": 9103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6729095665782663,
+      "learning_rate": 3.912672665220773e-06,
+      "loss": 0.7256,
+      "step": 9104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6360730275056805,
+      "learning_rate": 3.911661006602144e-06,
+      "loss": 0.8287,
+      "step": 9105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.05794611285528,
+      "learning_rate": 3.910649394751867e-06,
+      "loss": 0.6727,
+      "step": 9106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.017724533689281,
+      "learning_rate": 3.9096378297134115e-06,
+      "loss": 0.8269,
+      "step": 9107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.459638094706772,
+      "learning_rate": 3.908626311530245e-06,
+      "loss": 0.6355,
+      "step": 9108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.009358468300281,
+      "learning_rate": 3.907614840245836e-06,
+      "loss": 0.7238,
+      "step": 9109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2063361177707312,
+      "learning_rate": 3.906603415903652e-06,
+      "loss": 0.665,
+      "step": 9110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.509627970478491,
+      "learning_rate": 3.905592038547155e-06,
+      "loss": 0.759,
+      "step": 9111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.912729151447167,
+      "learning_rate": 3.904580708219804e-06,
+      "loss": 0.8715,
+      "step": 9112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9544380116777393,
+      "learning_rate": 3.903569424965059e-06,
+      "loss": 0.7223,
+      "step": 9113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0812043705749903,
+      "learning_rate": 3.90255818882638e-06,
+      "loss": 0.6456,
+      "step": 9114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.361481709109973,
+      "learning_rate": 3.901546999847217e-06,
+      "loss": 0.7639,
+      "step": 9115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.009576540726898,
+      "learning_rate": 3.9005358580710275e-06,
+      "loss": 0.848,
+      "step": 9116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.553711255234656,
+      "learning_rate": 3.89952476354126e-06,
+      "loss": 0.811,
+      "step": 9117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2525223873510622,
+      "learning_rate": 3.898513716301366e-06,
+      "loss": 0.7275,
+      "step": 9118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0208839697522514,
+      "learning_rate": 3.897502716394789e-06,
+      "loss": 0.7825,
+      "step": 9119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.861736179667669,
+      "learning_rate": 3.896491763864976e-06,
+      "loss": 0.5607,
+      "step": 9120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1970430495581126,
+      "learning_rate": 3.89548085875537e-06,
+      "loss": 0.9365,
+      "step": 9121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.72411578970059,
+      "learning_rate": 3.894470001109412e-06,
+      "loss": 0.7582,
+      "step": 9122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8837409421348037,
+      "learning_rate": 3.8934591909705385e-06,
+      "loss": 0.7961,
+      "step": 9123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.461262331564626,
+      "learning_rate": 3.892448428382189e-06,
+      "loss": 0.6845,
+      "step": 9124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.083963267603939,
+      "learning_rate": 3.8914377133877976e-06,
+      "loss": 0.784,
+      "step": 9125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.39606287509554,
+      "learning_rate": 3.890427046030796e-06,
+      "loss": 0.7581,
+      "step": 9126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3060780957840126,
+      "learning_rate": 3.889416426354614e-06,
+      "loss": 0.7135,
+      "step": 9127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0834946142857294,
+      "learning_rate": 3.888405854402684e-06,
+      "loss": 0.7115,
+      "step": 9128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8450121898946026,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.7069,
+      "step": 9129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1709148075664992,
+      "learning_rate": 3.886384853845274e-06,
+      "loss": 0.7366,
+      "step": 9130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.199485107561883,
+      "learning_rate": 3.885374425326642e-06,
+      "loss": 0.7491,
+      "step": 9131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4157906118056185,
+      "learning_rate": 3.884364044705955e-06,
+      "loss": 0.794,
+      "step": 9132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4930883736026836,
+      "learning_rate": 3.883353712026629e-06,
+      "loss": 0.8081,
+      "step": 9133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5273500383661864,
+      "learning_rate": 3.8823434273320794e-06,
+      "loss": 0.7841,
+      "step": 9134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0007567041565535,
+      "learning_rate": 3.881333190665723e-06,
+      "loss": 0.791,
+      "step": 9135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2255010895479146,
+      "learning_rate": 3.880323002070971e-06,
+      "loss": 0.6931,
+      "step": 9136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1848353820558293,
+      "learning_rate": 3.879312861591233e-06,
+      "loss": 0.8413,
+      "step": 9137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.834057471703215,
+      "learning_rate": 3.878302769269917e-06,
+      "loss": 0.7177,
+      "step": 9138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.78333402950909,
+      "learning_rate": 3.877292725150429e-06,
+      "loss": 1.0048,
+      "step": 9139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.186866944339806,
+      "learning_rate": 3.876282729276174e-06,
+      "loss": 0.8783,
+      "step": 9140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.101079726972309,
+      "learning_rate": 3.875272781690552e-06,
+      "loss": 0.902,
+      "step": 9141
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1611723674955408,
+      "learning_rate": 3.8742628824369624e-06,
+      "loss": 0.7144,
+      "step": 9142
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7361151920230329,
+      "learning_rate": 3.8732530315588065e-06,
+      "loss": 0.6626,
+      "step": 9143
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7108671170357956,
+      "learning_rate": 3.872243229099476e-06,
+      "loss": 0.8744,
+      "step": 9144
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.773532710129964,
+      "learning_rate": 3.871233475102365e-06,
+      "loss": 1.048,
+      "step": 9145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1291623432112714,
+      "learning_rate": 3.870223769610865e-06,
+      "loss": 0.8867,
+      "step": 9146
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.97983166313867,
+      "learning_rate": 3.869214112668368e-06,
+      "loss": 0.8339,
+      "step": 9147
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3165339677897347,
+      "learning_rate": 3.868204504318258e-06,
+      "loss": 0.9393,
+      "step": 9148
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8792961785803186,
+      "learning_rate": 3.86719494460392e-06,
+      "loss": 0.5977,
+      "step": 9149
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9177638602473959,
+      "learning_rate": 3.866185433568741e-06,
+      "loss": 0.7288,
+      "step": 9150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2096012459768515,
+      "learning_rate": 3.865175971256095e-06,
+      "loss": 0.9699,
+      "step": 9151
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9880388968470747,
+      "learning_rate": 3.864166557709368e-06,
+      "loss": 0.5654,
+      "step": 9152
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2472855396065388,
+      "learning_rate": 3.863157192971932e-06,
+      "loss": 0.8097,
+      "step": 9153
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0355699082543244,
+      "learning_rate": 3.8621478770871645e-06,
+      "loss": 0.6192,
+      "step": 9154
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0509166469433744,
+      "learning_rate": 3.861138610098435e-06,
+      "loss": 0.6741,
+      "step": 9155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0460483598044177,
+      "learning_rate": 3.8601293920491165e-06,
+      "loss": 0.8858,
+      "step": 9156
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2212976173220746,
+      "learning_rate": 3.859120222982578e-06,
+      "loss": 0.9006,
+      "step": 9157
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7666615728313473,
+      "learning_rate": 3.858111102942184e-06,
+      "loss": 0.703,
+      "step": 9158
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.094450650548191,
+      "learning_rate": 3.857102031971298e-06,
+      "loss": 1.0832,
+      "step": 9159
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.830340400697669,
+      "learning_rate": 3.856093010113282e-06,
+      "loss": 0.6736,
+      "step": 9160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2816092517527027,
+      "learning_rate": 3.855084037411501e-06,
+      "loss": 0.7541,
+      "step": 9161
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1621088803156288,
+      "learning_rate": 3.854075113909307e-06,
+      "loss": 0.7814,
+      "step": 9162
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8077429201697257,
+      "learning_rate": 3.853066239650055e-06,
+      "loss": 0.7272,
+      "step": 9163
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.30951262849271,
+      "learning_rate": 3.852057414677102e-06,
+      "loss": 0.7316,
+      "step": 9164
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0881583805168935,
+      "learning_rate": 3.851048639033801e-06,
+      "loss": 0.6457,
+      "step": 9165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.934538498109115,
+      "learning_rate": 3.850039912763496e-06,
+      "loss": 0.6693,
+      "step": 9166
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.543655454687977,
+      "learning_rate": 3.849031235909539e-06,
+      "loss": 0.8917,
+      "step": 9167
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3575689001357922,
+      "learning_rate": 3.848022608515273e-06,
+      "loss": 0.8961,
+      "step": 9168
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.165604158487334,
+      "learning_rate": 3.84701403062404e-06,
+      "loss": 0.8495,
+      "step": 9169
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0689004535850004,
+      "learning_rate": 3.846005502279182e-06,
+      "loss": 0.7432,
+      "step": 9170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.038019689797681,
+      "learning_rate": 3.844997023524038e-06,
+      "loss": 0.6446,
+      "step": 9171
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.604742423392683,
+      "learning_rate": 3.843988594401946e-06,
+      "loss": 1.0044,
+      "step": 9172
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2901657318913164,
+      "learning_rate": 3.842980214956236e-06,
+      "loss": 0.7165,
+      "step": 9173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1245224318816653,
+      "learning_rate": 3.841971885230243e-06,
+      "loss": 0.6872,
+      "step": 9174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.344279155541609,
+      "learning_rate": 3.840963605267298e-06,
+      "loss": 0.7545,
+      "step": 9175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0607274692642688,
+      "learning_rate": 3.83995537511073e-06,
+      "loss": 0.6667,
+      "step": 9176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1810225105595853,
+      "learning_rate": 3.83894719480386e-06,
+      "loss": 0.8059,
+      "step": 9177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9596370895727944,
+      "learning_rate": 3.837939064390016e-06,
+      "loss": 0.6754,
+      "step": 9178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9895068480930285,
+      "learning_rate": 3.83693098391252e-06,
+      "loss": 0.7222,
+      "step": 9179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2031816251593934,
+      "learning_rate": 3.835922953414688e-06,
+      "loss": 0.9367,
+      "step": 9180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.163483599544918,
+      "learning_rate": 3.83491497293984e-06,
+      "loss": 0.7885,
+      "step": 9181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.637134311043968,
+      "learning_rate": 3.833907042531289e-06,
+      "loss": 0.8289,
+      "step": 9182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9879337361510263,
+      "learning_rate": 3.832899162232352e-06,
+      "loss": 0.9402,
+      "step": 9183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.033431128584619,
+      "learning_rate": 3.8318913320863355e-06,
+      "loss": 0.9538,
+      "step": 9184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.219255056043763,
+      "learning_rate": 3.830883552136551e-06,
+      "loss": 0.7111,
+      "step": 9185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.27191694996171,
+      "learning_rate": 3.829875822426304e-06,
+      "loss": 0.8669,
+      "step": 9186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3634710082206243,
+      "learning_rate": 3.828868142998898e-06,
+      "loss": 0.8729,
+      "step": 9187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3299674838009725,
+      "learning_rate": 3.827860513897635e-06,
+      "loss": 0.7836,
+      "step": 9188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.768676392942795,
+      "learning_rate": 3.826852935165818e-06,
+      "loss": 0.717,
+      "step": 9189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9908862262364906,
+      "learning_rate": 3.825845406846743e-06,
+      "loss": 0.8974,
+      "step": 9190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1372578469492347,
+      "learning_rate": 3.8248379289837065e-06,
+      "loss": 0.6613,
+      "step": 9191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9460958287711927,
+      "learning_rate": 3.8238305016199995e-06,
+      "loss": 0.636,
+      "step": 9192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.758242150174255,
+      "learning_rate": 3.822823124798915e-06,
+      "loss": 0.63,
+      "step": 9193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5531043018625907,
+      "learning_rate": 3.8218157985637465e-06,
+      "loss": 0.5804,
+      "step": 9194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5187275283082564,
+      "learning_rate": 3.820808522957774e-06,
+      "loss": 0.846,
+      "step": 9195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.165664813221379,
+      "learning_rate": 3.819801298024286e-06,
+      "loss": 0.7436,
+      "step": 9196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.336308796988787,
+      "learning_rate": 3.818794123806565e-06,
+      "loss": 0.9644,
+      "step": 9197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2640964793825282,
+      "learning_rate": 3.81778700034789e-06,
+      "loss": 0.9847,
+      "step": 9198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1165508515073441,
+      "learning_rate": 3.816779927691542e-06,
+      "loss": 0.7349,
+      "step": 9199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5684431523732325,
+      "learning_rate": 3.815772905880794e-06,
+      "loss": 0.792,
+      "step": 9200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1643839283524877,
+      "learning_rate": 3.814765934958924e-06,
+      "loss": 0.6738,
+      "step": 9201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8345437886594962,
+      "learning_rate": 3.8137590149691992e-06,
+      "loss": 0.8731,
+      "step": 9202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.395841677136103,
+      "learning_rate": 3.8127521459548934e-06,
+      "loss": 0.8589,
+      "step": 9203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.27010217348373,
+      "learning_rate": 3.811745327959271e-06,
+      "loss": 0.6764,
+      "step": 9204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8541879673632573,
+      "learning_rate": 3.810738561025599e-06,
+      "loss": 0.8292,
+      "step": 9205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1989605543358186,
+      "learning_rate": 3.8097318451971375e-06,
+      "loss": 0.8627,
+      "step": 9206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7296722083697738,
+      "learning_rate": 3.80872518051715e-06,
+      "loss": 0.6543,
+      "step": 9207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1907806698269476,
+      "learning_rate": 3.807718567028897e-06,
+      "loss": 0.8002,
+      "step": 9208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.207144386265475,
+      "learning_rate": 3.8067120047756313e-06,
+      "loss": 0.6499,
+      "step": 9209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.092742562795255,
+      "learning_rate": 3.8057054938006072e-06,
+      "loss": 0.7906,
+      "step": 9210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.509031295213329,
+      "learning_rate": 3.804699034147078e-06,
+      "loss": 0.7948,
+      "step": 9211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2373249806633395,
+      "learning_rate": 3.803692625858295e-06,
+      "loss": 0.8873,
+      "step": 9212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.722336018497033,
+      "learning_rate": 3.8026862689775033e-06,
+      "loss": 0.6331,
+      "step": 9213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.12550495354284,
+      "learning_rate": 3.801679963547949e-06,
+      "loss": 0.6128,
+      "step": 9214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.41856726950689,
+      "learning_rate": 3.8006737096128763e-06,
+      "loss": 0.9817,
+      "step": 9215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1080969795666693,
+      "learning_rate": 3.7996675072155243e-06,
+      "loss": 0.8561,
+      "step": 9216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1528269482992224,
+      "learning_rate": 3.7986613563991327e-06,
+      "loss": 0.8055,
+      "step": 9217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2019605417778267,
+      "learning_rate": 3.7976552572069385e-06,
+      "loss": 0.9847,
+      "step": 9218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2408870888083954,
+      "learning_rate": 3.7966492096821773e-06,
+      "loss": 0.6542,
+      "step": 9219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1364261769040724,
+      "learning_rate": 3.795643213868077e-06,
+      "loss": 0.8471,
+      "step": 9220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8643083499448758,
+      "learning_rate": 3.7946372698078704e-06,
+      "loss": 0.9193,
+      "step": 9221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5080193798028545,
+      "learning_rate": 3.793631377544787e-06,
+      "loss": 0.8495,
+      "step": 9222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9876326084238792,
+      "learning_rate": 3.7926255371220487e-06,
+      "loss": 0.4809,
+      "step": 9223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5435025136318552,
+      "learning_rate": 3.7916197485828793e-06,
+      "loss": 0.6853,
+      "step": 9224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1631102697414777,
+      "learning_rate": 3.7906140119705006e-06,
+      "loss": 0.665,
+      "step": 9225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2140321756659254,
+      "learning_rate": 3.7896083273281324e-06,
+      "loss": 0.7752,
+      "step": 9226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0183345683974334,
+      "learning_rate": 3.7886026946989896e-06,
+      "loss": 0.8573,
+      "step": 9227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9967795000769,
+      "learning_rate": 3.787597114126286e-06,
+      "loss": 0.6974,
+      "step": 9228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.137651806880114,
+      "learning_rate": 3.786591585653235e-06,
+      "loss": 0.6836,
+      "step": 9229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.0371701576142405,
+      "learning_rate": 3.785586109323047e-06,
+      "loss": 0.8251,
+      "step": 9230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.164269308263464,
+      "learning_rate": 3.7845806851789275e-06,
+      "loss": 0.6428,
+      "step": 9231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.778135227336975,
+      "learning_rate": 3.783575313264083e-06,
+      "loss": 0.721,
+      "step": 9232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.830905690894576,
+      "learning_rate": 3.7825699936217183e-06,
+      "loss": 0.8062,
+      "step": 9233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9059238423507738,
+      "learning_rate": 3.7815647262950293e-06,
+      "loss": 0.8803,
+      "step": 9234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1369651521618276,
+      "learning_rate": 3.7805595113272186e-06,
+      "loss": 0.9559,
+      "step": 9235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.165611691263526,
+      "learning_rate": 3.7795543487614815e-06,
+      "loss": 0.7317,
+      "step": 9236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0959324848259784,
+      "learning_rate": 3.7785492386410138e-06,
+      "loss": 0.711,
+      "step": 9237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.523833023401021,
+      "learning_rate": 3.7775441810090043e-06,
+      "loss": 0.7679,
+      "step": 9238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.860485952635181,
+      "learning_rate": 3.7765391759086424e-06,
+      "loss": 0.9082,
+      "step": 9239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.251689100215959,
+      "learning_rate": 3.7755342233831188e-06,
+      "loss": 0.7583,
+      "step": 9240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1806461099486505,
+      "learning_rate": 3.7745293234756164e-06,
+      "loss": 0.9302,
+      "step": 9241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0513564532808592,
+      "learning_rate": 3.7735244762293183e-06,
+      "loss": 0.833,
+      "step": 9242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.357448861025969,
+      "learning_rate": 3.7725196816874045e-06,
+      "loss": 0.8776,
+      "step": 9243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8128696840043046,
+      "learning_rate": 3.771514939893055e-06,
+      "loss": 0.8259,
+      "step": 9244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.174106363100766,
+      "learning_rate": 3.7705102508894425e-06,
+      "loss": 0.7869,
+      "step": 9245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0263272485915373,
+      "learning_rate": 3.769505614719744e-06,
+      "loss": 0.9065,
+      "step": 9246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0140656599987345,
+      "learning_rate": 3.7685010314271287e-06,
+      "loss": 0.6138,
+      "step": 9247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.522814551482176,
+      "learning_rate": 3.7674965010547686e-06,
+      "loss": 0.7928,
+      "step": 9248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9011894229360962,
+      "learning_rate": 3.766492023645827e-06,
+      "loss": 0.6711,
+      "step": 9249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2855960004018017,
+      "learning_rate": 3.7654875992434714e-06,
+      "loss": 0.7402,
+      "step": 9250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.247121238412199,
+      "learning_rate": 3.764483227890864e-06,
+      "loss": 0.8625,
+      "step": 9251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8883574663643063,
+      "learning_rate": 3.763478909631162e-06,
+      "loss": 0.6902,
+      "step": 9252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7334884762461942,
+      "learning_rate": 3.7624746445075243e-06,
+      "loss": 0.8976,
+      "step": 9253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6275420838849217,
+      "learning_rate": 3.761470432563109e-06,
+      "loss": 0.9995,
+      "step": 9254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.163643020685944,
+      "learning_rate": 3.760466273841068e-06,
+      "loss": 0.8702,
+      "step": 9255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8009776672806415,
+      "learning_rate": 3.759462168384551e-06,
+      "loss": 0.7568,
+      "step": 9256
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2448790203441464,
+      "learning_rate": 3.758458116236707e-06,
+      "loss": 0.9581,
+      "step": 9257
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9771975303804532,
+      "learning_rate": 3.757454117440682e-06,
+      "loss": 0.8298,
+      "step": 9258
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8658029530722393,
+      "learning_rate": 3.7564501720396242e-06,
+      "loss": 0.8644,
+      "step": 9259
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9630896063809656,
+      "learning_rate": 3.75544628007667e-06,
+      "loss": 0.5613,
+      "step": 9260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09057644558698,
+      "learning_rate": 3.75444244159496e-06,
+      "loss": 0.8452,
+      "step": 9261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.024749528523783,
+      "learning_rate": 3.7534386566376345e-06,
+      "loss": 0.6699,
+      "step": 9262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2447048456005545,
+      "learning_rate": 3.7524349252478243e-06,
+      "loss": 0.8611,
+      "step": 9263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8361819835542312,
+      "learning_rate": 3.7514312474686643e-06,
+      "loss": 0.7319,
+      "step": 9264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2863260867207387,
+      "learning_rate": 3.7504276233432835e-06,
+      "loss": 0.6546,
+      "step": 9265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7922142174927655,
+      "learning_rate": 3.749424052914813e-06,
+      "loss": 0.8142,
+      "step": 9266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1775283198752695,
+      "learning_rate": 3.7484205362263737e-06,
+      "loss": 0.8716,
+      "step": 9267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0420696731935157,
+      "learning_rate": 3.747417073321092e-06,
+      "loss": 0.8365,
+      "step": 9268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9227087449578157,
+      "learning_rate": 3.74641366424209e-06,
+      "loss": 1.1002,
+      "step": 9269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.857881018177481,
+      "learning_rate": 3.7454103090324835e-06,
+      "loss": 0.8819,
+      "step": 9270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.927856934314404,
+      "learning_rate": 3.7444070077353896e-06,
+      "loss": 0.9757,
+      "step": 9271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.470368887871838,
+      "learning_rate": 3.743403760393922e-06,
+      "loss": 1.0237,
+      "step": 9272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3698478513728487,
+      "learning_rate": 3.7424005670511964e-06,
+      "loss": 0.8652,
+      "step": 9273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0576883301991695,
+      "learning_rate": 3.7413974277503183e-06,
+      "loss": 0.8217,
+      "step": 9274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7869188110002607,
+      "learning_rate": 3.740394342534394e-06,
+      "loss": 0.8519,
+      "step": 9275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.230206939183926,
+      "learning_rate": 3.739391311446531e-06,
+      "loss": 0.6968,
+      "step": 9276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2278578925010308,
+      "learning_rate": 3.738388334529832e-06,
+      "loss": 0.6189,
+      "step": 9277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.467961751895636,
+      "learning_rate": 3.7373854118273955e-06,
+      "loss": 0.6987,
+      "step": 9278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8359790026027627,
+      "learning_rate": 3.7363825433823187e-06,
+      "loss": 0.9353,
+      "step": 9279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1208291912566364,
+      "learning_rate": 3.7353797292377002e-06,
+      "loss": 0.6512,
+      "step": 9280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9288686995564588,
+      "learning_rate": 3.7343769694366294e-06,
+      "loss": 0.6731,
+      "step": 9281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1883452134668784,
+      "learning_rate": 3.7333742640221994e-06,
+      "loss": 0.7663,
+      "step": 9282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.855577089387591,
+      "learning_rate": 3.732371613037498e-06,
+      "loss": 0.5979,
+      "step": 9283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0032245688763117,
+      "learning_rate": 3.7313690165256134e-06,
+      "loss": 0.9446,
+      "step": 9284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1369311589441686,
+      "learning_rate": 3.730366474529625e-06,
+      "loss": 0.9319,
+      "step": 9285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1612668264718886,
+      "learning_rate": 3.7293639870926173e-06,
+      "loss": 0.6842,
+      "step": 9286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9243660213380174,
+      "learning_rate": 3.728361554257671e-06,
+      "loss": 0.6649,
+      "step": 9287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.87222602324093,
+      "learning_rate": 3.7273591760678594e-06,
+      "loss": 0.9959,
+      "step": 9288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.013801036021056,
+      "learning_rate": 3.7263568525662574e-06,
+      "loss": 0.7978,
+      "step": 9289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.230667332980842,
+      "learning_rate": 3.7253545837959383e-06,
+      "loss": 0.8116,
+      "step": 9290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8074795193445141,
+      "learning_rate": 3.7243523697999733e-06,
+      "loss": 0.9163,
+      "step": 9291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.267255165284485,
+      "learning_rate": 3.723350210621426e-06,
+      "loss": 0.936,
+      "step": 9292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0649226255119193,
+      "learning_rate": 3.722348106303364e-06,
+      "loss": 0.8753,
+      "step": 9293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.421929243082345,
+      "learning_rate": 3.7213460568888493e-06,
+      "loss": 0.7043,
+      "step": 9294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.06533139124424,
+      "learning_rate": 3.7203440624209438e-06,
+      "loss": 0.7227,
+      "step": 9295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.449289514451868,
+      "learning_rate": 3.7193421229427017e-06,
+      "loss": 0.9633,
+      "step": 9296
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0898141973942361,
+      "learning_rate": 3.7183402384971818e-06,
+      "loss": 0.6816,
+      "step": 9297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4834590295143952,
+      "learning_rate": 3.7173384091274372e-06,
+      "loss": 0.8626,
+      "step": 9298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8044397267315873,
+      "learning_rate": 3.716336634876516e-06,
+      "loss": 0.8781,
+      "step": 9299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.225324155427661,
+      "learning_rate": 3.7153349157874687e-06,
+      "loss": 0.9569,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.543089325404871,
+      "learning_rate": 3.7143332519033413e-06,
+      "loss": 0.6264,
+      "step": 9301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3785599732266713,
+      "learning_rate": 3.7133316432671794e-06,
+      "loss": 0.7976,
+      "step": 9302
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1766792409758846,
+      "learning_rate": 3.7123300899220193e-06,
+      "loss": 0.932,
+      "step": 9303
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1468565914237296,
+      "learning_rate": 3.711328591910904e-06,
+      "loss": 0.7344,
+      "step": 9304
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4754987079599986,
+      "learning_rate": 3.710327149276871e-06,
+      "loss": 0.8364,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.361109507136618,
+      "learning_rate": 3.70932576206295e-06,
+      "loss": 0.7107,
+      "step": 9306
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.695635917290531,
+      "learning_rate": 3.708324430312177e-06,
+      "loss": 0.8371,
+      "step": 9307
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3215044299088685,
+      "learning_rate": 3.7073231540675792e-06,
+      "loss": 0.7214,
+      "step": 9308
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1544099040018567,
+      "learning_rate": 3.7063219333721857e-06,
+      "loss": 0.9792,
+      "step": 9309
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1023946608310644,
+      "learning_rate": 3.7053207682690184e-06,
+      "loss": 0.7098,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0832184394853717,
+      "learning_rate": 3.704319658801102e-06,
+      "loss": 0.7526,
+      "step": 9311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9515047402295849,
+      "learning_rate": 3.7033186050114547e-06,
+      "loss": 0.9247,
+      "step": 9312
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3457064662344584,
+      "learning_rate": 3.7023176069430965e-06,
+      "loss": 0.8405,
+      "step": 9313
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5272242828639495,
+      "learning_rate": 3.7013166646390384e-06,
+      "loss": 0.9258,
+      "step": 9314
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1137918541815344,
+      "learning_rate": 3.700315778142297e-06,
+      "loss": 0.6306,
+      "step": 9315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390771151456257,
+      "learning_rate": 3.699314947495881e-06,
+      "loss": 0.8868,
+      "step": 9316
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.660776738150117,
+      "learning_rate": 3.698314172742799e-06,
+      "loss": 0.687,
+      "step": 9317
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.336855062094555,
+      "learning_rate": 3.6973134539260545e-06,
+      "loss": 0.7866,
+      "step": 9318
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9228952049219858,
+      "learning_rate": 3.6963127910886526e-06,
+      "loss": 0.829,
+      "step": 9319
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.144783089805191,
+      "learning_rate": 3.695312184273595e-06,
+      "loss": 0.6768,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.715691798925626,
+      "learning_rate": 3.694311633523878e-06,
+      "loss": 0.7408,
+      "step": 9321
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9012792495326691,
+      "learning_rate": 3.693311138882497e-06,
+      "loss": 0.9095,
+      "step": 9322
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8334350342528152,
+      "learning_rate": 3.6923107003924486e-06,
+      "loss": 0.6388,
+      "step": 9323
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2311200112097573,
+      "learning_rate": 3.691310318096719e-06,
+      "loss": 0.8232,
+      "step": 9324
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2964566984461277,
+      "learning_rate": 3.690309992038301e-06,
+      "loss": 0.8057,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.249104897548068,
+      "learning_rate": 3.689309722260179e-06,
+      "loss": 0.9905,
+      "step": 9326
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.735164824621521,
+      "learning_rate": 3.6883095088053387e-06,
+      "loss": 0.782,
+      "step": 9327
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.74415699002698,
+      "learning_rate": 3.6873093517167584e-06,
+      "loss": 0.7397,
+      "step": 9328
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0508631576341485,
+      "learning_rate": 3.6863092510374198e-06,
+      "loss": 0.8046,
+      "step": 9329
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9025172932906556,
+      "learning_rate": 3.685309206810297e-06,
+      "loss": 0.9514,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1544113437359083,
+      "learning_rate": 3.684309219078368e-06,
+      "loss": 0.9132,
+      "step": 9331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6531959980306628,
+      "learning_rate": 3.6833092878846e-06,
+      "loss": 0.8514,
+      "step": 9332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2418318690765395,
+      "learning_rate": 3.6823094132719645e-06,
+      "loss": 0.5779,
+      "step": 9333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.387563771316117,
+      "learning_rate": 3.68130959528343e-06,
+      "loss": 0.9162,
+      "step": 9334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.226637120464864,
+      "learning_rate": 3.680309833961958e-06,
+      "loss": 0.8211,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2436598341299703,
+      "learning_rate": 3.6793101293505106e-06,
+      "loss": 0.631,
+      "step": 9336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.253262740189453,
+      "learning_rate": 3.678310481492048e-06,
+      "loss": 0.8623,
+      "step": 9337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9511317998772608,
+      "learning_rate": 3.6773108904295294e-06,
+      "loss": 0.5946,
+      "step": 9338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8310960807681873,
+      "learning_rate": 3.6763113562059077e-06,
+      "loss": 0.8194,
+      "step": 9339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.818887946290182,
+      "learning_rate": 3.6753118788641327e-06,
+      "loss": 0.8192,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.236872350535009,
+      "learning_rate": 3.67431245844716e-06,
+      "loss": 0.8971,
+      "step": 9341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0584140564747881,
+      "learning_rate": 3.67331309499793e-06,
+      "loss": 0.6291,
+      "step": 9342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9456929796884546,
+      "learning_rate": 3.672313788559392e-06,
+      "loss": 0.8424,
+      "step": 9343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.087423512552722,
+      "learning_rate": 3.6713145391744877e-06,
+      "loss": 0.8278,
+      "step": 9344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9044591352962863,
+      "learning_rate": 3.6703153468861585e-06,
+      "loss": 1.0369,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8867928558791878,
+      "learning_rate": 3.6693162117373377e-06,
+      "loss": 0.9638,
+      "step": 9346
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.064608980829705,
+      "learning_rate": 3.668317133770963e-06,
+      "loss": 0.8435,
+      "step": 9347
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.090945095084608,
+      "learning_rate": 3.667318113029968e-06,
+      "loss": 0.8655,
+      "step": 9348
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7635383576945804,
+      "learning_rate": 3.6663191495572827e-06,
+      "loss": 0.7352,
+      "step": 9349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.116145751028497,
+      "learning_rate": 3.6653202433958313e-06,
+      "loss": 0.8151,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.519750638583921,
+      "learning_rate": 3.664321394588542e-06,
+      "loss": 0.826,
+      "step": 9351
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1401105697844933,
+      "learning_rate": 3.663322603178339e-06,
+      "loss": 0.7483,
+      "step": 9352
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9024141421652534,
+      "learning_rate": 3.66232386920814e-06,
+      "loss": 0.8024,
+      "step": 9353
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3870884103710734,
+      "learning_rate": 3.661325192720862e-06,
+      "loss": 0.8128,
+      "step": 9354
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.709245734777865,
+      "learning_rate": 3.660326573759422e-06,
+      "loss": 0.6554,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9079317538681826,
+      "learning_rate": 3.659328012366735e-06,
+      "loss": 0.8227,
+      "step": 9356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.690877907094511,
+      "learning_rate": 3.658329508585707e-06,
+      "loss": 0.7745,
+      "step": 9357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.279711265788607,
+      "learning_rate": 3.657331062459249e-06,
+      "loss": 1.0086,
+      "step": 9358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0853115218747416,
+      "learning_rate": 3.6563326740302664e-06,
+      "loss": 0.7735,
+      "step": 9359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1610510284013589,
+      "learning_rate": 3.6553343433416598e-06,
+      "loss": 0.5931,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9238649260709249,
+      "learning_rate": 3.6543360704363305e-06,
+      "loss": 0.7634,
+      "step": 9361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5343673569905176,
+      "learning_rate": 3.6533378553571786e-06,
+      "loss": 0.7676,
+      "step": 9362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3074864877047356,
+      "learning_rate": 3.6523396981470995e-06,
+      "loss": 0.6881,
+      "step": 9363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.591563886216224,
+      "learning_rate": 3.6513415988489824e-06,
+      "loss": 0.6924,
+      "step": 9364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5705187382365366,
+      "learning_rate": 3.6503435575057205e-06,
+      "loss": 0.9538,
+      "step": 9365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1117372505588956,
+      "learning_rate": 3.6493455741602035e-06,
+      "loss": 0.9585,
+      "step": 9366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0272847276289943,
+      "learning_rate": 3.6483476488553143e-06,
+      "loss": 0.795,
+      "step": 9367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6253807582965085,
+      "learning_rate": 3.6473497816339375e-06,
+      "loss": 0.8811,
+      "step": 9368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1853533793454063,
+      "learning_rate": 3.6463519725389516e-06,
+      "loss": 0.9235,
+      "step": 9369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0022378248933808,
+      "learning_rate": 3.645354221613239e-06,
+      "loss": 0.8621,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2849213242202677,
+      "learning_rate": 3.6443565288996698e-06,
+      "loss": 0.7561,
+      "step": 9371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.133819993811088,
+      "learning_rate": 3.6433588944411207e-06,
+      "loss": 0.8576,
+      "step": 9372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5380976977447283,
+      "learning_rate": 3.642361318280461e-06,
+      "loss": 0.8237,
+      "step": 9373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1788709211177701,
+      "learning_rate": 3.64136380046056e-06,
+      "loss": 0.7376,
+      "step": 9374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.154494249132971,
+      "learning_rate": 3.640366341024281e-06,
+      "loss": 0.8316,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.065103989906127,
+      "learning_rate": 3.639368940014489e-06,
+      "loss": 0.7372,
+      "step": 9376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.346123096839984,
+      "learning_rate": 3.638371597474044e-06,
+      "loss": 0.789,
+      "step": 9377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.251264037143011,
+      "learning_rate": 3.637374313445805e-06,
+      "loss": 0.6477,
+      "step": 9378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.095751383688326,
+      "learning_rate": 3.6363770879726247e-06,
+      "loss": 0.8747,
+      "step": 9379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6972398987854806,
+      "learning_rate": 3.635379921097359e-06,
+      "loss": 0.8184,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.60812138604823,
+      "learning_rate": 3.634382812862857e-06,
+      "loss": 0.6015,
+      "step": 9381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9679138448981843,
+      "learning_rate": 3.6333857633119673e-06,
+      "loss": 0.8043,
+      "step": 9382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3389821186707036,
+      "learning_rate": 3.632388772487534e-06,
+      "loss": 0.6805,
+      "step": 9383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7402180109755276,
+      "learning_rate": 3.6313918404324e-06,
+      "loss": 0.711,
+      "step": 9384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.300107049771488,
+      "learning_rate": 3.63039496718941e-06,
+      "loss": 0.7831,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2429055025960234,
+      "learning_rate": 3.6293981528013965e-06,
+      "loss": 0.8398,
+      "step": 9386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9837454379699259,
+      "learning_rate": 3.6284013973111962e-06,
+      "loss": 0.689,
+      "step": 9387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.372371289754901,
+      "learning_rate": 3.627404700761645e-06,
+      "loss": 0.9149,
+      "step": 9388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5209711473505254,
+      "learning_rate": 3.6264080631955683e-06,
+      "loss": 0.7403,
+      "step": 9389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2347418113016095,
+      "learning_rate": 3.6254114846557977e-06,
+      "loss": 0.8348,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0579665430021157,
+      "learning_rate": 3.624414965185156e-06,
+      "loss": 0.7676,
+      "step": 9391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.061467969048857,
+      "learning_rate": 3.623418504826468e-06,
+      "loss": 0.8859,
+      "step": 9392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.021403531368909,
+      "learning_rate": 3.622422103622551e-06,
+      "loss": 0.9035,
+      "step": 9393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4089790959344892,
+      "learning_rate": 3.621425761616224e-06,
+      "loss": 0.784,
+      "step": 9394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7444978429438436,
+      "learning_rate": 3.6204294788503034e-06,
+      "loss": 0.9519,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3503685891610093,
+      "learning_rate": 3.6194332553676026e-06,
+      "loss": 0.8085,
+      "step": 9396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9955914836666906,
+      "learning_rate": 3.618437091210927e-06,
+      "loss": 0.8695,
+      "step": 9397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.001568690859064,
+      "learning_rate": 3.617440986423086e-06,
+      "loss": 0.7847,
+      "step": 9398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7854857741787744,
+      "learning_rate": 3.616444941046887e-06,
+      "loss": 0.7376,
+      "step": 9399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.791075043216099,
+      "learning_rate": 3.615448955125129e-06,
+      "loss": 0.9096,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9467242604098705,
+      "learning_rate": 3.614453028700613e-06,
+      "loss": 0.7127,
+      "step": 9401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1115917573621465,
+      "learning_rate": 3.613457161816135e-06,
+      "loss": 0.7164,
+      "step": 9402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2253470162440292,
+      "learning_rate": 3.6124613545144937e-06,
+      "loss": 0.694,
+      "step": 9403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9080132683978954,
+      "learning_rate": 3.6114656068384767e-06,
+      "loss": 0.6822,
+      "step": 9404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.114623747623969,
+      "learning_rate": 3.610469918830874e-06,
+      "loss": 0.8756,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.53353136719168,
+      "learning_rate": 3.6094742905344753e-06,
+      "loss": 0.7851,
+      "step": 9406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.374965779635867,
+      "learning_rate": 3.6084787219920616e-06,
+      "loss": 0.8001,
+      "step": 9407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0437629003771676,
+      "learning_rate": 3.6074832132464165e-06,
+      "loss": 0.7366,
+      "step": 9408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.459838104645164,
+      "learning_rate": 3.6064877643403194e-06,
+      "loss": 0.821,
+      "step": 9409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.053686785107632,
+      "learning_rate": 3.605492375316548e-06,
+      "loss": 0.8718,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2989792492421666,
+      "learning_rate": 3.604497046217873e-06,
+      "loss": 0.8448,
+      "step": 9411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28122772258076,
+      "learning_rate": 3.6035017770870672e-06,
+      "loss": 1.0613,
+      "step": 9412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.092056138377427,
+      "learning_rate": 3.6025065679669025e-06,
+      "loss": 0.8084,
+      "step": 9413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.298880215408962,
+      "learning_rate": 3.601511418900143e-06,
+      "loss": 0.7894,
+      "step": 9414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3607349384031893,
+      "learning_rate": 3.600516329929551e-06,
+      "loss": 0.9763,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.048498968538042,
+      "learning_rate": 3.599521301097888e-06,
+      "loss": 0.6286,
+      "step": 9416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.279457065985713,
+      "learning_rate": 3.598526332447917e-06,
+      "loss": 0.7112,
+      "step": 9417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8985989802323255,
+      "learning_rate": 3.597531424022388e-06,
+      "loss": 1.0306,
+      "step": 9418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1120800992493076,
+      "learning_rate": 3.5965365758640587e-06,
+      "loss": 0.8447,
+      "step": 9419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6664364268375564,
+      "learning_rate": 3.5955417880156766e-06,
+      "loss": 0.8236,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0479869343074952,
+      "learning_rate": 3.594547060519995e-06,
+      "loss": 0.8351,
+      "step": 9421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.159227151740994,
+      "learning_rate": 3.5935523934197537e-06,
+      "loss": 0.8152,
+      "step": 9422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.43398248197263,
+      "learning_rate": 3.592557786757699e-06,
+      "loss": 0.9649,
+      "step": 9423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.269252308889,
+      "learning_rate": 3.591563240576572e-06,
+      "loss": 0.7718,
+      "step": 9424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.619890150419704,
+      "learning_rate": 3.590568754919107e-06,
+      "loss": 0.9431,
+      "step": 9425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9391640824666005,
+      "learning_rate": 3.5895743298280423e-06,
+      "loss": 0.952,
+      "step": 9426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4706494503372953,
+      "learning_rate": 3.5885799653461107e-06,
+      "loss": 0.9591,
+      "step": 9427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1370304449028423,
+      "learning_rate": 3.587585661516042e-06,
+      "loss": 0.9019,
+      "step": 9428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.321065774251458,
+      "learning_rate": 3.5865914183805606e-06,
+      "loss": 0.7479,
+      "step": 9429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1865958497296996,
+      "learning_rate": 3.585597235982394e-06,
+      "loss": 0.6297,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.905690153977676,
+      "learning_rate": 3.584603114364266e-06,
+      "loss": 0.8422,
+      "step": 9431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.203475907801648,
+      "learning_rate": 3.5836090535688944e-06,
+      "loss": 0.6429,
+      "step": 9432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8702824565171796,
+      "learning_rate": 3.582615053638996e-06,
+      "loss": 0.9723,
+      "step": 9433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.636430327209334,
+      "learning_rate": 3.581621114617284e-06,
+      "loss": 0.9077,
+      "step": 9434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0461066474013707,
+      "learning_rate": 3.580627236546474e-06,
+      "loss": 0.8116,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8971191988054517,
+      "learning_rate": 3.5796334194692704e-06,
+      "loss": 0.9111,
+      "step": 9436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.724092351537121,
+      "learning_rate": 3.578639663428384e-06,
+      "loss": 0.9099,
+      "step": 9437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2099577610235657,
+      "learning_rate": 3.5776459684665146e-06,
+      "loss": 0.692,
+      "step": 9438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6757550824747325,
+      "learning_rate": 3.5766523346263682e-06,
+      "loss": 0.7611,
+      "step": 9439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7404355746115552,
+      "learning_rate": 3.5756587619506388e-06,
+      "loss": 0.7271,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8537766498252444,
+      "learning_rate": 3.5746652504820263e-06,
+      "loss": 0.682,
+      "step": 9441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0452783710529028,
+      "learning_rate": 3.5736718002632222e-06,
+      "loss": 0.7654,
+      "step": 9442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.530243499353887,
+      "learning_rate": 3.572678411336916e-06,
+      "loss": 0.6888,
+      "step": 9443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1979753050948956,
+      "learning_rate": 3.571685083745798e-06,
+      "loss": 0.9245,
+      "step": 9444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0507883227702273,
+      "learning_rate": 3.5706918175325523e-06,
+      "loss": 0.9178,
+      "step": 9445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0497710370374054,
+      "learning_rate": 3.5696986127398646e-06,
+      "loss": 0.8404,
+      "step": 9446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6332466173460722,
+      "learning_rate": 3.5687054694104117e-06,
+      "loss": 0.8558,
+      "step": 9447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2712584511333427,
+      "learning_rate": 3.5677123875868713e-06,
+      "loss": 0.9693,
+      "step": 9448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.818094390426108,
+      "learning_rate": 3.56671936731192e-06,
+      "loss": 0.7298,
+      "step": 9449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9599304911880977,
+      "learning_rate": 3.5657264086282317e-06,
+      "loss": 0.8216,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9392286859419492,
+      "learning_rate": 3.564733511578472e-06,
+      "loss": 0.7484,
+      "step": 9451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.104311438881113,
+      "learning_rate": 3.5637406762053094e-06,
+      "loss": 0.6373,
+      "step": 9452
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.784468367285203,
+      "learning_rate": 3.562747902551411e-06,
+      "loss": 0.8894,
+      "step": 9453
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8935493095609566,
+      "learning_rate": 3.561755190659434e-06,
+      "loss": 0.6413,
+      "step": 9454
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.10819029907616,
+      "learning_rate": 3.5607625405720402e-06,
+      "loss": 0.8519,
+      "step": 9455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0337516293696063,
+      "learning_rate": 3.559769952331884e-06,
+      "loss": 0.7446,
+      "step": 9456
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0695170965632794,
+      "learning_rate": 3.5587774259816234e-06,
+      "loss": 0.6583,
+      "step": 9457
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.281704598266073,
+      "learning_rate": 3.557784961563904e-06,
+      "loss": 0.777,
+      "step": 9458
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1609637072646315,
+      "learning_rate": 3.556792559121377e-06,
+      "loss": 0.783,
+      "step": 9459
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.455690446132792,
+      "learning_rate": 3.5558002186966906e-06,
+      "loss": 0.8155,
+      "step": 9460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.637722360778467,
+      "learning_rate": 3.554807940332483e-06,
+      "loss": 0.6676,
+      "step": 9461
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.339023671284605,
+      "learning_rate": 3.5538157240713964e-06,
+      "loss": 0.8365,
+      "step": 9462
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8650025462569197,
+      "learning_rate": 3.5528235699560697e-06,
+      "loss": 0.9752,
+      "step": 9463
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3354906964533266,
+      "learning_rate": 3.5518314780291384e-06,
+      "loss": 0.8437,
+      "step": 9464
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0044796368639837,
+      "learning_rate": 3.5508394483332333e-06,
+      "loss": 0.641,
+      "step": 9465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9545752477011917,
+      "learning_rate": 3.549847480910983e-06,
+      "loss": 0.6674,
+      "step": 9466
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1827554380140546,
+      "learning_rate": 3.548855575805016e-06,
+      "loss": 0.9202,
+      "step": 9467
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0753996447225456,
+      "learning_rate": 3.547863733057959e-06,
+      "loss": 0.7253,
+      "step": 9468
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9876314066971155,
+      "learning_rate": 3.5468719527124294e-06,
+      "loss": 0.8677,
+      "step": 9469
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1215120278611745,
+      "learning_rate": 3.5458802348110488e-06,
+      "loss": 0.7166,
+      "step": 9470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5599653599844903,
+      "learning_rate": 3.544888579396435e-06,
+      "loss": 0.8765,
+      "step": 9471
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.564471205588936,
+      "learning_rate": 3.543896986511197e-06,
+      "loss": 0.7191,
+      "step": 9472
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.24702214766598,
+      "learning_rate": 3.5429054561979485e-06,
+      "loss": 0.7933,
+      "step": 9473
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.810278680970197,
+      "learning_rate": 3.541913988499299e-06,
+      "loss": 0.5675,
+      "step": 9474
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1532620521929666,
+      "learning_rate": 3.5409225834578536e-06,
+      "loss": 0.847,
+      "step": 9475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.811868596811013,
+      "learning_rate": 3.5399312411162124e-06,
+      "loss": 0.9067,
+      "step": 9476
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4661922336432265,
+      "learning_rate": 3.5389399615169774e-06,
+      "loss": 0.7851,
+      "step": 9477
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8795871291173163,
+      "learning_rate": 3.5379487447027483e-06,
+      "loss": 0.7816,
+      "step": 9478
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2155852435957404,
+      "learning_rate": 3.5369575907161167e-06,
+      "loss": 0.8147,
+      "step": 9479
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.409428882500439,
+      "learning_rate": 3.5359664995996746e-06,
+      "loss": 0.9085,
+      "step": 9480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.670234619101944,
+      "learning_rate": 3.5349754713960126e-06,
+      "loss": 0.7092,
+      "step": 9481
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.221430025233516,
+      "learning_rate": 3.5339845061477195e-06,
+      "loss": 0.8799,
+      "step": 9482
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8388624331016599,
+      "learning_rate": 3.5329936038973756e-06,
+      "loss": 0.6843,
+      "step": 9483
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.287189743895215,
+      "learning_rate": 3.5320027646875643e-06,
+      "loss": 0.8446,
+      "step": 9484
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.196014819005786,
+      "learning_rate": 3.5310119885608625e-06,
+      "loss": 0.6429,
+      "step": 9485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0540296920508627,
+      "learning_rate": 3.5300212755598497e-06,
+      "loss": 0.7607,
+      "step": 9486
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.017742657549854,
+      "learning_rate": 3.5290306257270935e-06,
+      "loss": 0.6268,
+      "step": 9487
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.482868674078097,
+      "learning_rate": 3.5280400391051693e-06,
+      "loss": 0.929,
+      "step": 9488
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0398301737036615,
+      "learning_rate": 3.5270495157366434e-06,
+      "loss": 0.6342,
+      "step": 9489
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.524095502995187,
+      "learning_rate": 3.5260590556640783e-06,
+      "loss": 0.6998,
+      "step": 9490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0720058042532883,
+      "learning_rate": 3.5250686589300387e-06,
+      "loss": 0.7833,
+      "step": 9491
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1030212493293496,
+      "learning_rate": 3.524078325577084e-06,
+      "loss": 0.7934,
+      "step": 9492
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.693641880625892,
+      "learning_rate": 3.5230880556477707e-06,
+      "loss": 0.7804,
+      "step": 9493
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.916578014080249,
+      "learning_rate": 3.5220978491846534e-06,
+      "loss": 0.8594,
+      "step": 9494
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2822652527939893,
+      "learning_rate": 3.521107706230282e-06,
+      "loss": 0.7528,
+      "step": 9495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4439366345333244,
+      "learning_rate": 3.5201176268272058e-06,
+      "loss": 0.7917,
+      "step": 9496
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3851635331798122,
+      "learning_rate": 3.519127611017973e-06,
+      "loss": 0.7902,
+      "step": 9497
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0304424406818375,
+      "learning_rate": 3.5181376588451244e-06,
+      "loss": 0.6687,
+      "step": 9498
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7161272257007543,
+      "learning_rate": 3.517147770351199e-06,
+      "loss": 1.0787,
+      "step": 9499
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4325783579080222,
+      "learning_rate": 3.5161579455787388e-06,
+      "loss": 1.0201,
+      "step": 9500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0263862446197987,
+      "learning_rate": 3.5151681845702747e-06,
+      "loss": 0.8553,
+      "step": 9501
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.590508701972718,
+      "learning_rate": 3.5141784873683418e-06,
+      "loss": 0.7424,
+      "step": 9502
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2966908893387035,
+      "learning_rate": 3.513188854015466e-06,
+      "loss": 0.6924,
+      "step": 9503
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6138094515975276,
+      "learning_rate": 3.5121992845541797e-06,
+      "loss": 0.87,
+      "step": 9504
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.278204866214059,
+      "learning_rate": 3.5112097790270017e-06,
+      "loss": 0.9181,
+      "step": 9505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1152341536263344,
+      "learning_rate": 3.5102203374764555e-06,
+      "loss": 0.8224,
+      "step": 9506
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1709038526611533,
+      "learning_rate": 3.5092309599450596e-06,
+      "loss": 0.6867,
+      "step": 9507
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0575042431142863,
+      "learning_rate": 3.50824164647533e-06,
+      "loss": 0.927,
+      "step": 9508
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8486181358408715,
+      "learning_rate": 3.507252397109777e-06,
+      "loss": 0.5468,
+      "step": 9509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0822698686918644,
+      "learning_rate": 3.506263211890914e-06,
+      "loss": 0.8125,
+      "step": 9510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2351624472963192,
+      "learning_rate": 3.505274090861249e-06,
+      "loss": 0.6571,
+      "step": 9511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6498809446617255,
+      "learning_rate": 3.504285034063285e-06,
+      "loss": 0.8739,
+      "step": 9512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.56074636756126,
+      "learning_rate": 3.503296041539522e-06,
+      "loss": 0.8494,
+      "step": 9513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2817501619062286,
+      "learning_rate": 3.5023071133324627e-06,
+      "loss": 0.6547,
+      "step": 9514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2404028114685848,
+      "learning_rate": 3.501318249484604e-06,
+      "loss": 0.7406,
+      "step": 9515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.64137931097485,
+      "learning_rate": 3.5003294500384367e-06,
+      "loss": 0.9209,
+      "step": 9516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.141599811586741,
+      "learning_rate": 3.4993407150364513e-06,
+      "loss": 0.847,
+      "step": 9517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5363613311346724,
+      "learning_rate": 3.4983520445211403e-06,
+      "loss": 0.8829,
+      "step": 9518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.06381886136898,
+      "learning_rate": 3.497363438534984e-06,
+      "loss": 0.7099,
+      "step": 9519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.467259790470079,
+      "learning_rate": 3.496374897120467e-06,
+      "loss": 0.7118,
+      "step": 9520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0499511766080216,
+      "learning_rate": 3.495386420320071e-06,
+      "loss": 0.7843,
+      "step": 9521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8792429555348638,
+      "learning_rate": 3.4943980081762717e-06,
+      "loss": 0.7793,
+      "step": 9522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6780237580373987,
+      "learning_rate": 3.493409660731542e-06,
+      "loss": 0.9333,
+      "step": 9523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.089487953450703,
+      "learning_rate": 3.4924213780283545e-06,
+      "loss": 0.6886,
+      "step": 9524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2881864827571388,
+      "learning_rate": 3.4914331601091793e-06,
+      "loss": 0.8111,
+      "step": 9525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9264750044324728,
+      "learning_rate": 3.4904450070164803e-06,
+      "loss": 0.7741,
+      "step": 9526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0402299570501268,
+      "learning_rate": 3.4894569187927204e-06,
+      "loss": 0.8043,
+      "step": 9527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.052291813139687,
+      "learning_rate": 3.48846889548036e-06,
+      "loss": 1.0941,
+      "step": 9528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8786110096434743,
+      "learning_rate": 3.4874809371218608e-06,
+      "loss": 0.8871,
+      "step": 9529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2364584895491824,
+      "learning_rate": 3.486493043759673e-06,
+      "loss": 0.6458,
+      "step": 9530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.108715303916919,
+      "learning_rate": 3.4855052154362485e-06,
+      "loss": 0.9012,
+      "step": 9531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.393269555986215,
+      "learning_rate": 3.4845174521940384e-06,
+      "loss": 0.6074,
+      "step": 9532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.96284944091816,
+      "learning_rate": 3.4835297540754896e-06,
+      "loss": 0.6566,
+      "step": 9533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3752069388686405,
+      "learning_rate": 3.4825421211230437e-06,
+      "loss": 0.7045,
+      "step": 9534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4000931373941046,
+      "learning_rate": 3.481554553379143e-06,
+      "loss": 0.8303,
+      "step": 9535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0034588396512527,
+      "learning_rate": 3.4805670508862255e-06,
+      "loss": 0.8294,
+      "step": 9536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0979458270827465,
+      "learning_rate": 3.4795796136867237e-06,
+      "loss": 0.8267,
+      "step": 9537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.176421096923227,
+      "learning_rate": 3.4785922418230722e-06,
+      "loss": 0.6806,
+      "step": 9538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.933522707243515,
+      "learning_rate": 3.4776049353377016e-06,
+      "loss": 0.6599,
+      "step": 9539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6398445834941502,
+      "learning_rate": 3.476617694273039e-06,
+      "loss": 0.6653,
+      "step": 9540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.149970177842705,
+      "learning_rate": 3.4756305186715046e-06,
+      "loss": 0.7612,
+      "step": 9541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8387945361151887,
+      "learning_rate": 3.474643408575521e-06,
+      "loss": 0.8801,
+      "step": 9542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1911581438399312,
+      "learning_rate": 3.47365636402751e-06,
+      "loss": 0.6001,
+      "step": 9543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3124610825734786,
+      "learning_rate": 3.4726693850698824e-06,
+      "loss": 0.8271,
+      "step": 9544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5796670072656163,
+      "learning_rate": 3.4716824717450538e-06,
+      "loss": 0.7805,
+      "step": 9545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.13472732516539,
+      "learning_rate": 3.4706956240954317e-06,
+      "loss": 0.9119,
+      "step": 9546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0802039312253,
+      "learning_rate": 3.4697088421634266e-06,
+      "loss": 0.8145,
+      "step": 9547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.094634090255983,
+      "learning_rate": 3.4687221259914394e-06,
+      "loss": 0.9033,
+      "step": 9548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.261328963639171,
+      "learning_rate": 3.467735475621873e-06,
+      "loss": 0.9801,
+      "step": 9549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3727430109531005,
+      "learning_rate": 3.4667488910971252e-06,
+      "loss": 0.7748,
+      "step": 9550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0741703072667868,
+      "learning_rate": 3.4657623724595944e-06,
+      "loss": 1.0265,
+      "step": 9551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.553177483049563,
+      "learning_rate": 3.4647759197516687e-06,
+      "loss": 0.9021,
+      "step": 9552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5867653362193026,
+      "learning_rate": 3.4637895330157423e-06,
+      "loss": 0.5829,
+      "step": 9553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3305657505824176,
+      "learning_rate": 3.4628032122942024e-06,
+      "loss": 0.899,
+      "step": 9554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.398872470806113,
+      "learning_rate": 3.461816957629429e-06,
+      "loss": 0.7363,
+      "step": 9555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.223403062335116,
+      "learning_rate": 3.4608307690638075e-06,
+      "loss": 0.8526,
+      "step": 9556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9972443618271563,
+      "learning_rate": 3.459844646639717e-06,
+      "loss": 0.8624,
+      "step": 9557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0487832486553317,
+      "learning_rate": 3.4588585903995314e-06,
+      "loss": 0.7572,
+      "step": 9558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.335727829737713,
+      "learning_rate": 3.4578726003856245e-06,
+      "loss": 0.892,
+      "step": 9559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.156678227610281,
+      "learning_rate": 3.4568866766403665e-06,
+      "loss": 0.8339,
+      "step": 9560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1427050026729204,
+      "learning_rate": 3.4559008192061254e-06,
+      "loss": 1.0498,
+      "step": 9561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1936402122670904,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.8527,
+      "step": 9562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0840667899616019,
+      "learning_rate": 3.453929303440145e-06,
+      "loss": 0.78,
+      "step": 9563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9383979425327067,
+      "learning_rate": 3.4529436451931263e-06,
+      "loss": 0.9127,
+      "step": 9564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2897213170039918,
+      "learning_rate": 3.4519580534265674e-06,
+      "loss": 0.6964,
+      "step": 9565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4214870390849486,
+      "learning_rate": 3.4509725281828155e-06,
+      "loss": 0.713,
+      "step": 9566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.491822813692517,
+      "learning_rate": 3.449987069504226e-06,
+      "loss": 0.8587,
+      "step": 9567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2012717346648056,
+      "learning_rate": 3.4490016774331426e-06,
+      "loss": 0.9239,
+      "step": 9568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.973519148527666,
+      "learning_rate": 3.448016352011914e-06,
+      "loss": 0.8265,
+      "step": 9569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1260307032108603,
+      "learning_rate": 3.4470310932828766e-06,
+      "loss": 0.6484,
+      "step": 9570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.558276785389996,
+      "learning_rate": 3.4460459012883716e-06,
+      "loss": 0.9414,
+      "step": 9571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9406557158600604,
+      "learning_rate": 3.4450607760707373e-06,
+      "loss": 0.8042,
+      "step": 9572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.303511257213764,
+      "learning_rate": 3.4440757176723037e-06,
+      "loss": 0.8432,
+      "step": 9573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265038320807143,
+      "learning_rate": 3.4430907261354e-06,
+      "loss": 0.7791,
+      "step": 9574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.202985473097859,
+      "learning_rate": 3.442105801502356e-06,
+      "loss": 0.8009,
+      "step": 9575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3651528475082144,
+      "learning_rate": 3.441120943815497e-06,
+      "loss": 0.832,
+      "step": 9576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0933833397754764,
+      "learning_rate": 3.440136153117142e-06,
+      "loss": 1.064,
+      "step": 9577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1345483553795,
+      "learning_rate": 3.4391514294496097e-06,
+      "loss": 0.5762,
+      "step": 9578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.799086624993349,
+      "learning_rate": 3.438166772855218e-06,
+      "loss": 0.6844,
+      "step": 9579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4907763993079315,
+      "learning_rate": 3.4371821833762776e-06,
+      "loss": 0.7248,
+      "step": 9580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8773259589885318,
+      "learning_rate": 3.4361976610550996e-06,
+      "loss": 0.5615,
+      "step": 9581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.138206975066467,
+      "learning_rate": 3.43521320593399e-06,
+      "loss": 0.976,
+      "step": 9582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.934506294977628,
+      "learning_rate": 3.4342288180552556e-06,
+      "loss": 0.8827,
+      "step": 9583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.884743926491262,
+      "learning_rate": 3.4332444974611946e-06,
+      "loss": 0.5879,
+      "step": 9584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.444204655559446,
+      "learning_rate": 3.4322602441941064e-06,
+      "loss": 0.9059,
+      "step": 9585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9650231241124656,
+      "learning_rate": 3.4312760582962877e-06,
+      "loss": 0.8762,
+      "step": 9586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.854154223300558,
+      "learning_rate": 3.4302919398100325e-06,
+      "loss": 0.7952,
+      "step": 9587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4641818313926585,
+      "learning_rate": 3.4293078887776264e-06,
+      "loss": 0.6972,
+      "step": 9588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.049652146104433,
+      "learning_rate": 3.428323905241358e-06,
+      "loss": 0.7201,
+      "step": 9589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.013650355788179,
+      "learning_rate": 3.427339989243514e-06,
+      "loss": 0.6414,
+      "step": 9590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.17900734245979,
+      "learning_rate": 3.426356140826372e-06,
+      "loss": 0.7701,
+      "step": 9591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.695606722284027,
+      "learning_rate": 3.4253723600322105e-06,
+      "loss": 0.7596,
+      "step": 9592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.849177827068188,
+      "learning_rate": 3.424388646903305e-06,
+      "loss": 0.7373,
+      "step": 9593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3829529148693167,
+      "learning_rate": 3.4234050014819308e-06,
+      "loss": 0.7993,
+      "step": 9594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.042912842774713,
+      "learning_rate": 3.422421423810354e-06,
+      "loss": 0.6068,
+      "step": 9595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7903817486539693,
+      "learning_rate": 3.421437913930842e-06,
+      "loss": 0.6807,
+      "step": 9596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8511252971500627,
+      "learning_rate": 3.420454471885659e-06,
+      "loss": 0.7185,
+      "step": 9597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3379767668668188,
+      "learning_rate": 3.4194710977170635e-06,
+      "loss": 0.8459,
+      "step": 9598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.351406875946986,
+      "learning_rate": 3.4184877914673155e-06,
+      "loss": 0.9447,
+      "step": 9599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1594417333278937,
+      "learning_rate": 3.41750455317867e-06,
+      "loss": 0.6938,
+      "step": 9600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.371500374753652,
+      "learning_rate": 3.4165213828933797e-06,
+      "loss": 0.6575,
+      "step": 9601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1246824149033916,
+      "learning_rate": 3.4155382806536897e-06,
+      "loss": 0.9314,
+      "step": 9602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.545330426104065,
+      "learning_rate": 3.4145552465018483e-06,
+      "loss": 0.8523,
+      "step": 9603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.492825386656528,
+      "learning_rate": 3.4135722804801004e-06,
+      "loss": 0.8489,
+      "step": 9604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1106974270930488,
+      "learning_rate": 3.4125893826306855e-06,
+      "loss": 0.6838,
+      "step": 9605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5603442660356888,
+      "learning_rate": 3.411606552995838e-06,
+      "loss": 0.8735,
+      "step": 9606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.325510071171768,
+      "learning_rate": 3.4106237916177954e-06,
+      "loss": 0.775,
+      "step": 9607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1120341457841305,
+      "learning_rate": 3.4096410985387898e-06,
+      "loss": 0.8928,
+      "step": 9608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.545498780673469,
+      "learning_rate": 3.4086584738010455e-06,
+      "loss": 0.7586,
+      "step": 9609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3925703055777108,
+      "learning_rate": 3.4076759174467923e-06,
+      "loss": 0.6653,
+      "step": 9610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8337334382923745,
+      "learning_rate": 3.4066934295182496e-06,
+      "loss": 0.7878,
+      "step": 9611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2012020418577536,
+      "learning_rate": 3.4057110100576408e-06,
+      "loss": 0.6447,
+      "step": 9612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1173032596812784,
+      "learning_rate": 3.4047286591071794e-06,
+      "loss": 0.6024,
+      "step": 9613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1563398407620282,
+      "learning_rate": 3.4037463767090807e-06,
+      "loss": 0.6969,
+      "step": 9614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.301650356338058,
+      "learning_rate": 3.402764162905554e-06,
+      "loss": 0.7739,
+      "step": 9615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7995677538792998,
+      "learning_rate": 3.401782017738811e-06,
+      "loss": 0.8408,
+      "step": 9616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0918109765467054,
+      "learning_rate": 3.400799941251052e-06,
+      "loss": 0.7224,
+      "step": 9617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140562496609572,
+      "learning_rate": 3.3998179334844823e-06,
+      "loss": 0.8325,
+      "step": 9618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2442779715100616,
+      "learning_rate": 3.3988359944812997e-06,
+      "loss": 0.8608,
+      "step": 9619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.923858747278879,
+      "learning_rate": 3.3978541242837003e-06,
+      "loss": 0.8919,
+      "step": 9620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.591609775663181,
+      "learning_rate": 3.396872322933876e-06,
+      "loss": 0.8447,
+      "step": 9621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2341312946195206,
+      "learning_rate": 3.395890590474019e-06,
+      "loss": 0.6838,
+      "step": 9622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2891989618085,
+      "learning_rate": 3.3949089269463178e-06,
+      "loss": 0.8205,
+      "step": 9623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.374022513741114,
+      "learning_rate": 3.3939273323929533e-06,
+      "loss": 0.6924,
+      "step": 9624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.093028972903149,
+      "learning_rate": 3.3929458068561073e-06,
+      "loss": 0.8259,
+      "step": 9625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0928626375587,
+      "learning_rate": 3.391964350377961e-06,
+      "loss": 0.9128,
+      "step": 9626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3218129073019083,
+      "learning_rate": 3.3909829630006865e-06,
+      "loss": 0.7675,
+      "step": 9627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.536726811690236,
+      "learning_rate": 3.390001644766458e-06,
+      "loss": 0.8686,
+      "step": 9628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0280665265325144,
+      "learning_rate": 3.3890203957174437e-06,
+      "loss": 0.9777,
+      "step": 9629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4667784513897955,
+      "learning_rate": 3.388039215895813e-06,
+      "loss": 0.8906,
+      "step": 9630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.491533827105458,
+      "learning_rate": 3.3870581053437244e-06,
+      "loss": 0.8134,
+      "step": 9631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.487464846446084,
+      "learning_rate": 3.3860770641033417e-06,
+      "loss": 0.7358,
+      "step": 9632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3855412573527417,
+      "learning_rate": 3.385096092216823e-06,
+      "loss": 0.6894,
+      "step": 9633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4853432060412017,
+      "learning_rate": 3.3841151897263234e-06,
+      "loss": 0.8677,
+      "step": 9634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9711978571073523,
+      "learning_rate": 3.3831343566739904e-06,
+      "loss": 0.7277,
+      "step": 9635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2225894944406583,
+      "learning_rate": 3.3821535931019744e-06,
+      "loss": 0.8926,
+      "step": 9636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9809636187802115,
+      "learning_rate": 3.3811728990524245e-06,
+      "loss": 0.9671,
+      "step": 9637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.233534846697212,
+      "learning_rate": 3.380192274567479e-06,
+      "loss": 0.6766,
+      "step": 9638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5721923972744323,
+      "learning_rate": 3.379211719689278e-06,
+      "loss": 0.7954,
+      "step": 9639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.871942117247422,
+      "learning_rate": 3.3782312344599598e-06,
+      "loss": 0.6683,
+      "step": 9640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9787726460800261,
+      "learning_rate": 3.3772508189216586e-06,
+      "loss": 0.6483,
+      "step": 9641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0918012282986123,
+      "learning_rate": 3.376270473116503e-06,
+      "loss": 0.6704,
+      "step": 9642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.554020028872242,
+      "learning_rate": 3.37529019708662e-06,
+      "loss": 0.7726,
+      "step": 9643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.621354093346709,
+      "learning_rate": 3.3743099908741385e-06,
+      "loss": 0.9126,
+      "step": 9644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8469056529434433,
+      "learning_rate": 3.3733298545211748e-06,
+      "loss": 0.7608,
+      "step": 9645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.358634700330488,
+      "learning_rate": 3.37234978806985e-06,
+      "loss": 1.0057,
+      "step": 9646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2538574597078003,
+      "learning_rate": 3.3713697915622806e-06,
+      "loss": 0.7951,
+      "step": 9647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.377983401121129,
+      "learning_rate": 3.3703898650405796e-06,
+      "loss": 0.795,
+      "step": 9648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.047382522519133,
+      "learning_rate": 3.3694100085468535e-06,
+      "loss": 0.9748,
+      "step": 9649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.445188805545184,
+      "learning_rate": 3.3684302221232106e-06,
+      "loss": 0.8281,
+      "step": 9650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6725380885028795,
+      "learning_rate": 3.3674505058117556e-06,
+      "loss": 0.8674,
+      "step": 9651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9686300858539743,
+      "learning_rate": 3.36647085965459e-06,
+      "loss": 1.0301,
+      "step": 9652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.977820297861815,
+      "learning_rate": 3.365491283693807e-06,
+      "loss": 0.7748,
+      "step": 9653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.983978110604627,
+      "learning_rate": 3.364511777971504e-06,
+      "loss": 0.66,
+      "step": 9654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.253292808918471,
+      "learning_rate": 3.363532342529775e-06,
+      "loss": 0.9218,
+      "step": 9655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9678187688510316,
+      "learning_rate": 3.3625529774107028e-06,
+      "loss": 1.0051,
+      "step": 9656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8249239748144988,
+      "learning_rate": 3.361573682656377e-06,
+      "loss": 0.8307,
+      "step": 9657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2763528931390247,
+      "learning_rate": 3.360594458308879e-06,
+      "loss": 0.7716,
+      "step": 9658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9799948060032122,
+      "learning_rate": 3.3596153044102897e-06,
+      "loss": 0.9991,
+      "step": 9659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8257306138623661,
+      "learning_rate": 3.358636221002682e-06,
+      "loss": 0.7831,
+      "step": 9660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.047482831162953,
+      "learning_rate": 3.357657208128132e-06,
+      "loss": 0.9622,
+      "step": 9661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9783857694468843,
+      "learning_rate": 3.356678265828711e-06,
+      "loss": 0.8035,
+      "step": 9662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1995666073589601,
+      "learning_rate": 3.3556993941464834e-06,
+      "loss": 0.636,
+      "step": 9663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.760429701313237,
+      "learning_rate": 3.354720593123514e-06,
+      "loss": 0.7142,
+      "step": 9664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0325458192581227,
+      "learning_rate": 3.353741862801866e-06,
+      "loss": 0.613,
+      "step": 9665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3740162818887967,
+      "learning_rate": 3.352763203223598e-06,
+      "loss": 0.6705,
+      "step": 9666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6844457465015896,
+      "learning_rate": 3.351784614430761e-06,
+      "loss": 0.8242,
+      "step": 9667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9822980099729643,
+      "learning_rate": 3.3508060964654106e-06,
+      "loss": 0.7042,
+      "step": 9668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2164172220640252,
+      "learning_rate": 3.349827649369596e-06,
+      "loss": 0.8399,
+      "step": 9669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.849802424064979,
+      "learning_rate": 3.348849273185362e-06,
+      "loss": 0.7061,
+      "step": 9670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0830929313543933,
+      "learning_rate": 3.3478709679547527e-06,
+      "loss": 0.7743,
+      "step": 9671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5446246438263493,
+      "learning_rate": 3.346892733719806e-06,
+      "loss": 0.9442,
+      "step": 9672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1586813720153335,
+      "learning_rate": 3.3459145705225627e-06,
+      "loss": 0.6828,
+      "step": 9673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.807539379446746,
+      "learning_rate": 3.3449364784050515e-06,
+      "loss": 0.7738,
+      "step": 9674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4928876400078077,
+      "learning_rate": 3.343958457409307e-06,
+      "loss": 0.7837,
+      "step": 9675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0605752218622997,
+      "learning_rate": 3.3429805075773547e-06,
+      "loss": 0.7937,
+      "step": 9676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.17253363312198,
+      "learning_rate": 3.3420026289512232e-06,
+      "loss": 0.7158,
+      "step": 9677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3273565100884075,
+      "learning_rate": 3.341024821572928e-06,
+      "loss": 0.8808,
+      "step": 9678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8171300675834363,
+      "learning_rate": 3.3400470854844925e-06,
+      "loss": 0.8787,
+      "step": 9679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.348091240811608,
+      "learning_rate": 3.3390694207279308e-06,
+      "loss": 0.7397,
+      "step": 9680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1185160827360807,
+      "learning_rate": 3.3380918273452557e-06,
+      "loss": 0.8581,
+      "step": 9681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.180378188508615,
+      "learning_rate": 3.3371143053784737e-06,
+      "loss": 0.8551,
+      "step": 9682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0321636461429975,
+      "learning_rate": 3.336136854869595e-06,
+      "loss": 0.8493,
+      "step": 9683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1546403894165278,
+      "learning_rate": 3.3351594758606222e-06,
+      "loss": 0.7475,
+      "step": 9684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4924752492063007,
+      "learning_rate": 3.334182168393554e-06,
+      "loss": 0.8134,
+      "step": 9685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09440003597001,
+      "learning_rate": 3.3332049325103867e-06,
+      "loss": 0.7326,
+      "step": 9686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.080227783560281,
+      "learning_rate": 3.332227768253116e-06,
+      "loss": 0.66,
+      "step": 9687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.072103170053193,
+      "learning_rate": 3.3312506756637343e-06,
+      "loss": 0.8878,
+      "step": 9688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1009656904216287,
+      "learning_rate": 3.3302736547842263e-06,
+      "loss": 0.784,
+      "step": 9689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5363436949467735,
+      "learning_rate": 3.3292967056565766e-06,
+      "loss": 0.7129,
+      "step": 9690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.792256035821822,
+      "learning_rate": 3.3283198283227715e-06,
+      "loss": 0.794,
+      "step": 9691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.335931990683661,
+      "learning_rate": 3.3273430228247834e-06,
+      "loss": 0.9122,
+      "step": 9692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.129215873476766,
+      "learning_rate": 3.3263662892045924e-06,
+      "loss": 0.7821,
+      "step": 9693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4431947238922174,
+      "learning_rate": 3.3253896275041677e-06,
+      "loss": 0.8957,
+      "step": 9694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3199216557769637,
+      "learning_rate": 3.324413037765483e-06,
+      "loss": 0.7792,
+      "step": 9695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.031454871709978,
+      "learning_rate": 3.3234365200304997e-06,
+      "loss": 0.7509,
+      "step": 9696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.39730048998098,
+      "learning_rate": 3.322460074341183e-06,
+      "loss": 0.561,
+      "step": 9697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.313446230192015,
+      "learning_rate": 3.321483700739494e-06,
+      "loss": 0.8717,
+      "step": 9698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0415828829957354,
+      "learning_rate": 3.3205073992673885e-06,
+      "loss": 0.552,
+      "step": 9699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.225829089804333,
+      "learning_rate": 3.31953116996682e-06,
+      "loss": 0.7264,
+      "step": 9700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1137793505644873,
+      "learning_rate": 3.3185550128797387e-06,
+      "loss": 0.7341,
+      "step": 9701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0646035705416175,
+      "learning_rate": 3.317578928048096e-06,
+      "loss": 0.7816,
+      "step": 9702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2054494586408513,
+      "learning_rate": 3.3166029155138324e-06,
+      "loss": 0.8267,
+      "step": 9703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.617096233267351,
+      "learning_rate": 3.3156269753188895e-06,
+      "loss": 0.777,
+      "step": 9704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.192843882871488,
+      "learning_rate": 3.314651107505207e-06,
+      "loss": 0.7601,
+      "step": 9705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1172138838029007,
+      "learning_rate": 3.313675312114722e-06,
+      "loss": 0.5698,
+      "step": 9706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5043397201750643,
+      "learning_rate": 3.312699589189362e-06,
+      "loss": 0.6883,
+      "step": 9707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.192714719543629,
+      "learning_rate": 3.31172393877106e-06,
+      "loss": 0.8573,
+      "step": 9708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7004441617009785,
+      "learning_rate": 3.310748360901741e-06,
+      "loss": 0.7052,
+      "step": 9709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0640299499223858,
+      "learning_rate": 3.3097728556233256e-06,
+      "loss": 0.9303,
+      "step": 9710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.366729328210049,
+      "learning_rate": 3.308797422977734e-06,
+      "loss": 0.8839,
+      "step": 9711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0029556852207184,
+      "learning_rate": 3.307822063006886e-06,
+      "loss": 0.7297,
+      "step": 9712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.315833877597001,
+      "learning_rate": 3.3068467757526927e-06,
+      "loss": 0.821,
+      "step": 9713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0434192946628453,
+      "learning_rate": 3.3058715612570623e-06,
+      "loss": 0.7057,
+      "step": 9714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8011926719112683,
+      "learning_rate": 3.3048964195619044e-06,
+      "loss": 0.7362,
+      "step": 9715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0467037316482033,
+      "learning_rate": 3.303921350709124e-06,
+      "loss": 0.7546,
+      "step": 9716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1228590629674802,
+      "learning_rate": 3.3029463547406198e-06,
+      "loss": 0.693,
+      "step": 9717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.298163472827359,
+      "learning_rate": 3.30197143169829e-06,
+      "loss": 0.8848,
+      "step": 9718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0620914162109902,
+      "learning_rate": 3.300996581624028e-06,
+      "loss": 0.671,
+      "step": 9719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9959381748282716,
+      "learning_rate": 3.30002180455973e-06,
+      "loss": 0.5588,
+      "step": 9720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8784228927082476,
+      "learning_rate": 3.299047100547278e-06,
+      "loss": 0.857,
+      "step": 9721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1350179640799962,
+      "learning_rate": 3.2980724696285626e-06,
+      "loss": 0.7876,
+      "step": 9722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4204277871965405,
+      "learning_rate": 3.2970979118454616e-06,
+      "loss": 0.9041,
+      "step": 9723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2858743211030976,
+      "learning_rate": 3.2961234272398578e-06,
+      "loss": 0.8635,
+      "step": 9724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0373576188240987,
+      "learning_rate": 3.2951490158536238e-06,
+      "loss": 0.7626,
+      "step": 9725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7319930407776156,
+      "learning_rate": 3.2941746777286344e-06,
+      "loss": 0.766,
+      "step": 9726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2400297071581066,
+      "learning_rate": 3.2932004129067592e-06,
+      "loss": 0.7566,
+      "step": 9727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1476531780482926,
+      "learning_rate": 3.2922262214298613e-06,
+      "loss": 0.6466,
+      "step": 9728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.60744180975779,
+      "learning_rate": 3.291252103339806e-06,
+      "loss": 0.679,
+      "step": 9729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.176680048475889,
+      "learning_rate": 3.2902780586784542e-06,
+      "loss": 0.8661,
+      "step": 9730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7327727842489637,
+      "learning_rate": 3.2893040874876624e-06,
+      "loss": 0.7858,
+      "step": 9731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.143369226778852,
+      "learning_rate": 3.2883301898092833e-06,
+      "loss": 0.6096,
+      "step": 9732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.985080680709704,
+      "learning_rate": 3.287356365685168e-06,
+      "loss": 0.8001,
+      "step": 9733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1358020684701864,
+      "learning_rate": 3.2863826151571654e-06,
+      "loss": 0.9548,
+      "step": 9734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.828181333048525,
+      "learning_rate": 3.285408938267116e-06,
+      "loss": 0.7957,
+      "step": 9735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5066551832297708,
+      "learning_rate": 3.284435335056865e-06,
+      "loss": 0.8469,
+      "step": 9736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.105290385309383,
+      "learning_rate": 3.283461805568246e-06,
+      "loss": 0.8245,
+      "step": 9737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9090682421610823,
+      "learning_rate": 3.2824883498430992e-06,
+      "loss": 0.7257,
+      "step": 9738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1908524739595663,
+      "learning_rate": 3.2815149679232507e-06,
+      "loss": 0.8588,
+      "step": 9739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.221645098840919,
+      "learning_rate": 3.2805416598505323e-06,
+      "loss": 0.777,
+      "step": 9740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5227325762572104,
+      "learning_rate": 3.2795684256667674e-06,
+      "loss": 0.7152,
+      "step": 9741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.086769522158854,
+      "learning_rate": 3.278595265413781e-06,
+      "loss": 0.8044,
+      "step": 9742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3775830921787238,
+      "learning_rate": 3.277622179133387e-06,
+      "loss": 0.8372,
+      "step": 9743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.741268477978389,
+      "learning_rate": 3.276649166867406e-06,
+      "loss": 0.7163,
+      "step": 9744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.37693160834685,
+      "learning_rate": 3.2756762286576482e-06,
+      "loss": 0.9992,
+      "step": 9745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.00444382576347,
+      "learning_rate": 3.2747033645459226e-06,
+      "loss": 0.941,
+      "step": 9746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.946085061389348,
+      "learning_rate": 3.2737305745740353e-06,
+      "loss": 0.6808,
+      "step": 9747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46544378584042,
+      "learning_rate": 3.2727578587837893e-06,
+      "loss": 0.7491,
+      "step": 9748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6463559903502296,
+      "learning_rate": 3.271785217216987e-06,
+      "loss": 0.6737,
+      "step": 9749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.144065214398935,
+      "learning_rate": 3.2708126499154225e-06,
+      "loss": 0.6032,
+      "step": 9750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1530913536893164,
+      "learning_rate": 3.2698401569208883e-06,
+      "loss": 0.6709,
+      "step": 9751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.996676691795696,
+      "learning_rate": 3.2688677382751754e-06,
+      "loss": 0.7722,
+      "step": 9752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9580566483646469,
+      "learning_rate": 3.2678953940200744e-06,
+      "loss": 0.8567,
+      "step": 9753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.345258222555342,
+      "learning_rate": 3.266923124197363e-06,
+      "loss": 0.8093,
+      "step": 9754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1443880199267005,
+      "learning_rate": 3.265950928848825e-06,
+      "loss": 0.7393,
+      "step": 9755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.19074278725566,
+      "learning_rate": 3.264978808016239e-06,
+      "loss": 1.0126,
+      "step": 9756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0881085987892205,
+      "learning_rate": 3.264006761741376e-06,
+      "loss": 0.7139,
+      "step": 9757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203506476217006,
+      "learning_rate": 3.2630347900660094e-06,
+      "loss": 0.7787,
+      "step": 9758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.561766447336766,
+      "learning_rate": 3.2620628930319065e-06,
+      "loss": 0.7426,
+      "step": 9759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7354745342574052,
+      "learning_rate": 3.261091070680833e-06,
+      "loss": 0.7793,
+      "step": 9760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2043521156734665,
+      "learning_rate": 3.260119323054546e-06,
+      "loss": 0.9438,
+      "step": 9761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.683541121146456,
+      "learning_rate": 3.2591476501948075e-06,
+      "loss": 0.8754,
+      "step": 9762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9209782225885867,
+      "learning_rate": 3.258176052143374e-06,
+      "loss": 0.6756,
+      "step": 9763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.917959915767974,
+      "learning_rate": 3.257204528941993e-06,
+      "loss": 0.8549,
+      "step": 9764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.131968911726866,
+      "learning_rate": 3.256233080632414e-06,
+      "loss": 0.9095,
+      "step": 9765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6836531258013947,
+      "learning_rate": 3.255261707256383e-06,
+      "loss": 0.7237,
+      "step": 9766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.276723898396048,
+      "learning_rate": 3.2542904088556437e-06,
+      "loss": 0.6743,
+      "step": 9767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9635481955081846,
+      "learning_rate": 3.2533191854719336e-06,
+      "loss": 0.9689,
+      "step": 9768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.281758490937354,
+      "learning_rate": 3.2523480371469863e-06,
+      "loss": 0.8451,
+      "step": 9769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0705937885740482,
+      "learning_rate": 3.251376963922537e-06,
+      "loss": 0.5606,
+      "step": 9770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9342981007290323,
+      "learning_rate": 3.250405965840315e-06,
+      "loss": 0.7481,
+      "step": 9771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6785913272220228,
+      "learning_rate": 3.249435042942043e-06,
+      "loss": 0.9301,
+      "step": 9772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.217997445321142,
+      "learning_rate": 3.2484641952694473e-06,
+      "loss": 0.6206,
+      "step": 9773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1372296525767416,
+      "learning_rate": 3.2474934228642475e-06,
+      "loss": 0.7139,
+      "step": 9774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.831067960555166,
+      "learning_rate": 3.246522725768156e-06,
+      "loss": 0.8933,
+      "step": 9775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.431606906576559,
+      "learning_rate": 3.2455521040228883e-06,
+      "loss": 0.6909,
+      "step": 9776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1970695911909184,
+      "learning_rate": 3.2445815576701544e-06,
+      "loss": 0.7705,
+      "step": 9777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.136585218896138,
+      "learning_rate": 3.2436110867516625e-06,
+      "loss": 0.6943,
+      "step": 9778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.203276154484336,
+      "learning_rate": 3.242640691309111e-06,
+      "loss": 0.6511,
+      "step": 9779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.99976072319589,
+      "learning_rate": 3.2416703713842036e-06,
+      "loss": 0.7935,
+      "step": 9780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.789225343041913,
+      "learning_rate": 3.240700127018638e-06,
+      "loss": 0.7306,
+      "step": 9781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2473545535578996,
+      "learning_rate": 3.239729958254104e-06,
+      "loss": 0.9145,
+      "step": 9782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1637295738315667,
+      "learning_rate": 3.238759865132295e-06,
+      "loss": 1.0414,
+      "step": 9783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.284824541071247,
+      "learning_rate": 3.2377898476948964e-06,
+      "loss": 0.767,
+      "step": 9784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9699512329256592,
+      "learning_rate": 3.2368199059835947e-06,
+      "loss": 0.7216,
+      "step": 9785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.018444799647761,
+      "learning_rate": 3.235850040040066e-06,
+      "loss": 0.7491,
+      "step": 9786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.049574319151596,
+      "learning_rate": 3.2348802499059915e-06,
+      "loss": 0.9382,
+      "step": 9787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6750810927632216,
+      "learning_rate": 3.2339105356230426e-06,
+      "loss": 0.8132,
+      "step": 9788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2722286883416816,
+      "learning_rate": 3.2329408972328934e-06,
+      "loss": 0.8638,
+      "step": 9789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1657741396209715,
+      "learning_rate": 3.2319713347772075e-06,
+      "loss": 0.8277,
+      "step": 9790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9852137764690965,
+      "learning_rate": 3.231001848297651e-06,
+      "loss": 0.8823,
+      "step": 9791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8509675288727987,
+      "learning_rate": 3.230032437835887e-06,
+      "loss": 0.9234,
+      "step": 9792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8092011102698224,
+      "learning_rate": 3.2290631034335684e-06,
+      "loss": 0.7484,
+      "step": 9793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7002795809939104,
+      "learning_rate": 3.2280938451323524e-06,
+      "loss": 1.0604,
+      "step": 9794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0387736316252822,
+      "learning_rate": 3.2271246629738917e-06,
+      "loss": 0.8792,
+      "step": 9795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1803265927424942,
+      "learning_rate": 3.2261555569998316e-06,
+      "loss": 0.7077,
+      "step": 9796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.225200397596633,
+      "learning_rate": 3.2251865272518187e-06,
+      "loss": 0.8536,
+      "step": 9797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.202992378776367,
+      "learning_rate": 3.224217573771492e-06,
+      "loss": 0.8496,
+      "step": 9798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2447039621118727,
+      "learning_rate": 3.223248696600493e-06,
+      "loss": 0.7866,
+      "step": 9799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4017495568844645,
+      "learning_rate": 3.2222798957804524e-06,
+      "loss": 0.8524,
+      "step": 9800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7522996814785372,
+      "learning_rate": 3.2213111713530056e-06,
+      "loss": 0.7681,
+      "step": 9801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.037597789162419,
+      "learning_rate": 3.2203425233597773e-06,
+      "loss": 0.6419,
+      "step": 9802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.242420682333744,
+      "learning_rate": 3.2193739518423972e-06,
+      "loss": 0.8926,
+      "step": 9803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3694707618703856,
+      "learning_rate": 3.2184054568424817e-06,
+      "loss": 0.8483,
+      "step": 9804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0515113133560137,
+      "learning_rate": 3.217437038401652e-06,
+      "loss": 0.6004,
+      "step": 9805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3141697060773696,
+      "learning_rate": 3.2164686965615223e-06,
+      "loss": 0.9263,
+      "step": 9806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.047157141397534,
+      "learning_rate": 3.215500431363706e-06,
+      "loss": 0.8423,
+      "step": 9807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.314667561849928,
+      "learning_rate": 3.2145322428498093e-06,
+      "loss": 0.8418,
+      "step": 9808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7169643991595436,
+      "learning_rate": 3.2135641310614383e-06,
+      "loss": 0.7985,
+      "step": 9809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.734382922982129,
+      "learning_rate": 3.2125960960401974e-06,
+      "loss": 0.6504,
+      "step": 9810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.218449686757701,
+      "learning_rate": 3.211628137827682e-06,
+      "loss": 0.8338,
+      "step": 9811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.41967727455423,
+      "learning_rate": 3.2106602564654877e-06,
+      "loss": 0.8789,
+      "step": 9812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.560025137029196,
+      "learning_rate": 3.2096924519952067e-06,
+      "loss": 0.8987,
+      "step": 9813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.266062495422454,
+      "learning_rate": 3.20872472445843e-06,
+      "loss": 0.713,
+      "step": 9814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.857995423766254,
+      "learning_rate": 3.207757073896741e-06,
+      "loss": 0.8046,
+      "step": 9815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9652848609961073,
+      "learning_rate": 3.2067895003517207e-06,
+      "loss": 0.7981,
+      "step": 9816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5307095553449774,
+      "learning_rate": 3.2058220038649513e-06,
+      "loss": 0.6622,
+      "step": 9817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.447147511821753,
+      "learning_rate": 3.2048545844780044e-06,
+      "loss": 0.6755,
+      "step": 9818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3837324117215704,
+      "learning_rate": 3.203887242232455e-06,
+      "loss": 0.8087,
+      "step": 9819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1060898759817754,
+      "learning_rate": 3.2029199771698693e-06,
+      "loss": 0.6147,
+      "step": 9820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0404311349746713,
+      "learning_rate": 3.2019527893318177e-06,
+      "loss": 0.8639,
+      "step": 9821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.047467884902662,
+      "learning_rate": 3.2009856787598558e-06,
+      "loss": 0.6954,
+      "step": 9822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9159641348016458,
+      "learning_rate": 3.2000186454955474e-06,
+      "loss": 0.7859,
+      "step": 9823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.534528999148765,
+      "learning_rate": 3.1990516895804467e-06,
+      "loss": 0.8242,
+      "step": 9824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2709655755970872,
+      "learning_rate": 3.198084811056107e-06,
+      "loss": 0.8681,
+      "step": 9825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.325779063513634,
+      "learning_rate": 3.197118009964074e-06,
+      "loss": 0.7449,
+      "step": 9826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.760348691052534,
+      "learning_rate": 3.1961512863458953e-06,
+      "loss": 0.7716,
+      "step": 9827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1216675045646447,
+      "learning_rate": 3.195184640243115e-06,
+      "loss": 0.7069,
+      "step": 9828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.884388379567947,
+      "learning_rate": 3.1942180716972698e-06,
+      "loss": 0.6529,
+      "step": 9829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0947719306788968,
+      "learning_rate": 3.1932515807498945e-06,
+      "loss": 0.8569,
+      "step": 9830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2379200870207154,
+      "learning_rate": 3.1922851674425226e-06,
+      "loss": 0.837,
+      "step": 9831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.070368535822446,
+      "learning_rate": 3.191318831816685e-06,
+      "loss": 0.9676,
+      "step": 9832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.082680410064498,
+      "learning_rate": 3.190352573913903e-06,
+      "loss": 0.7775,
+      "step": 9833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8993390363964353,
+      "learning_rate": 3.189386393775703e-06,
+      "loss": 0.6659,
+      "step": 9834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0917196000177403,
+      "learning_rate": 3.1884202914436024e-06,
+      "loss": 0.7957,
+      "step": 9835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0580279668248695,
+      "learning_rate": 3.1874542669591148e-06,
+      "loss": 0.6537,
+      "step": 9836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4417658001876896,
+      "learning_rate": 3.1864883203637544e-06,
+      "loss": 0.7101,
+      "step": 9837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.20142368537744,
+      "learning_rate": 3.18552245169903e-06,
+      "loss": 0.6839,
+      "step": 9838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.135185731592059,
+      "learning_rate": 3.1845566610064487e-06,
+      "loss": 0.6337,
+      "step": 9839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.778213262730036,
+      "learning_rate": 3.1835909483275083e-06,
+      "loss": 0.9061,
+      "step": 9840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2681910108161247,
+      "learning_rate": 3.1826253137037112e-06,
+      "loss": 0.6086,
+      "step": 9841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8094847745736398,
+      "learning_rate": 3.1816597571765517e-06,
+      "loss": 0.7427,
+      "step": 9842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.094536939775288,
+      "learning_rate": 3.180694278787524e-06,
+      "loss": 0.8083,
+      "step": 9843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3108961586246735,
+      "learning_rate": 3.179728878578112e-06,
+      "loss": 0.6662,
+      "step": 9844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.529765128175181,
+      "learning_rate": 3.1787635565898046e-06,
+      "loss": 0.8515,
+      "step": 9845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1166675202039786,
+      "learning_rate": 3.1777983128640855e-06,
+      "loss": 0.8697,
+      "step": 9846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.204363418348013,
+      "learning_rate": 3.1768331474424285e-06,
+      "loss": 0.8548,
+      "step": 9847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4022590256549403,
+      "learning_rate": 3.1758680603663127e-06,
+      "loss": 0.6827,
+      "step": 9848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8907293870127588,
+      "learning_rate": 3.1749030516772084e-06,
+      "loss": 0.6966,
+      "step": 9849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2631472725423376,
+      "learning_rate": 3.1739381214165864e-06,
+      "loss": 0.7981,
+      "step": 9850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.021016906956167,
+      "learning_rate": 3.1729732696259085e-06,
+      "loss": 0.8225,
+      "step": 9851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8414139448922713,
+      "learning_rate": 3.172008496346639e-06,
+      "loss": 0.7246,
+      "step": 9852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7322065460384943,
+      "learning_rate": 3.1710438016202362e-06,
+      "loss": 0.752,
+      "step": 9853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.17193643211781,
+      "learning_rate": 3.170079185488153e-06,
+      "loss": 0.7346,
+      "step": 9854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.106038947257729,
+      "learning_rate": 3.1691146479918423e-06,
+      "loss": 0.9503,
+      "step": 9855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.299556562391252,
+      "learning_rate": 3.168150189172754e-06,
+      "loss": 0.8892,
+      "step": 9856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8807949519092146,
+      "learning_rate": 3.1671858090723315e-06,
+      "loss": 0.8007,
+      "step": 9857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8683779520739145,
+      "learning_rate": 3.1662215077320174e-06,
+      "loss": 0.6957,
+      "step": 9858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8013266309217704,
+      "learning_rate": 3.165257285193248e-06,
+      "loss": 0.758,
+      "step": 9859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0532454753385294,
+      "learning_rate": 3.1642931414974586e-06,
+      "loss": 0.7452,
+      "step": 9860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.54198723756524,
+      "learning_rate": 3.1633290766860834e-06,
+      "loss": 1.0155,
+      "step": 9861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.072745092003371,
+      "learning_rate": 3.1623650908005476e-06,
+      "loss": 0.7138,
+      "step": 9862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6862932276712956,
+      "learning_rate": 3.1614011838822755e-06,
+      "loss": 0.8773,
+      "step": 9863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.091044451675503,
+      "learning_rate": 3.1604373559726915e-06,
+      "loss": 0.5934,
+      "step": 9864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359653019663565,
+      "learning_rate": 3.1594736071132086e-06,
+      "loss": 0.6552,
+      "step": 9865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15457608341658,
+      "learning_rate": 3.1585099373452454e-06,
+      "loss": 0.9424,
+      "step": 9866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.191855807203065,
+      "learning_rate": 3.15754634671021e-06,
+      "loss": 0.9139,
+      "step": 9867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.686993865151094,
+      "learning_rate": 3.1565828352495144e-06,
+      "loss": 0.7299,
+      "step": 9868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3905347989921517,
+      "learning_rate": 3.1556194030045563e-06,
+      "loss": 0.8116,
+      "step": 9869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.712179416490835,
+      "learning_rate": 3.154656050016742e-06,
+      "loss": 0.7618,
+      "step": 9870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2016537801022342,
+      "learning_rate": 3.153692776327466e-06,
+      "loss": 0.8811,
+      "step": 9871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.136408525733823,
+      "learning_rate": 3.1527295819781256e-06,
+      "loss": 0.6954,
+      "step": 9872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.06745523696451,
+      "learning_rate": 3.151766467010107e-06,
+      "loss": 0.7778,
+      "step": 9873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7769461451156543,
+      "learning_rate": 3.1508034314647994e-06,
+      "loss": 0.8519,
+      "step": 9874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.166774319522694,
+      "learning_rate": 3.149840475383589e-06,
+      "loss": 0.8708,
+      "step": 9875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9196788888132796,
+      "learning_rate": 3.1488775988078525e-06,
+      "loss": 0.834,
+      "step": 9876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2697545410995903,
+      "learning_rate": 3.1479148017789673e-06,
+      "loss": 0.7808,
+      "step": 9877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.284795462185913,
+      "learning_rate": 3.146952084338308e-06,
+      "loss": 0.6925,
+      "step": 9878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0701008728542947,
+      "learning_rate": 3.1459894465272467e-06,
+      "loss": 0.8874,
+      "step": 9879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0375037737734452,
+      "learning_rate": 3.1450268883871467e-06,
+      "loss": 0.6662,
+      "step": 9880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3851512915589494,
+      "learning_rate": 3.1440644099593716e-06,
+      "loss": 0.7324,
+      "step": 9881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.489492936083141,
+      "learning_rate": 3.143102011285284e-06,
+      "loss": 0.892,
+      "step": 9882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2627278658330274,
+      "learning_rate": 3.142139692406237e-06,
+      "loss": 0.6805,
+      "step": 9883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140615213282456,
+      "learning_rate": 3.1411774533635854e-06,
+      "loss": 0.8368,
+      "step": 9884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3758973683539852,
+      "learning_rate": 3.1402152941986786e-06,
+      "loss": 0.7873,
+      "step": 9885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.183304047836956,
+      "learning_rate": 3.1392532149528643e-06,
+      "loss": 0.9961,
+      "step": 9886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8432096689935393,
+      "learning_rate": 3.1382912156674807e-06,
+      "loss": 0.74,
+      "step": 9887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2572415677761732,
+      "learning_rate": 3.13732929638387e-06,
+      "loss": 0.729,
+      "step": 9888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2916188377887443,
+      "learning_rate": 3.136367457143369e-06,
+      "loss": 0.7598,
+      "step": 9889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.232377153000455,
+      "learning_rate": 3.1354056979873104e-06,
+      "loss": 0.6137,
+      "step": 9890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8888838172978712,
+      "learning_rate": 3.134444018957019e-06,
+      "loss": 0.7359,
+      "step": 9891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1521774424016356,
+      "learning_rate": 3.1334824200938228e-06,
+      "loss": 0.6885,
+      "step": 9892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8623811004764403,
+      "learning_rate": 3.132520901439047e-06,
+      "loss": 0.8171,
+      "step": 9893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.007875974278303,
+      "learning_rate": 3.1315594630340052e-06,
+      "loss": 0.7012,
+      "step": 9894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1155963207549786,
+      "learning_rate": 3.1305981049200134e-06,
+      "loss": 0.7898,
+      "step": 9895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3196750905582504,
+      "learning_rate": 3.129636827138385e-06,
+      "loss": 0.9244,
+      "step": 9896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4809996200330855,
+      "learning_rate": 3.128675629730429e-06,
+      "loss": 0.7963,
+      "step": 9897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2264201032908122,
+      "learning_rate": 3.1277145127374475e-06,
+      "loss": 0.7051,
+      "step": 9898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8205257764740157,
+      "learning_rate": 3.1267534762007435e-06,
+      "loss": 0.6721,
+      "step": 9899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.542968830134796,
+      "learning_rate": 3.125792520161616e-06,
+      "loss": 0.7807,
+      "step": 9900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1161318815349843,
+      "learning_rate": 3.1248316446613563e-06,
+      "loss": 0.6575,
+      "step": 9901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.624879816366736,
+      "learning_rate": 3.1238708497412566e-06,
+      "loss": 0.7699,
+      "step": 9902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46684308413079,
+      "learning_rate": 3.1229101354426063e-06,
+      "loss": 1.0413,
+      "step": 9903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.173558387482104,
+      "learning_rate": 3.1219495018066888e-06,
+      "loss": 0.6659,
+      "step": 9904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1254060566885395,
+      "learning_rate": 3.1209889488747813e-06,
+      "loss": 1.0242,
+      "step": 9905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2350417274292464,
+      "learning_rate": 3.120028476688164e-06,
+      "loss": 0.7254,
+      "step": 9906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.097240884783352,
+      "learning_rate": 3.119068085288111e-06,
+      "loss": 0.7269,
+      "step": 9907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9715111822560487,
+      "learning_rate": 3.118107774715892e-06,
+      "loss": 0.9202,
+      "step": 9908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9380249814867394,
+      "learning_rate": 3.1171475450127717e-06,
+      "loss": 0.8703,
+      "step": 9909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.082456870419681,
+      "learning_rate": 3.1161873962200146e-06,
+      "loss": 0.8364,
+      "step": 9910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9532131442051202,
+      "learning_rate": 3.1152273283788824e-06,
+      "loss": 0.8213,
+      "step": 9911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2466978206881656,
+      "learning_rate": 3.114267341530627e-06,
+      "loss": 0.8714,
+      "step": 9912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1559927828309404,
+      "learning_rate": 3.1133074357165055e-06,
+      "loss": 0.5182,
+      "step": 9913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2659802379121077,
+      "learning_rate": 3.112347610977764e-06,
+      "loss": 0.8136,
+      "step": 9914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0555474361503516,
+      "learning_rate": 3.111387867355652e-06,
+      "loss": 0.7738,
+      "step": 9915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9599594611727114,
+      "learning_rate": 3.110428204891407e-06,
+      "loss": 0.934,
+      "step": 9916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3396914811196843,
+      "learning_rate": 3.109468623626272e-06,
+      "loss": 0.8535,
+      "step": 9917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.645138347311512,
+      "learning_rate": 3.1085091236014808e-06,
+      "loss": 0.6888,
+      "step": 9918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8422143787695884,
+      "learning_rate": 3.1075497048582635e-06,
+      "loss": 0.7323,
+      "step": 9919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8695787412954388,
+      "learning_rate": 3.1065903674378507e-06,
+      "loss": 0.7584,
+      "step": 9920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.217753693140532,
+      "learning_rate": 3.1056311113814675e-06,
+      "loss": 0.7048,
+      "step": 9921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.103950867730755,
+      "learning_rate": 3.1046719367303334e-06,
+      "loss": 0.6703,
+      "step": 9922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.172217675121069,
+      "learning_rate": 3.1037128435256688e-06,
+      "loss": 0.7441,
+      "step": 9923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1013947793729857,
+      "learning_rate": 3.102753831808685e-06,
+      "loss": 0.607,
+      "step": 9924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0274481813659566,
+      "learning_rate": 3.101794901620594e-06,
+      "loss": 0.8734,
+      "step": 9925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5341823614257737,
+      "learning_rate": 3.1008360530026053e-06,
+      "loss": 0.7643,
+      "step": 9926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5248754453311313,
+      "learning_rate": 3.099877285995921e-06,
+      "loss": 0.7711,
+      "step": 9927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9451218918943722,
+      "learning_rate": 3.09891860064174e-06,
+      "loss": 0.5874,
+      "step": 9928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2148617211989374,
+      "learning_rate": 3.097959996981263e-06,
+      "loss": 0.857,
+      "step": 9929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3469438005862657,
+      "learning_rate": 3.0970014750556786e-06,
+      "loss": 0.8986,
+      "step": 9930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.642463236074611,
+      "learning_rate": 3.0960430349061797e-06,
+      "loss": 0.7857,
+      "step": 9931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3642581243484413,
+      "learning_rate": 3.0950846765739518e-06,
+      "loss": 0.7316,
+      "step": 9932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9632783738170754,
+      "learning_rate": 3.0941264001001796e-06,
+      "loss": 0.7292,
+      "step": 9933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.71859949793478,
+      "learning_rate": 3.093168205526038e-06,
+      "loss": 0.6845,
+      "step": 9934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9854175010909758,
+      "learning_rate": 3.0922100928927062e-06,
+      "loss": 0.6521,
+      "step": 9935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2017361640038615,
+      "learning_rate": 3.0912520622413573e-06,
+      "loss": 0.6865,
+      "step": 9936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3368243938818556,
+      "learning_rate": 3.090294113613157e-06,
+      "loss": 0.7547,
+      "step": 9937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9940214577976592,
+      "learning_rate": 3.0893362470492716e-06,
+      "loss": 0.883,
+      "step": 9938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2016465896995285,
+      "learning_rate": 3.0883784625908618e-06,
+      "loss": 0.8024,
+      "step": 9939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.156279004528396,
+      "learning_rate": 3.0874207602790895e-06,
+      "loss": 0.8456,
+      "step": 9940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1726216090653074,
+      "learning_rate": 3.086463140155106e-06,
+      "loss": 0.7033,
+      "step": 9941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0876092514211644,
+      "learning_rate": 3.085505602260061e-06,
+      "loss": 0.8895,
+      "step": 9942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.995945124105358,
+      "learning_rate": 3.0845481466351047e-06,
+      "loss": 0.8257,
+      "step": 9943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1449939202774817,
+      "learning_rate": 3.083590773321383e-06,
+      "loss": 0.7114,
+      "step": 9944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6566679762324705,
+      "learning_rate": 3.082633482360032e-06,
+      "loss": 1.0675,
+      "step": 9945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4865224170150215,
+      "learning_rate": 3.0816762737921896e-06,
+      "loss": 0.8684,
+      "step": 9946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9916675027754516,
+      "learning_rate": 3.0807191476589926e-06,
+      "loss": 0.7595,
+      "step": 9947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2226371763368657,
+      "learning_rate": 3.0797621040015656e-06,
+      "loss": 0.5522,
+      "step": 9948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6179617103878248,
+      "learning_rate": 3.0788051428610377e-06,
+      "loss": 0.8672,
+      "step": 9949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3233081103280018,
+      "learning_rate": 3.077848264278533e-06,
+      "loss": 0.8125,
+      "step": 9950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.006443294868599,
+      "learning_rate": 3.07689146829517e-06,
+      "loss": 0.7499,
+      "step": 9951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9893273571380685,
+      "learning_rate": 3.0759347549520614e-06,
+      "loss": 0.791,
+      "step": 9952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1412001033754557,
+      "learning_rate": 3.074978124290322e-06,
+      "loss": 0.6837,
+      "step": 9953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0969748575479183,
+      "learning_rate": 3.0740215763510617e-06,
+      "loss": 0.8171,
+      "step": 9954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0736312060246136,
+      "learning_rate": 3.0730651111753824e-06,
+      "loss": 0.6912,
+      "step": 9955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.770483055505484,
+      "learning_rate": 3.072108728804386e-06,
+      "loss": 0.6967,
+      "step": 9956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0253100193541305,
+      "learning_rate": 3.0711524292791707e-06,
+      "loss": 0.6548,
+      "step": 9957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0617853914553441,
+      "learning_rate": 3.0701962126408334e-06,
+      "loss": 0.6662,
+      "step": 9958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0654793541230854,
+      "learning_rate": 3.069240078930461e-06,
+      "loss": 0.8786,
+      "step": 9959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.236833372979482,
+      "learning_rate": 3.0682840281891433e-06,
+      "loss": 0.8649,
+      "step": 9960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9287069740930327,
+      "learning_rate": 3.0673280604579623e-06,
+      "loss": 0.725,
+      "step": 9961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9831108737670132,
+      "learning_rate": 3.0663721757780006e-06,
+      "loss": 0.801,
+      "step": 9962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4257019193361917,
+      "learning_rate": 3.0654163741903315e-06,
+      "loss": 0.8663,
+      "step": 9963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.788243126623811,
+      "learning_rate": 3.0644606557360303e-06,
+      "loss": 0.7006,
+      "step": 9964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.818469126405356,
+      "learning_rate": 3.0635050204561658e-06,
+      "loss": 0.7022,
+      "step": 9965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9310722788232515,
+      "learning_rate": 3.0625494683918012e-06,
+      "loss": 0.6773,
+      "step": 9966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1875532521068042,
+      "learning_rate": 3.0615939995840017e-06,
+      "loss": 0.6208,
+      "step": 9967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.314982581812669,
+      "learning_rate": 3.0606386140738253e-06,
+      "loss": 0.7088,
+      "step": 9968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2598380219283163,
+      "learning_rate": 3.0596833119023283e-06,
+      "loss": 0.8296,
+      "step": 9969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9079011723927786,
+      "learning_rate": 3.058728093110559e-06,
+      "loss": 0.8518,
+      "step": 9970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.376290147433472,
+      "learning_rate": 3.0577729577395663e-06,
+      "loss": 0.8555,
+      "step": 9971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1646302702980138,
+      "learning_rate": 3.056817905830397e-06,
+      "loss": 0.6309,
+      "step": 9972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.045200903617157,
+      "learning_rate": 3.055862937424089e-06,
+      "loss": 0.6762,
+      "step": 9973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3325232329385917,
+      "learning_rate": 3.054908052561681e-06,
+      "loss": 0.7774,
+      "step": 9974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2863334183674016,
+      "learning_rate": 3.053953251284205e-06,
+      "loss": 0.902,
+      "step": 9975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.422602754276735,
+      "learning_rate": 3.0529985336326933e-06,
+      "loss": 0.8382,
+      "step": 9976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0891199147940527,
+      "learning_rate": 3.0520438996481694e-06,
+      "loss": 0.759,
+      "step": 9977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0720099572540234,
+      "learning_rate": 3.0510893493716574e-06,
+      "loss": 0.6701,
+      "step": 9978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2243097500488074,
+      "learning_rate": 3.0501348828441767e-06,
+      "loss": 0.8429,
+      "step": 9979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.914428773717523,
+      "learning_rate": 3.0491805001067443e-06,
+      "loss": 0.7817,
+      "step": 9980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9663103520665672,
+      "learning_rate": 3.048226201200369e-06,
+      "loss": 0.7792,
+      "step": 9981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8678501133584433,
+      "learning_rate": 3.047271986166061e-06,
+      "loss": 0.7822,
+      "step": 9982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.268150454108207,
+      "learning_rate": 3.0463178550448246e-06,
+      "loss": 0.8441,
+      "step": 9983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9424135251484624,
+      "learning_rate": 3.0453638078776614e-06,
+      "loss": 0.7284,
+      "step": 9984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1664540131528005,
+      "learning_rate": 3.044409844705568e-06,
+      "loss": 0.8344,
+      "step": 9985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.018583713427205,
+      "learning_rate": 3.043455965569538e-06,
+      "loss": 1.0827,
+      "step": 9986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8404586603852016,
+      "learning_rate": 3.042502170510565e-06,
+      "loss": 0.7629,
+      "step": 9987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2072478445772044,
+      "learning_rate": 3.041548459569633e-06,
+      "loss": 0.8143,
+      "step": 9988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3066801296073627,
+      "learning_rate": 3.0405948327877233e-06,
+      "loss": 0.6704,
+      "step": 9989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1188598724945757,
+      "learning_rate": 3.039641290205818e-06,
+      "loss": 0.7905,
+      "step": 9990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7114823634803096,
+      "learning_rate": 3.038687831864895e-06,
+      "loss": 0.7811,
+      "step": 9991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1137412819494334,
+      "learning_rate": 3.037734457805922e-06,
+      "loss": 0.6633,
+      "step": 9992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1976696401596376,
+      "learning_rate": 3.036781168069869e-06,
+      "loss": 0.7691,
+      "step": 9993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.197548489422567,
+      "learning_rate": 3.0358279626977034e-06,
+      "loss": 0.8494,
+      "step": 9994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0317272233318924,
+      "learning_rate": 3.0348748417303826e-06,
+      "loss": 0.6452,
+      "step": 9995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.525507661262753,
+      "learning_rate": 3.033921805208867e-06,
+      "loss": 0.8292,
+      "step": 9996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.860273920777992,
+      "learning_rate": 3.0329688531741097e-06,
+      "loss": 0.9181,
+      "step": 9997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0504601831610887,
+      "learning_rate": 3.032015985667064e-06,
+      "loss": 0.933,
+      "step": 9998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1351182396526052,
+      "learning_rate": 3.0310632027286717e-06,
+      "loss": 0.6562,
+      "step": 9999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.714229604887517,
+      "learning_rate": 3.030110504399879e-06,
+      "loss": 0.8243,
+      "step": 10000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1334264238983445,
+      "learning_rate": 3.029157890721627e-06,
+      "loss": 0.8908,
+      "step": 10001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3204905308479185,
+      "learning_rate": 3.0282053617348485e-06,
+      "loss": 1.0788,
+      "step": 10002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2536776918552075,
+      "learning_rate": 3.027252917480476e-06,
+      "loss": 0.6531,
+      "step": 10003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2866886225786898,
+      "learning_rate": 3.026300557999439e-06,
+      "loss": 0.6978,
+      "step": 10004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0972671946863493,
+      "learning_rate": 3.0253482833326654e-06,
+      "loss": 0.8632,
+      "step": 10005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.055556046662613,
+      "learning_rate": 3.0243960935210726e-06,
+      "loss": 0.8968,
+      "step": 10006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.002941864306829,
+      "learning_rate": 3.023443988605578e-06,
+      "loss": 0.735,
+      "step": 10007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.295051843609817,
+      "learning_rate": 3.0224919686270982e-06,
+      "loss": 0.5855,
+      "step": 10008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5551001965056903,
+      "learning_rate": 3.021540033626544e-06,
+      "loss": 0.8969,
+      "step": 10009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1246754135444155,
+      "learning_rate": 3.0205881836448186e-06,
+      "loss": 0.5902,
+      "step": 10010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.012601150838781,
+      "learning_rate": 3.0196364187228287e-06,
+      "loss": 0.8074,
+      "step": 10011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.332926429515522,
+      "learning_rate": 3.0186847389014738e-06,
+      "loss": 0.7156,
+      "step": 10012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.349584602141382,
+      "learning_rate": 3.0177331442216455e-06,
+      "loss": 0.7519,
+      "step": 10013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375083609912801,
+      "learning_rate": 3.0167816347242396e-06,
+      "loss": 0.7154,
+      "step": 10014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1871434637722715,
+      "learning_rate": 3.0158302104501435e-06,
+      "loss": 0.9418,
+      "step": 10015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8063195898033553,
+      "learning_rate": 3.0148788714402444e-06,
+      "loss": 0.8505,
+      "step": 10016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9794958646361334,
+      "learning_rate": 3.0139276177354188e-06,
+      "loss": 0.8424,
+      "step": 10017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9387889695233418,
+      "learning_rate": 3.0129764493765466e-06,
+      "loss": 0.8191,
+      "step": 10018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.554064569566472,
+      "learning_rate": 3.012025366404504e-06,
+      "loss": 0.7408,
+      "step": 10019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9330512024468307,
+      "learning_rate": 3.0110743688601578e-06,
+      "loss": 0.8286,
+      "step": 10020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3483997256220617,
+      "learning_rate": 3.0101234567843747e-06,
+      "loss": 0.8864,
+      "step": 10021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.116833549373285,
+      "learning_rate": 3.009172630218018e-06,
+      "loss": 0.8718,
+      "step": 10022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9422953306569164,
+      "learning_rate": 3.0082218892019487e-06,
+      "loss": 0.8488,
+      "step": 10023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.213482310788114,
+      "learning_rate": 3.00727123377702e-06,
+      "loss": 0.7506,
+      "step": 10024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1187322116999705,
+      "learning_rate": 3.0063206639840846e-06,
+      "loss": 0.7692,
+      "step": 10025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.59358082982745,
+      "learning_rate": 3.00537017986399e-06,
+      "loss": 0.8376,
+      "step": 10026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7679396033286703,
+      "learning_rate": 3.0044197814575825e-06,
+      "loss": 0.8419,
+      "step": 10027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0920507267215214,
+      "learning_rate": 3.0034694688056998e-06,
+      "loss": 0.7048,
+      "step": 10028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8971871381160449,
+      "learning_rate": 3.002519241949181e-06,
+      "loss": 0.6684,
+      "step": 10029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.852032045092656,
+      "learning_rate": 3.0015691009288605e-06,
+      "loss": 0.7201,
+      "step": 10030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.270490145855429,
+      "learning_rate": 3.0006190457855643e-06,
+      "loss": 0.8921,
+      "step": 10031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9964445183127405,
+      "learning_rate": 2.99966907656012e-06,
+      "loss": 0.5927,
+      "step": 10032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3755770710029767,
+      "learning_rate": 2.9987191932933523e-06,
+      "loss": 0.7079,
+      "step": 10033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3551670248408265,
+      "learning_rate": 2.997769396026078e-06,
+      "loss": 0.8736,
+      "step": 10034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.356494591387185,
+      "learning_rate": 2.9968196847991117e-06,
+      "loss": 0.868,
+      "step": 10035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.006076362513453,
+      "learning_rate": 2.995870059653263e-06,
+      "loss": 0.7925,
+      "step": 10036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.49507939879327,
+      "learning_rate": 2.994920520629344e-06,
+      "loss": 0.8065,
+      "step": 10037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.007985542453315,
+      "learning_rate": 2.9939710677681545e-06,
+      "loss": 0.6547,
+      "step": 10038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.517407349766819,
+      "learning_rate": 2.9930217011104957e-06,
+      "loss": 0.8122,
+      "step": 10039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1130655833026968,
+      "learning_rate": 2.992072420697164e-06,
+      "loss": 0.6598,
+      "step": 10040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8325885088542995,
+      "learning_rate": 2.991123226568954e-06,
+      "loss": 0.7614,
+      "step": 10041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.197122266806058,
+      "learning_rate": 2.990174118766651e-06,
+      "loss": 1.0463,
+      "step": 10042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5061409286955474,
+      "learning_rate": 2.9892250973310433e-06,
+      "loss": 0.9068,
+      "step": 10043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9619344465645707,
+      "learning_rate": 2.98827616230291e-06,
+      "loss": 0.8165,
+      "step": 10044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9500740366769176,
+      "learning_rate": 2.987327313723033e-06,
+      "loss": 0.8631,
+      "step": 10045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1157725824951115,
+      "learning_rate": 2.9863785516321806e-06,
+      "loss": 0.8076,
+      "step": 10046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.574648500965558,
+      "learning_rate": 2.9854298760711285e-06,
+      "loss": 0.7523,
+      "step": 10047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3031343681226963,
+      "learning_rate": 2.9844812870806417e-06,
+      "loss": 0.6959,
+      "step": 10048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2890128585293628,
+      "learning_rate": 2.9835327847014816e-06,
+      "loss": 0.8577,
+      "step": 10049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5952841985149635,
+      "learning_rate": 2.982584368974408e-06,
+      "loss": 0.5813,
+      "step": 10050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.097004566283194,
+      "learning_rate": 2.9816360399401777e-06,
+      "loss": 0.795,
+      "step": 10051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0080925393052818,
+      "learning_rate": 2.980687797639543e-06,
+      "loss": 0.7472,
+      "step": 10052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5041590662748665,
+      "learning_rate": 2.979739642113251e-06,
+      "loss": 0.7308,
+      "step": 10053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3764441584927867,
+      "learning_rate": 2.9787915734020446e-06,
+      "loss": 1.0269,
+      "step": 10054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7219443011832745,
+      "learning_rate": 2.9778435915466674e-06,
+      "loss": 0.7963,
+      "step": 10055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.800869860813262,
+      "learning_rate": 2.9768956965878526e-06,
+      "loss": 0.6826,
+      "step": 10056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.256626125000081,
+      "learning_rate": 2.9759478885663362e-06,
+      "loss": 0.7427,
+      "step": 10057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0901107545768904,
+      "learning_rate": 2.975000167522846e-06,
+      "loss": 0.8165,
+      "step": 10058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.014770614176012,
+      "learning_rate": 2.9740525334981105e-06,
+      "loss": 0.9505,
+      "step": 10059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.464810781887076,
+      "learning_rate": 2.9731049865328486e-06,
+      "loss": 0.6367,
+      "step": 10060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6475705594052434,
+      "learning_rate": 2.9721575266677783e-06,
+      "loss": 0.974,
+      "step": 10061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.112364136816919,
+      "learning_rate": 2.971210153943617e-06,
+      "loss": 0.6827,
+      "step": 10062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2982423325519346,
+      "learning_rate": 2.970262868401075e-06,
+      "loss": 0.7884,
+      "step": 10063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9118377028658629,
+      "learning_rate": 2.9693156700808556e-06,
+      "loss": 0.8462,
+      "step": 10064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.082430403305921,
+      "learning_rate": 2.968368559023664e-06,
+      "loss": 0.6768,
+      "step": 10065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.368393414302988,
+      "learning_rate": 2.967421535270203e-06,
+      "loss": 0.7905,
+      "step": 10066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3682530476557613,
+      "learning_rate": 2.966474598861164e-06,
+      "loss": 0.7069,
+      "step": 10067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1021355656648715,
+      "learning_rate": 2.965527749837239e-06,
+      "loss": 0.8592,
+      "step": 10068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9746372142002886,
+      "learning_rate": 2.9645809882391187e-06,
+      "loss": 0.7795,
+      "step": 10069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.841618058715109,
+      "learning_rate": 2.9636343141074873e-06,
+      "loss": 0.8921,
+      "step": 10070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6928348469608183,
+      "learning_rate": 2.962687727483024e-06,
+      "loss": 0.6857,
+      "step": 10071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.525878149004601,
+      "learning_rate": 2.961741228406405e-06,
+      "loss": 0.7014,
+      "step": 10072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1385978470791813,
+      "learning_rate": 2.9607948169183077e-06,
+      "loss": 0.7168,
+      "step": 10073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.56358048287627,
+      "learning_rate": 2.959848493059396e-06,
+      "loss": 0.7987,
+      "step": 10074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0759753729267687,
+      "learning_rate": 2.958902256870339e-06,
+      "loss": 0.8295,
+      "step": 10075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8782103490558733,
+      "learning_rate": 2.9579561083917984e-06,
+      "loss": 0.7403,
+      "step": 10076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3531605564623295,
+      "learning_rate": 2.9570100476644325e-06,
+      "loss": 0.9776,
+      "step": 10077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7679820227178005,
+      "learning_rate": 2.9560640747288925e-06,
+      "loss": 0.7689,
+      "step": 10078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.640711936279134,
+      "learning_rate": 2.9551181896258317e-06,
+      "loss": 0.8821,
+      "step": 10079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8561994320027835,
+      "learning_rate": 2.9541723923958975e-06,
+      "loss": 0.7991,
+      "step": 10080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1345903434466935,
+      "learning_rate": 2.9532266830797336e-06,
+      "loss": 0.7035,
+      "step": 10081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.210315421624237,
+      "learning_rate": 2.9522810617179742e-06,
+      "loss": 0.8371,
+      "step": 10082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.105115992661098,
+      "learning_rate": 2.951335528351259e-06,
+      "loss": 0.8439,
+      "step": 10083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0199978608714133,
+      "learning_rate": 2.9503900830202202e-06,
+      "loss": 0.7268,
+      "step": 10084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9570434292760206,
+      "learning_rate": 2.9494447257654836e-06,
+      "loss": 0.6867,
+      "step": 10085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1059321173887224,
+      "learning_rate": 2.948499456627674e-06,
+      "loss": 0.8153,
+      "step": 10086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1654063448513075,
+      "learning_rate": 2.94755427564741e-06,
+      "loss": 0.6986,
+      "step": 10087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2074205681321428,
+      "learning_rate": 2.9466091828653127e-06,
+      "loss": 0.668,
+      "step": 10088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0632964702460668,
+      "learning_rate": 2.9456641783219897e-06,
+      "loss": 0.6274,
+      "step": 10089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9914319759326897,
+      "learning_rate": 2.9447192620580532e-06,
+      "loss": 0.8014,
+      "step": 10090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2264022585759293,
+      "learning_rate": 2.9437744341141084e-06,
+      "loss": 0.7861,
+      "step": 10091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0875874321906802,
+      "learning_rate": 2.9428296945307533e-06,
+      "loss": 0.6755,
+      "step": 10092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2512939663160316,
+      "learning_rate": 2.9418850433485877e-06,
+      "loss": 0.7964,
+      "step": 10093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1746117901200286,
+      "learning_rate": 2.9409404806082077e-06,
+      "loss": 1.0632,
+      "step": 10094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.151186869911108,
+      "learning_rate": 2.939996006350201e-06,
+      "loss": 0.8921,
+      "step": 10095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1517718020134073,
+      "learning_rate": 2.939051620615152e-06,
+      "loss": 0.9612,
+      "step": 10096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8648838358367024,
+      "learning_rate": 2.938107323443645e-06,
+      "loss": 0.5956,
+      "step": 10097
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09723786402822,
+      "learning_rate": 2.9371631148762592e-06,
+      "loss": 0.9369,
+      "step": 10098
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8914655245304652,
+      "learning_rate": 2.936218994953568e-06,
+      "loss": 0.611,
+      "step": 10099
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5313537636221333,
+      "learning_rate": 2.935274963716144e-06,
+      "loss": 0.8503,
+      "step": 10100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.739855837377948,
+      "learning_rate": 2.934331021204551e-06,
+      "loss": 0.7645,
+      "step": 10101
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.088105231783765,
+      "learning_rate": 2.9333871674593573e-06,
+      "loss": 0.8329,
+      "step": 10102
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8507223393634142,
+      "learning_rate": 2.9324434025211173e-06,
+      "loss": 0.7045,
+      "step": 10103
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.928623705876265,
+      "learning_rate": 2.93149972643039e-06,
+      "loss": 0.8859,
+      "step": 10104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8689626485544015,
+      "learning_rate": 2.930556139227726e-06,
+      "loss": 0.66,
+      "step": 10105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5297662163578982,
+      "learning_rate": 2.929612640953675e-06,
+      "loss": 0.9107,
+      "step": 10106
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8032057444368457,
+      "learning_rate": 2.9286692316487776e-06,
+      "loss": 0.6816,
+      "step": 10107
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.325377691741749,
+      "learning_rate": 2.9277259113535774e-06,
+      "loss": 0.7627,
+      "step": 10108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.14832696159006,
+      "learning_rate": 2.9267826801086103e-06,
+      "loss": 0.6962,
+      "step": 10109
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.569672330893283,
+      "learning_rate": 2.9258395379544085e-06,
+      "loss": 0.7819,
+      "step": 10110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.872295111002846,
+      "learning_rate": 2.9248964849315004e-06,
+      "loss": 0.7871,
+      "step": 10111
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.106750739538373,
+      "learning_rate": 2.9239535210804115e-06,
+      "loss": 0.7436,
+      "step": 10112
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7910260374489186,
+      "learning_rate": 2.923010646441666e-06,
+      "loss": 0.7474,
+      "step": 10113
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9467108798551345,
+      "learning_rate": 2.9220678610557773e-06,
+      "loss": 0.851,
+      "step": 10114
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.93637809400708,
+      "learning_rate": 2.9211251649632587e-06,
+      "loss": 0.8679,
+      "step": 10115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.496171284905132,
+      "learning_rate": 2.920182558204623e-06,
+      "loss": 0.8204,
+      "step": 10116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.74608073116204,
+      "learning_rate": 2.919240040820376e-06,
+      "loss": 0.6974,
+      "step": 10117
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.017103731788993,
+      "learning_rate": 2.9182976128510177e-06,
+      "loss": 0.7501,
+      "step": 10118
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0601697926127214,
+      "learning_rate": 2.9173552743370454e-06,
+      "loss": 0.6605,
+      "step": 10119
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.974499168121157,
+      "learning_rate": 2.9164130253189594e-06,
+      "loss": 0.8556,
+      "step": 10120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0700185193631304,
+      "learning_rate": 2.9154708658372433e-06,
+      "loss": 0.7068,
+      "step": 10121
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.298985508788198,
+      "learning_rate": 2.9145287959323852e-06,
+      "loss": 0.9738,
+      "step": 10122
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1584098626629191,
+      "learning_rate": 2.913586815644871e-06,
+      "loss": 0.6186,
+      "step": 10123
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1660786011421385,
+      "learning_rate": 2.912644925015179e-06,
+      "loss": 0.8601,
+      "step": 10124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8824862644018914,
+      "learning_rate": 2.9117031240837825e-06,
+      "loss": 1.0323,
+      "step": 10125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.976117841008764,
+      "learning_rate": 2.910761412891153e-06,
+      "loss": 0.7739,
+      "step": 10126
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0695822086907523,
+      "learning_rate": 2.9098197914777575e-06,
+      "loss": 0.8567,
+      "step": 10127
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8219328001127257,
+      "learning_rate": 2.9088782598840646e-06,
+      "loss": 0.8487,
+      "step": 10128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.95733382271522,
+      "learning_rate": 2.9079368181505263e-06,
+      "loss": 0.5934,
+      "step": 10129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8424756756307126,
+      "learning_rate": 2.9069954663176036e-06,
+      "loss": 0.7226,
+      "step": 10130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9747898409132847,
+      "learning_rate": 2.9060542044257487e-06,
+      "loss": 0.7122,
+      "step": 10131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0835706900067312,
+      "learning_rate": 2.9051130325154053e-06,
+      "loss": 0.797,
+      "step": 10132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9698366872037871,
+      "learning_rate": 2.904171950627022e-06,
+      "loss": 0.8403,
+      "step": 10133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6967715142344035,
+      "learning_rate": 2.9032309588010372e-06,
+      "loss": 0.718,
+      "step": 10134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2031771737903627,
+      "learning_rate": 2.9022900570778876e-06,
+      "loss": 0.6335,
+      "step": 10135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.911184292629522,
+      "learning_rate": 2.9013492454980074e-06,
+      "loss": 0.824,
+      "step": 10136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.438390908580349,
+      "learning_rate": 2.9004085241018208e-06,
+      "loss": 0.7235,
+      "step": 10137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.881039116314047,
+      "learning_rate": 2.899467892929761e-06,
+      "loss": 0.8086,
+      "step": 10138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4823000352027162,
+      "learning_rate": 2.8985273520222414e-06,
+      "loss": 0.8393,
+      "step": 10139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3250031077520985,
+      "learning_rate": 2.8975869014196797e-06,
+      "loss": 0.8128,
+      "step": 10140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2447540820533312,
+      "learning_rate": 2.896646541162493e-06,
+      "loss": 0.6939,
+      "step": 10141
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5061691970370497,
+      "learning_rate": 2.895706271291089e-06,
+      "loss": 0.9174,
+      "step": 10142
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.405548496004923,
+      "learning_rate": 2.894766091845873e-06,
+      "loss": 0.8685,
+      "step": 10143
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9637435216611823,
+      "learning_rate": 2.893826002867247e-06,
+      "loss": 0.834,
+      "step": 10144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2745295914684154,
+      "learning_rate": 2.8928860043956062e-06,
+      "loss": 0.8716,
+      "step": 10145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.38407105970583,
+      "learning_rate": 2.8919460964713507e-06,
+      "loss": 0.761,
+      "step": 10146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.169424503424921,
+      "learning_rate": 2.891006279134862e-06,
+      "loss": 0.7488,
+      "step": 10147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9827258847929268,
+      "learning_rate": 2.8900665524265326e-06,
+      "loss": 0.81,
+      "step": 10148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0314458857168223,
+      "learning_rate": 2.889126916386744e-06,
+      "loss": 0.7855,
+      "step": 10149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.212951030721876,
+      "learning_rate": 2.88818737105587e-06,
+      "loss": 0.8124,
+      "step": 10150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1609347452271817,
+      "learning_rate": 2.8872479164742896e-06,
+      "loss": 0.7178,
+      "step": 10151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2250218242843267,
+      "learning_rate": 2.886308552682371e-06,
+      "loss": 0.7471,
+      "step": 10152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.797150987544391,
+      "learning_rate": 2.885369279720482e-06,
+      "loss": 0.816,
+      "step": 10153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.770946254302735,
+      "learning_rate": 2.884430097628984e-06,
+      "loss": 0.7801,
+      "step": 10154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.488974942064696,
+      "learning_rate": 2.883491006448235e-06,
+      "loss": 0.6944,
+      "step": 10155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9773030316944238,
+      "learning_rate": 2.8825520062185945e-06,
+      "loss": 0.9484,
+      "step": 10156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3878014801806726,
+      "learning_rate": 2.881613096980407e-06,
+      "loss": 0.8435,
+      "step": 10157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4293826316500247,
+      "learning_rate": 2.8806742787740237e-06,
+      "loss": 0.8505,
+      "step": 10158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1037654261875165,
+      "learning_rate": 2.879735551639787e-06,
+      "loss": 0.8525,
+      "step": 10159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.744427740197232,
+      "learning_rate": 2.8787969156180353e-06,
+      "loss": 0.9413,
+      "step": 10160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2056576463898723,
+      "learning_rate": 2.877858370749105e-06,
+      "loss": 0.8098,
+      "step": 10161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3271358836623093,
+      "learning_rate": 2.8769199170733265e-06,
+      "loss": 0.852,
+      "step": 10162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.22407650637073,
+      "learning_rate": 2.8759815546310254e-06,
+      "loss": 0.7693,
+      "step": 10163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8668672927395555,
+      "learning_rate": 2.8750432834625312e-06,
+      "loss": 0.6524,
+      "step": 10164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.689429562310233,
+      "learning_rate": 2.874105103608156e-06,
+      "loss": 0.7742,
+      "step": 10165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5488355625695505,
+      "learning_rate": 2.8731670151082213e-06,
+      "loss": 0.802,
+      "step": 10166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.156699082347053,
+      "learning_rate": 2.872229018003036e-06,
+      "loss": 0.8484,
+      "step": 10167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3348030195115979,
+      "learning_rate": 2.871291112332909e-06,
+      "loss": 0.6526,
+      "step": 10168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5365888353413975,
+      "learning_rate": 2.8703532981381437e-06,
+      "loss": 0.8393,
+      "step": 10169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4877411451688394,
+      "learning_rate": 2.869415575459038e-06,
+      "loss": 0.9487,
+      "step": 10170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7426463566787167,
+      "learning_rate": 2.8684779443358945e-06,
+      "loss": 0.6378,
+      "step": 10171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.997422532335954,
+      "learning_rate": 2.8675404048089987e-06,
+      "loss": 0.8941,
+      "step": 10172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0633777101739903,
+      "learning_rate": 2.8666029569186393e-06,
+      "loss": 0.7159,
+      "step": 10173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3846617787903406,
+      "learning_rate": 2.8656656007051055e-06,
+      "loss": 0.8119,
+      "step": 10174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.02296165337143,
+      "learning_rate": 2.864728336208671e-06,
+      "loss": 0.6347,
+      "step": 10175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1569548606478093,
+      "learning_rate": 2.8637911634696165e-06,
+      "loss": 0.7291,
+      "step": 10176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.919937004606498,
+      "learning_rate": 2.862854082528214e-06,
+      "loss": 0.8713,
+      "step": 10177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.436548928745199,
+      "learning_rate": 2.861917093424731e-06,
+      "loss": 0.9931,
+      "step": 10178
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.998501426663559,
+      "learning_rate": 2.860980196199432e-06,
+      "loss": 0.755,
+      "step": 10179
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.138158950743284,
+      "learning_rate": 2.8600433908925764e-06,
+      "loss": 0.886,
+      "step": 10180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6499346862310316,
+      "learning_rate": 2.859106677544424e-06,
+      "loss": 0.7988,
+      "step": 10181
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3730125997318696,
+      "learning_rate": 2.8581700561952274e-06,
+      "loss": 0.8427,
+      "step": 10182
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0336236735751063,
+      "learning_rate": 2.8572335268852297e-06,
+      "loss": 0.8524,
+      "step": 10183
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5974701772696895,
+      "learning_rate": 2.8562970896546815e-06,
+      "loss": 0.7506,
+      "step": 10184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.052685867527936,
+      "learning_rate": 2.855360744543822e-06,
+      "loss": 0.6974,
+      "step": 10185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0576179704011213,
+      "learning_rate": 2.8544244915928876e-06,
+      "loss": 0.7481,
+      "step": 10186
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4080800155018345,
+      "learning_rate": 2.8534883308421114e-06,
+      "loss": 0.7906,
+      "step": 10187
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0847646576772878,
+      "learning_rate": 2.852552262331719e-06,
+      "loss": 0.6165,
+      "step": 10188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9554318393727697,
+      "learning_rate": 2.8516162861019437e-06,
+      "loss": 0.7913,
+      "step": 10189
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5207012548391345,
+      "learning_rate": 2.850680402192999e-06,
+      "loss": 0.9623,
+      "step": 10190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2078132645338178,
+      "learning_rate": 2.8497446106451022e-06,
+      "loss": 0.8616,
+      "step": 10191
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4602794325773223,
+      "learning_rate": 2.8488089114984725e-06,
+      "loss": 0.8684,
+      "step": 10192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.974987215144586,
+      "learning_rate": 2.8478733047933103e-06,
+      "loss": 0.692,
+      "step": 10193
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.943320860284785,
+      "learning_rate": 2.846937790569828e-06,
+      "loss": 0.7594,
+      "step": 10194
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3648546849617533,
+      "learning_rate": 2.8460023688682237e-06,
+      "loss": 0.9278,
+      "step": 10195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.422018967352534,
+      "learning_rate": 2.8450670397286945e-06,
+      "loss": 0.6819,
+      "step": 10196
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.749948751036242,
+      "learning_rate": 2.8441318031914344e-06,
+      "loss": 0.7642,
+      "step": 10197
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.23228672468918,
+      "learning_rate": 2.8431966592966297e-06,
+      "loss": 0.8695,
+      "step": 10198
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9606810032594617,
+      "learning_rate": 2.84226160808447e-06,
+      "loss": 0.854,
+      "step": 10199
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8846099488963315,
+      "learning_rate": 2.8413266495951366e-06,
+      "loss": 0.6923,
+      "step": 10200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1608930960903976,
+      "learning_rate": 2.8403917838688007e-06,
+      "loss": 0.8573,
+      "step": 10201
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1045437658875303,
+      "learning_rate": 2.8394570109456413e-06,
+      "loss": 0.9863,
+      "step": 10202
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9796757444499127,
+      "learning_rate": 2.838522330865826e-06,
+      "loss": 0.8023,
+      "step": 10203
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0062443157045868,
+      "learning_rate": 2.837587743669521e-06,
+      "loss": 0.5827,
+      "step": 10204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.063905775931427,
+      "learning_rate": 2.8366532493968858e-06,
+      "loss": 0.8106,
+      "step": 10205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0554747829939686,
+      "learning_rate": 2.835718848088076e-06,
+      "loss": 0.8288,
+      "step": 10206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.144140338610433,
+      "learning_rate": 2.8347845397832517e-06,
+      "loss": 0.5904,
+      "step": 10207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.073294660119589,
+      "learning_rate": 2.8338503245225544e-06,
+      "loss": 0.7577,
+      "step": 10208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8469161410550725,
+      "learning_rate": 2.8329162023461355e-06,
+      "loss": 0.8191,
+      "step": 10209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.012203468978386,
+      "learning_rate": 2.8319821732941345e-06,
+      "loss": 0.7207,
+      "step": 10210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421965981008037,
+      "learning_rate": 2.8310482374066843e-06,
+      "loss": 1.0483,
+      "step": 10211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3938862125660716,
+      "learning_rate": 2.830114394723924e-06,
+      "loss": 0.7459,
+      "step": 10212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2009890270234487,
+      "learning_rate": 2.8291806452859803e-06,
+      "loss": 0.8779,
+      "step": 10213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.478145192027823,
+      "learning_rate": 2.82824698913298e-06,
+      "loss": 0.727,
+      "step": 10214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.126357608212503,
+      "learning_rate": 2.827313426305042e-06,
+      "loss": 0.84,
+      "step": 10215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1141684288568985,
+      "learning_rate": 2.826379956842284e-06,
+      "loss": 0.7331,
+      "step": 10216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.818416579902949,
+      "learning_rate": 2.825446580784822e-06,
+      "loss": 0.7861,
+      "step": 10217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9833476614541872,
+      "learning_rate": 2.824513298172763e-06,
+      "loss": 0.7733,
+      "step": 10218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5381920785686436,
+      "learning_rate": 2.823580109046212e-06,
+      "loss": 0.9308,
+      "step": 10219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3412822769031516,
+      "learning_rate": 2.822647013445272e-06,
+      "loss": 0.879,
+      "step": 10220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.381308161590982,
+      "learning_rate": 2.821714011410039e-06,
+      "loss": 0.8012,
+      "step": 10221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.169413402709129,
+      "learning_rate": 2.820781102980606e-06,
+      "loss": 0.872,
+      "step": 10222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.090510929031838,
+      "learning_rate": 2.8198482881970626e-06,
+      "loss": 0.6448,
+      "step": 10223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3656768024857193,
+      "learning_rate": 2.8189155670994913e-06,
+      "loss": 0.6171,
+      "step": 10224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7045706095288429,
+      "learning_rate": 2.8179829397279795e-06,
+      "loss": 0.6551,
+      "step": 10225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3579424712110995,
+      "learning_rate": 2.8170504061225966e-06,
+      "loss": 1.0418,
+      "step": 10226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0526369576739139,
+      "learning_rate": 2.8161179663234215e-06,
+      "loss": 0.7602,
+      "step": 10227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 35.98661157803628,
+      "learning_rate": 2.8151856203705215e-06,
+      "loss": 0.8299,
+      "step": 10228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9980107568864542,
+      "learning_rate": 2.814253368303961e-06,
+      "loss": 0.7272,
+      "step": 10229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0913512152009215,
+      "learning_rate": 2.813321210163801e-06,
+      "loss": 0.8529,
+      "step": 10230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.225096200151031,
+      "learning_rate": 2.8123891459900967e-06,
+      "loss": 0.7062,
+      "step": 10231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0321206172857784,
+      "learning_rate": 2.8114571758229064e-06,
+      "loss": 0.8588,
+      "step": 10232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.83752357396922,
+      "learning_rate": 2.8105252997022737e-06,
+      "loss": 0.6996,
+      "step": 10233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.741316743623248,
+      "learning_rate": 2.809593517668243e-06,
+      "loss": 0.6958,
+      "step": 10234
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.366547173693852,
+      "learning_rate": 2.80866182976086e-06,
+      "loss": 0.8084,
+      "step": 10235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6585650235013123,
+      "learning_rate": 2.8077302360201575e-06,
+      "loss": 0.973,
+      "step": 10236
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.663840672637759,
+      "learning_rate": 2.80679873648617e-06,
+      "loss": 0.7289,
+      "step": 10237
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3649657024661288,
+      "learning_rate": 2.8058673311989247e-06,
+      "loss": 0.6688,
+      "step": 10238
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2793494040654947,
+      "learning_rate": 2.804936020198447e-06,
+      "loss": 0.6905,
+      "step": 10239
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6705881057194456,
+      "learning_rate": 2.8040048035247576e-06,
+      "loss": 0.9433,
+      "step": 10240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.514597879693151,
+      "learning_rate": 2.8030736812178717e-06,
+      "loss": 0.9213,
+      "step": 10241
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5246621290012157,
+      "learning_rate": 2.8021426533178007e-06,
+      "loss": 0.7144,
+      "step": 10242
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2064816929641102,
+      "learning_rate": 2.8012117198645596e-06,
+      "loss": 0.7158,
+      "step": 10243
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7553321539057727,
+      "learning_rate": 2.800280880898143e-06,
+      "loss": 0.8667,
+      "step": 10244
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.149651824983142,
+      "learning_rate": 2.7993501364585584e-06,
+      "loss": 0.941,
+      "step": 10245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.879573929192793,
+      "learning_rate": 2.7984194865857994e-06,
+      "loss": 0.7168,
+      "step": 10246
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.028184106752837,
+      "learning_rate": 2.7974889313198584e-06,
+      "loss": 0.844,
+      "step": 10247
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0169293585396955,
+      "learning_rate": 2.796558470700723e-06,
+      "loss": 0.7292,
+      "step": 10248
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7269047800680695,
+      "learning_rate": 2.795628104768376e-06,
+      "loss": 0.8916,
+      "step": 10249
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.655604886653266,
+      "learning_rate": 2.7946978335628023e-06,
+      "loss": 0.8132,
+      "step": 10250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.635934021061421,
+      "learning_rate": 2.793767657123972e-06,
+      "loss": 1.0791,
+      "step": 10251
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7573288526944657,
+      "learning_rate": 2.792837575491857e-06,
+      "loss": 0.6616,
+      "step": 10252
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3752863219151457,
+      "learning_rate": 2.7919075887064296e-06,
+      "loss": 0.9245,
+      "step": 10253
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.638900349674776,
+      "learning_rate": 2.79097769680765e-06,
+      "loss": 1.0168,
+      "step": 10254
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8278874192685914,
+      "learning_rate": 2.790047899835479e-06,
+      "loss": 0.6756,
+      "step": 10255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.963811775960579,
+      "learning_rate": 2.789118197829872e-06,
+      "loss": 0.7732,
+      "step": 10256
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.608231170524332,
+      "learning_rate": 2.7881885908307794e-06,
+      "loss": 0.6173,
+      "step": 10257
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2414738525746905,
+      "learning_rate": 2.7872590788781483e-06,
+      "loss": 0.7859,
+      "step": 10258
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.165813904396864,
+      "learning_rate": 2.7863296620119217e-06,
+      "loss": 0.9271,
+      "step": 10259
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.975910763383893,
+      "learning_rate": 2.7854003402720416e-06,
+      "loss": 0.9128,
+      "step": 10260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.210065499773632,
+      "learning_rate": 2.7844711136984425e-06,
+      "loss": 0.7029,
+      "step": 10261
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.066585095937297,
+      "learning_rate": 2.7835419823310507e-06,
+      "loss": 0.7763,
+      "step": 10262
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.618259361902893,
+      "learning_rate": 2.7826129462097974e-06,
+      "loss": 0.7144,
+      "step": 10263
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.142557918449433,
+      "learning_rate": 2.781684005374604e-06,
+      "loss": 0.803,
+      "step": 10264
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5952612671157187,
+      "learning_rate": 2.7807551598653905e-06,
+      "loss": 0.9055,
+      "step": 10265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.154719837407582,
+      "learning_rate": 2.7798264097220694e-06,
+      "loss": 0.8796,
+      "step": 10266
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9932403046351297,
+      "learning_rate": 2.7788977549845503e-06,
+      "loss": 0.7343,
+      "step": 10267
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8277912127972855,
+      "learning_rate": 2.777969195692745e-06,
+      "loss": 0.8086,
+      "step": 10268
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.008791038780106,
+      "learning_rate": 2.777040731886549e-06,
+      "loss": 0.799,
+      "step": 10269
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9955311449812672,
+      "learning_rate": 2.7761123636058645e-06,
+      "loss": 0.7427,
+      "step": 10270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.636267688039592,
+      "learning_rate": 2.775184090890585e-06,
+      "loss": 0.9975,
+      "step": 10271
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8467490689777166,
+      "learning_rate": 2.7742559137806e-06,
+      "loss": 0.6881,
+      "step": 10272
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7166795369863521,
+      "learning_rate": 2.773327832315796e-06,
+      "loss": 0.6742,
+      "step": 10273
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4904737660353478,
+      "learning_rate": 2.7723998465360537e-06,
+      "loss": 0.7186,
+      "step": 10274
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.500682182820254,
+      "learning_rate": 2.771471956481251e-06,
+      "loss": 0.8473,
+      "step": 10275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9892007741135027,
+      "learning_rate": 2.770544162191261e-06,
+      "loss": 0.8751,
+      "step": 10276
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.760591158320006,
+      "learning_rate": 2.7696164637059527e-06,
+      "loss": 0.8659,
+      "step": 10277
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.726526287760526,
+      "learning_rate": 2.7686888610651936e-06,
+      "loss": 0.9601,
+      "step": 10278
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.364369580718908,
+      "learning_rate": 2.7677613543088432e-06,
+      "loss": 0.7305,
+      "step": 10279
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.247374944517337,
+      "learning_rate": 2.76683394347676e-06,
+      "loss": 0.6423,
+      "step": 10280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2314067472071857,
+      "learning_rate": 2.7659066286087947e-06,
+      "loss": 0.7578,
+      "step": 10281
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.728666029190709,
+      "learning_rate": 2.7649794097447956e-06,
+      "loss": 0.8387,
+      "step": 10282
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5687523666657675,
+      "learning_rate": 2.7640522869246134e-06,
+      "loss": 0.758,
+      "step": 10283
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1185466602717904,
+      "learning_rate": 2.7631252601880816e-06,
+      "loss": 0.7473,
+      "step": 10284
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2245187318995734,
+      "learning_rate": 2.762198329575037e-06,
+      "loss": 0.6577,
+      "step": 10285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28411829495199,
+      "learning_rate": 2.7612714951253183e-06,
+      "loss": 0.6846,
+      "step": 10286
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3882735892965954,
+      "learning_rate": 2.7603447568787456e-06,
+      "loss": 0.7212,
+      "step": 10287
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0813801914756012,
+      "learning_rate": 2.759418114875148e-06,
+      "loss": 0.6561,
+      "step": 10288
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.973098219932926,
+      "learning_rate": 2.7584915691543444e-06,
+      "loss": 0.7134,
+      "step": 10289
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.449362440051057,
+      "learning_rate": 2.7575651197561504e-06,
+      "loss": 0.7674,
+      "step": 10290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.131182322660912,
+      "learning_rate": 2.7566387667203776e-06,
+      "loss": 0.7743,
+      "step": 10291
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0593884432235505,
+      "learning_rate": 2.7557125100868327e-06,
+      "loss": 0.5727,
+      "step": 10292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2191524601122645,
+      "learning_rate": 2.75478634989532e-06,
+      "loss": 0.62,
+      "step": 10293
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.606979870295436,
+      "learning_rate": 2.753860286185637e-06,
+      "loss": 0.8689,
+      "step": 10294
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.594189763554873,
+      "learning_rate": 2.7529343189975788e-06,
+      "loss": 0.8915,
+      "step": 10295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.936412587251289,
+      "learning_rate": 2.7520084483709397e-06,
+      "loss": 0.9021,
+      "step": 10296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.951097642197017,
+      "learning_rate": 2.7510826743455037e-06,
+      "loss": 0.8841,
+      "step": 10297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.829462506118236,
+      "learning_rate": 2.7501569969610544e-06,
+      "loss": 0.6747,
+      "step": 10298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.042528087634352,
+      "learning_rate": 2.7492314162573687e-06,
+      "loss": 0.7847,
+      "step": 10299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9642052952031718,
+      "learning_rate": 2.74830593227422e-06,
+      "loss": 0.6295,
+      "step": 10300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.877549593446492,
+      "learning_rate": 2.7473805450513846e-06,
+      "loss": 0.7358,
+      "step": 10301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.63967140223689,
+      "learning_rate": 2.7464552546286217e-06,
+      "loss": 0.7185,
+      "step": 10302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.267395670442103,
+      "learning_rate": 2.7455300610456935e-06,
+      "loss": 0.7714,
+      "step": 10303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6775708063567276,
+      "learning_rate": 2.744604964342364e-06,
+      "loss": 0.7295,
+      "step": 10304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.168585975306279,
+      "learning_rate": 2.743679964558378e-06,
+      "loss": 0.7873,
+      "step": 10305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6920188586673923,
+      "learning_rate": 2.7427550617334905e-06,
+      "loss": 0.7988,
+      "step": 10306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296526441349824,
+      "learning_rate": 2.741830255907445e-06,
+      "loss": 0.8408,
+      "step": 10307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9074112608761238,
+      "learning_rate": 2.740905547119983e-06,
+      "loss": 0.7981,
+      "step": 10308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9112148270751252,
+      "learning_rate": 2.7399809354108415e-06,
+      "loss": 0.7583,
+      "step": 10309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0972837106498425,
+      "learning_rate": 2.73905642081975e-06,
+      "loss": 0.642,
+      "step": 10310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.048070943111501,
+      "learning_rate": 2.7381320033864434e-06,
+      "loss": 0.7447,
+      "step": 10311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8020317360811804,
+      "learning_rate": 2.73720768315064e-06,
+      "loss": 0.7609,
+      "step": 10312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7550960894729526,
+      "learning_rate": 2.7362834601520603e-06,
+      "loss": 0.7758,
+      "step": 10313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.416812839531527,
+      "learning_rate": 2.735359334430424e-06,
+      "loss": 0.9062,
+      "step": 10314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1790999603410697,
+      "learning_rate": 2.73443530602544e-06,
+      "loss": 0.727,
+      "step": 10315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.092520571173503,
+      "learning_rate": 2.733511374976817e-06,
+      "loss": 0.7846,
+      "step": 10316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7065186286393639,
+      "learning_rate": 2.7325875413242585e-06,
+      "loss": 0.6382,
+      "step": 10317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1548664569632403,
+      "learning_rate": 2.7316638051074605e-06,
+      "loss": 0.702,
+      "step": 10318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7328092668446917,
+      "learning_rate": 2.7307401663661247e-06,
+      "loss": 0.6834,
+      "step": 10319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1426714489690106,
+      "learning_rate": 2.729816625139934e-06,
+      "loss": 0.7548,
+      "step": 10320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0194820286852724,
+      "learning_rate": 2.728893181468581e-06,
+      "loss": 0.6352,
+      "step": 10321
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0670761133924063,
+      "learning_rate": 2.7279698353917476e-06,
+      "loss": 0.7756,
+      "step": 10322
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.243985233455772,
+      "learning_rate": 2.727046586949107e-06,
+      "loss": 0.837,
+      "step": 10323
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.99986604925008,
+      "learning_rate": 2.7261234361803383e-06,
+      "loss": 0.907,
+      "step": 10324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0008991763960005,
+      "learning_rate": 2.72520038312511e-06,
+      "loss": 0.8825,
+      "step": 10325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0562591987432999,
+      "learning_rate": 2.7242774278230876e-06,
+      "loss": 0.7396,
+      "step": 10326
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5583748457655298,
+      "learning_rate": 2.723354570313932e-06,
+      "loss": 0.9456,
+      "step": 10327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9803893778730408,
+      "learning_rate": 2.7224318106372995e-06,
+      "loss": 0.9012,
+      "step": 10328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1639269577321152,
+      "learning_rate": 2.721509148832847e-06,
+      "loss": 0.6454,
+      "step": 10329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6382872610860675,
+      "learning_rate": 2.7205865849402187e-06,
+      "loss": 0.9972,
+      "step": 10330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9136466333715423,
+      "learning_rate": 2.7196641189990623e-06,
+      "loss": 0.692,
+      "step": 10331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.225672590463188,
+      "learning_rate": 2.7187417510490176e-06,
+      "loss": 0.6721,
+      "step": 10332
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9304614727564013,
+      "learning_rate": 2.7178194811297204e-06,
+      "loss": 0.8218,
+      "step": 10333
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2367148477364687,
+      "learning_rate": 2.7168973092808025e-06,
+      "loss": 0.9631,
+      "step": 10334
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9789558155133609,
+      "learning_rate": 2.7159752355418927e-06,
+      "loss": 0.716,
+      "step": 10335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0030133118179405,
+      "learning_rate": 2.7150532599526114e-06,
+      "loss": 0.9743,
+      "step": 10336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2724809939872674,
+      "learning_rate": 2.7141313825525845e-06,
+      "loss": 0.7072,
+      "step": 10337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1639358260228314,
+      "learning_rate": 2.7132096033814193e-06,
+      "loss": 0.7997,
+      "step": 10338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.138976597432661,
+      "learning_rate": 2.7122879224787315e-06,
+      "loss": 0.8225,
+      "step": 10339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.17341229539948,
+      "learning_rate": 2.7113663398841296e-06,
+      "loss": 0.824,
+      "step": 10340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.073222102372347,
+      "learning_rate": 2.7104448556372086e-06,
+      "loss": 0.8092,
+      "step": 10341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.954219676629857,
+      "learning_rate": 2.709523469777573e-06,
+      "loss": 0.8813,
+      "step": 10342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8238722038417872,
+      "learning_rate": 2.7086021823448155e-06,
+      "loss": 0.7183,
+      "step": 10343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1987646522207585,
+      "learning_rate": 2.7076809933785254e-06,
+      "loss": 0.7442,
+      "step": 10344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2172404715259615,
+      "learning_rate": 2.7067599029182884e-06,
+      "loss": 0.6656,
+      "step": 10345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2078171417521775,
+      "learning_rate": 2.7058389110036835e-06,
+      "loss": 0.9481,
+      "step": 10346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8437465408294489,
+      "learning_rate": 2.704918017674294e-06,
+      "loss": 0.7611,
+      "step": 10347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0617232230083113,
+      "learning_rate": 2.7039972229696847e-06,
+      "loss": 0.6324,
+      "step": 10348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.377699056057053,
+      "learning_rate": 2.70307652692943e-06,
+      "loss": 0.6983,
+      "step": 10349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7807857289553286,
+      "learning_rate": 2.702155929593093e-06,
+      "loss": 0.7758,
+      "step": 10350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.119826826631277,
+      "learning_rate": 2.7012354310002333e-06,
+      "loss": 0.8696,
+      "step": 10351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1723817758824766,
+      "learning_rate": 2.700315031190406e-06,
+      "loss": 0.6686,
+      "step": 10352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.140304379215633,
+      "learning_rate": 2.6993947302031643e-06,
+      "loss": 0.739,
+      "step": 10353
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0993957813183703,
+      "learning_rate": 2.6984745280780524e-06,
+      "loss": 0.6482,
+      "step": 10354
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2078142937788643,
+      "learning_rate": 2.6975544248546195e-06,
+      "loss": 0.6371,
+      "step": 10355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9581891664046946,
+      "learning_rate": 2.696634420572397e-06,
+      "loss": 0.7681,
+      "step": 10356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.101398402655754,
+      "learning_rate": 2.6957145152709253e-06,
+      "loss": 0.987,
+      "step": 10357
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.05891027677017,
+      "learning_rate": 2.694794708989733e-06,
+      "loss": 0.7344,
+      "step": 10358
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3358805409461185,
+      "learning_rate": 2.6938750017683457e-06,
+      "loss": 0.81,
+      "step": 10359
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9911892208505972,
+      "learning_rate": 2.692955393646286e-06,
+      "loss": 0.7017,
+      "step": 10360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6024007978533223,
+      "learning_rate": 2.6920358846630688e-06,
+      "loss": 0.6365,
+      "step": 10361
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2797250930188637,
+      "learning_rate": 2.691116474858214e-06,
+      "loss": 0.7672,
+      "step": 10362
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0009171335256877,
+      "learning_rate": 2.690197164271224e-06,
+      "loss": 0.7213,
+      "step": 10363
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1741140758979403,
+      "learning_rate": 2.6892779529416045e-06,
+      "loss": 0.76,
+      "step": 10364
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.035054239768266,
+      "learning_rate": 2.6883588409088597e-06,
+      "loss": 0.8813,
+      "step": 10365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9129185833130617,
+      "learning_rate": 2.6874398282124837e-06,
+      "loss": 0.7949,
+      "step": 10366
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2254391430341522,
+      "learning_rate": 2.686520914891968e-06,
+      "loss": 1.0328,
+      "step": 10367
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.87364946269296,
+      "learning_rate": 2.6856021009868016e-06,
+      "loss": 0.8538,
+      "step": 10368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28486337401729,
+      "learning_rate": 2.6846833865364674e-06,
+      "loss": 0.7017,
+      "step": 10369
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.996957593514187,
+      "learning_rate": 2.6837647715804444e-06,
+      "loss": 0.6299,
+      "step": 10370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4780602587073086,
+      "learning_rate": 2.682846256158206e-06,
+      "loss": 0.8903,
+      "step": 10371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5998330689790143,
+      "learning_rate": 2.6819278403092258e-06,
+      "loss": 1.0119,
+      "step": 10372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.383939988420904,
+      "learning_rate": 2.6810095240729715e-06,
+      "loss": 0.9282,
+      "step": 10373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3242400284369882,
+      "learning_rate": 2.6800913074888984e-06,
+      "loss": 0.6806,
+      "step": 10374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.249111112365853,
+      "learning_rate": 2.679173190596471e-06,
+      "loss": 0.6817,
+      "step": 10375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.802669883244295,
+      "learning_rate": 2.6782551734351396e-06,
+      "loss": 0.8032,
+      "step": 10376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.012494641775374,
+      "learning_rate": 2.6773372560443543e-06,
+      "loss": 0.7309,
+      "step": 10377
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3636578421750465,
+      "learning_rate": 2.6764194384635607e-06,
+      "loss": 0.6957,
+      "step": 10378
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7607989768566528,
+      "learning_rate": 2.6755017207321964e-06,
+      "loss": 0.7021,
+      "step": 10379
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.413024546199233,
+      "learning_rate": 2.6745841028897034e-06,
+      "loss": 0.7078,
+      "step": 10380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1924407760362274,
+      "learning_rate": 2.6736665849755073e-06,
+      "loss": 0.6605,
+      "step": 10381
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.377137834771019,
+      "learning_rate": 2.6727491670290405e-06,
+      "loss": 0.6559,
+      "step": 10382
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1118552005241322,
+      "learning_rate": 2.6718318490897256e-06,
+      "loss": 0.7116,
+      "step": 10383
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7397286237621006,
+      "learning_rate": 2.6709146311969813e-06,
+      "loss": 0.9125,
+      "step": 10384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.51936860540553,
+      "learning_rate": 2.669997513390222e-06,
+      "loss": 0.8897,
+      "step": 10385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8469957459732462,
+      "learning_rate": 2.669080495708859e-06,
+      "loss": 0.8807,
+      "step": 10386
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.325714243355846,
+      "learning_rate": 2.668163578192298e-06,
+      "loss": 0.7417,
+      "step": 10387
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.53286036487062,
+      "learning_rate": 2.6672467608799413e-06,
+      "loss": 0.8529,
+      "step": 10388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.020912961473751,
+      "learning_rate": 2.666330043811185e-06,
+      "loss": 0.8801,
+      "step": 10389
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.012020973413883,
+      "learning_rate": 2.6654134270254263e-06,
+      "loss": 0.9553,
+      "step": 10390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9898173544595126,
+      "learning_rate": 2.6644969105620535e-06,
+      "loss": 0.7719,
+      "step": 10391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.274297688221916,
+      "learning_rate": 2.663580494460446e-06,
+      "loss": 0.843,
+      "step": 10392
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2367592659040336,
+      "learning_rate": 2.6626641787599904e-06,
+      "loss": 0.7114,
+      "step": 10393
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3516999470141786,
+      "learning_rate": 2.66174796350006e-06,
+      "loss": 0.9407,
+      "step": 10394
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.492934857017271,
+      "learning_rate": 2.660831848720028e-06,
+      "loss": 0.7946,
+      "step": 10395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.774920021681413,
+      "learning_rate": 2.6599158344592606e-06,
+      "loss": 0.8059,
+      "step": 10396
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7333921251861129,
+      "learning_rate": 2.658999920757119e-06,
+      "loss": 0.7506,
+      "step": 10397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7783758683548783,
+      "learning_rate": 2.6580841076529697e-06,
+      "loss": 0.6418,
+      "step": 10398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.034376996593674,
+      "learning_rate": 2.657168395186157e-06,
+      "loss": 0.7081,
+      "step": 10399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.527158394598497,
+      "learning_rate": 2.656252783396038e-06,
+      "loss": 0.7656,
+      "step": 10400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4960657007371982,
+      "learning_rate": 2.655337272321957e-06,
+      "loss": 0.7221,
+      "step": 10401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4891707158051037,
+      "learning_rate": 2.654421862003256e-06,
+      "loss": 0.6656,
+      "step": 10402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.12780269769482,
+      "learning_rate": 2.653506552479271e-06,
+      "loss": 0.8961,
+      "step": 10403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.260482207594546,
+      "learning_rate": 2.6525913437893346e-06,
+      "loss": 0.7977,
+      "step": 10404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.496280912992727,
+      "learning_rate": 2.651676235972777e-06,
+      "loss": 0.7321,
+      "step": 10405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2303585148169307,
+      "learning_rate": 2.650761229068921e-06,
+      "loss": 0.6064,
+      "step": 10406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2829101869114017,
+      "learning_rate": 2.649846323117086e-06,
+      "loss": 0.8888,
+      "step": 10407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.205084979315622,
+      "learning_rate": 2.6489315181565896e-06,
+      "loss": 0.6253,
+      "step": 10408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2382161130519367,
+      "learning_rate": 2.648016814226742e-06,
+      "loss": 0.8106,
+      "step": 10409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.306633461234485,
+      "learning_rate": 2.6471022113668505e-06,
+      "loss": 0.7275,
+      "step": 10410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5174168898894114,
+      "learning_rate": 2.646187709616217e-06,
+      "loss": 0.7147,
+      "step": 10411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7324715964492725,
+      "learning_rate": 2.645273309014139e-06,
+      "loss": 0.7995,
+      "step": 10412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3713835846679414,
+      "learning_rate": 2.6443590095999127e-06,
+      "loss": 0.8434,
+      "step": 10413
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.95252674137498,
+      "learning_rate": 2.6434448114128252e-06,
+      "loss": 0.7846,
+      "step": 10414
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1370885746261967,
+      "learning_rate": 2.642530714492161e-06,
+      "loss": 0.7088,
+      "step": 10415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9861373978137578,
+      "learning_rate": 2.6416167188772052e-06,
+      "loss": 0.8441,
+      "step": 10416
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0737348058273204,
+      "learning_rate": 2.6407028246072287e-06,
+      "loss": 0.7201,
+      "step": 10417
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4906639363457725,
+      "learning_rate": 2.6397890317215067e-06,
+      "loss": 0.7555,
+      "step": 10418
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.047499450701141,
+      "learning_rate": 2.6388753402593083e-06,
+      "loss": 1.0195,
+      "step": 10419
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.433037067126614,
+      "learning_rate": 2.637961750259894e-06,
+      "loss": 0.7999,
+      "step": 10420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7720202702277013,
+      "learning_rate": 2.6370482617625236e-06,
+      "loss": 0.7458,
+      "step": 10421
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9530965128763684,
+      "learning_rate": 2.636134874806451e-06,
+      "loss": 0.8175,
+      "step": 10422
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.094211955654157,
+      "learning_rate": 2.6352215894309306e-06,
+      "loss": 0.8676,
+      "step": 10423
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0108213192849824,
+      "learning_rate": 2.6343084056752032e-06,
+      "loss": 0.5999,
+      "step": 10424
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7954312150905134,
+      "learning_rate": 2.6333953235785105e-06,
+      "loss": 0.829,
+      "step": 10425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2728786255588322,
+      "learning_rate": 2.632482343180094e-06,
+      "loss": 0.8865,
+      "step": 10426
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0402384057873477,
+      "learning_rate": 2.631569464519183e-06,
+      "loss": 0.9444,
+      "step": 10427
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0435761148778027,
+      "learning_rate": 2.6306566876350072e-06,
+      "loss": 0.8587,
+      "step": 10428
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6541982179045362,
+      "learning_rate": 2.6297440125667904e-06,
+      "loss": 0.9258,
+      "step": 10429
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1563619734902693,
+      "learning_rate": 2.6288314393537522e-06,
+      "loss": 0.7916,
+      "step": 10430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.150888049181431,
+      "learning_rate": 2.627918968035108e-06,
+      "loss": 0.8756,
+      "step": 10431
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.778730201222039,
+      "learning_rate": 2.627006598650067e-06,
+      "loss": 0.8428,
+      "step": 10432
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.100617716132582,
+      "learning_rate": 2.6260943312378384e-06,
+      "loss": 0.6888,
+      "step": 10433
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8310295529924276,
+      "learning_rate": 2.6251821658376265e-06,
+      "loss": 0.7807,
+      "step": 10434
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0050279484492592,
+      "learning_rate": 2.624270102488621e-06,
+      "loss": 0.9599,
+      "step": 10435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6444960457072866,
+      "learning_rate": 2.6233581412300222e-06,
+      "loss": 0.6634,
+      "step": 10436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8973568145742026,
+      "learning_rate": 2.6224462821010185e-06,
+      "loss": 0.9135,
+      "step": 10437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4649646963460143,
+      "learning_rate": 2.621534525140792e-06,
+      "loss": 0.5728,
+      "step": 10438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9745648797913675,
+      "learning_rate": 2.620622870388524e-06,
+      "loss": 0.6726,
+      "step": 10439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7522152578249393,
+      "learning_rate": 2.619711317883389e-06,
+      "loss": 0.9424,
+      "step": 10440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.566726269814572,
+      "learning_rate": 2.6187998676645633e-06,
+      "loss": 0.8111,
+      "step": 10441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.259866715521706,
+      "learning_rate": 2.617888519771207e-06,
+      "loss": 0.8517,
+      "step": 10442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1713705061177975,
+      "learning_rate": 2.616977274242488e-06,
+      "loss": 0.8819,
+      "step": 10443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2056501375914674,
+      "learning_rate": 2.616066131117563e-06,
+      "loss": 0.7901,
+      "step": 10444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9593998713088372,
+      "learning_rate": 2.6151550904355853e-06,
+      "loss": 0.7649,
+      "step": 10445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1156253715285838,
+      "learning_rate": 2.6142441522357055e-06,
+      "loss": 0.6851,
+      "step": 10446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3449679647872945,
+      "learning_rate": 2.6133333165570667e-06,
+      "loss": 0.9515,
+      "step": 10447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.251881106269531,
+      "learning_rate": 2.612422583438811e-06,
+      "loss": 0.8922,
+      "step": 10448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3273716102885063,
+      "learning_rate": 2.6115119529200748e-06,
+      "loss": 0.7776,
+      "step": 10449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2712189615590876,
+      "learning_rate": 2.6106014250399868e-06,
+      "loss": 0.8381,
+      "step": 10450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.396600633904221,
+      "learning_rate": 2.6096909998376794e-06,
+      "loss": 0.9463,
+      "step": 10451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8894334196346427,
+      "learning_rate": 2.6087806773522755e-06,
+      "loss": 0.8031,
+      "step": 10452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6840924326818971,
+      "learning_rate": 2.6078704576228876e-06,
+      "loss": 0.7273,
+      "step": 10453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.244319715194118,
+      "learning_rate": 2.6069603406886347e-06,
+      "loss": 0.8675,
+      "step": 10454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.552510497529658,
+      "learning_rate": 2.606050326588626e-06,
+      "loss": 0.7965,
+      "step": 10455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9247830981576814,
+      "learning_rate": 2.605140415361967e-06,
+      "loss": 0.7422,
+      "step": 10456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2372672071829993,
+      "learning_rate": 2.6042306070477576e-06,
+      "loss": 0.9783,
+      "step": 10457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7981919527207118,
+      "learning_rate": 2.6033209016850926e-06,
+      "loss": 0.6298,
+      "step": 10458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2051245190955746,
+      "learning_rate": 2.60241129931307e-06,
+      "loss": 0.7954,
+      "step": 10459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3796539768167446,
+      "learning_rate": 2.601501799970769e-06,
+      "loss": 0.7915,
+      "step": 10460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28639234137578,
+      "learning_rate": 2.6005924036972807e-06,
+      "loss": 0.7091,
+      "step": 10461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.892677253693117,
+      "learning_rate": 2.59968311053168e-06,
+      "loss": 0.786,
+      "step": 10462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3106432338595257,
+      "learning_rate": 2.5987739205130413e-06,
+      "loss": 0.7342,
+      "step": 10463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.127945912997889,
+      "learning_rate": 2.597864833680436e-06,
+      "loss": 0.873,
+      "step": 10464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4012267231962245,
+      "learning_rate": 2.596955850072928e-06,
+      "loss": 0.8189,
+      "step": 10465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7362079610950123,
+      "learning_rate": 2.5960469697295794e-06,
+      "loss": 0.75,
+      "step": 10466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3514760333456834,
+      "learning_rate": 2.595138192689446e-06,
+      "loss": 0.8239,
+      "step": 10467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8213398710648796,
+      "learning_rate": 2.594229518991579e-06,
+      "loss": 0.7735,
+      "step": 10468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.072166964408358,
+      "learning_rate": 2.593320948675029e-06,
+      "loss": 0.6965,
+      "step": 10469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8393286431352152,
+      "learning_rate": 2.5924124817788377e-06,
+      "loss": 0.7711,
+      "step": 10470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7708093970501695,
+      "learning_rate": 2.591504118342044e-06,
+      "loss": 0.7427,
+      "step": 10471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.047302404048681,
+      "learning_rate": 2.5905958584036826e-06,
+      "loss": 0.7004,
+      "step": 10472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.739649803128196,
+      "learning_rate": 2.5896877020027813e-06,
+      "loss": 0.7264,
+      "step": 10473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.107005001947943,
+      "learning_rate": 2.588779649178371e-06,
+      "loss": 0.7743,
+      "step": 10474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.066196527787757,
+      "learning_rate": 2.5878716999694675e-06,
+      "loss": 0.9466,
+      "step": 10475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.687597736799287,
+      "learning_rate": 2.5869638544150876e-06,
+      "loss": 0.6994,
+      "step": 10476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.224820211873988,
+      "learning_rate": 2.586056112554248e-06,
+      "loss": 0.7688,
+      "step": 10477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4161599633792172,
+      "learning_rate": 2.5851484744259503e-06,
+      "loss": 0.6505,
+      "step": 10478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.79512112024695,
+      "learning_rate": 2.5842409400692026e-06,
+      "loss": 0.8207,
+      "step": 10479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5862414082340197,
+      "learning_rate": 2.5833335095230016e-06,
+      "loss": 0.9518,
+      "step": 10480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0542803726480345,
+      "learning_rate": 2.5824261828263424e-06,
+      "loss": 0.8404,
+      "step": 10481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.953421586323113,
+      "learning_rate": 2.5815189600182145e-06,
+      "loss": 0.7155,
+      "step": 10482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.120368137027317,
+      "learning_rate": 2.5806118411376e-06,
+      "loss": 0.757,
+      "step": 10483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.179817043660737,
+      "learning_rate": 2.579704826223488e-06,
+      "loss": 0.6589,
+      "step": 10484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3007131686991085,
+      "learning_rate": 2.578797915314848e-06,
+      "loss": 0.9471,
+      "step": 10485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.546078474515981,
+      "learning_rate": 2.577891108450651e-06,
+      "loss": 1.0733,
+      "step": 10486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1268589549071786,
+      "learning_rate": 2.5769844056698702e-06,
+      "loss": 1.0646,
+      "step": 10487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8249799971575245,
+      "learning_rate": 2.576077807011465e-06,
+      "loss": 0.8469,
+      "step": 10488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8313819462820848,
+      "learning_rate": 2.575171312514395e-06,
+      "loss": 0.8097,
+      "step": 10489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7685051360983173,
+      "learning_rate": 2.5742649222176147e-06,
+      "loss": 0.7752,
+      "step": 10490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9571366746234158,
+      "learning_rate": 2.573358636160071e-06,
+      "loss": 0.8197,
+      "step": 10491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.204223774397855,
+      "learning_rate": 2.572452454380715e-06,
+      "loss": 0.844,
+      "step": 10492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1967224381302635,
+      "learning_rate": 2.571546376918479e-06,
+      "loss": 0.6608,
+      "step": 10493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3487243368845365,
+      "learning_rate": 2.570640403812306e-06,
+      "loss": 0.864,
+      "step": 10494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3141425864830922,
+      "learning_rate": 2.569734535101127e-06,
+      "loss": 0.7777,
+      "step": 10495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0969057670889315,
+      "learning_rate": 2.5688287708238647e-06,
+      "loss": 0.6305,
+      "step": 10496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.923585428505296,
+      "learning_rate": 2.5679231110194454e-06,
+      "loss": 0.9549,
+      "step": 10497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1995610791442783,
+      "learning_rate": 2.567017555726787e-06,
+      "loss": 0.751,
+      "step": 10498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2660151548755727,
+      "learning_rate": 2.5661121049848026e-06,
+      "loss": 0.8641,
+      "step": 10499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.253966803849376,
+      "learning_rate": 2.5652067588324015e-06,
+      "loss": 0.6919,
+      "step": 10500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8070323479306885,
+      "learning_rate": 2.564301517308487e-06,
+      "loss": 0.8721,
+      "step": 10501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.247709109243232,
+      "learning_rate": 2.5633963804519623e-06,
+      "loss": 0.616,
+      "step": 10502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9232906353070955,
+      "learning_rate": 2.5624913483017246e-06,
+      "loss": 0.8311,
+      "step": 10503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1176407479013784,
+      "learning_rate": 2.5615864208966573e-06,
+      "loss": 0.9377,
+      "step": 10504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.881360730543118,
+      "learning_rate": 2.560681598275655e-06,
+      "loss": 0.8802,
+      "step": 10505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9812984587532092,
+      "learning_rate": 2.559776880477596e-06,
+      "loss": 0.7839,
+      "step": 10506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8667669733506855,
+      "learning_rate": 2.55887226754136e-06,
+      "loss": 0.8643,
+      "step": 10507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.576718276447197,
+      "learning_rate": 2.5579677595058196e-06,
+      "loss": 0.5916,
+      "step": 10508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.045270966678765,
+      "learning_rate": 2.55706335640984e-06,
+      "loss": 0.7694,
+      "step": 10509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.819404313589488,
+      "learning_rate": 2.5561590582922934e-06,
+      "loss": 0.774,
+      "step": 10510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3201838941444817,
+      "learning_rate": 2.5552548651920317e-06,
+      "loss": 0.8111,
+      "step": 10511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.979100638393201,
+      "learning_rate": 2.5543507771479137e-06,
+      "loss": 0.6751,
+      "step": 10512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9594285813051904,
+      "learning_rate": 2.553446794198793e-06,
+      "loss": 0.7331,
+      "step": 10513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8444823783675013,
+      "learning_rate": 2.552542916383507e-06,
+      "loss": 0.7765,
+      "step": 10514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9199974311410113,
+      "learning_rate": 2.551639143740905e-06,
+      "loss": 0.8078,
+      "step": 10515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0202776321773765,
+      "learning_rate": 2.5507354763098226e-06,
+      "loss": 0.5453,
+      "step": 10516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.51758283806975,
+      "learning_rate": 2.5498319141290904e-06,
+      "loss": 0.7254,
+      "step": 10517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8340992685490083,
+      "learning_rate": 2.548928457237538e-06,
+      "loss": 0.8109,
+      "step": 10518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0963905801687972,
+      "learning_rate": 2.5480251056739874e-06,
+      "loss": 0.7145,
+      "step": 10519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6593823126072502,
+      "learning_rate": 2.5471218594772602e-06,
+      "loss": 0.8344,
+      "step": 10520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4013841434421925,
+      "learning_rate": 2.5462187186861697e-06,
+      "loss": 0.755,
+      "step": 10521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0843365687198074,
+      "learning_rate": 2.5453156833395254e-06,
+      "loss": 0.6463,
+      "step": 10522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3368566171227747,
+      "learning_rate": 2.544412753476134e-06,
+      "loss": 0.6791,
+      "step": 10523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0917865521611967,
+      "learning_rate": 2.543509929134794e-06,
+      "loss": 0.7753,
+      "step": 10524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.013246315627442,
+      "learning_rate": 2.5426072103543043e-06,
+      "loss": 0.9573,
+      "step": 10525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.406641321343057,
+      "learning_rate": 2.5417045971734556e-06,
+      "loss": 0.7172,
+      "step": 10526
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.50605786659914,
+      "learning_rate": 2.5408020896310333e-06,
+      "loss": 0.8915,
+      "step": 10527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.156162181594672,
+      "learning_rate": 2.5398996877658256e-06,
+      "loss": 0.6835,
+      "step": 10528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.758870031340736,
+      "learning_rate": 2.5389973916166037e-06,
+      "loss": 0.7391,
+      "step": 10529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.415597916113764,
+      "learning_rate": 2.5380952012221473e-06,
+      "loss": 1.0348,
+      "step": 10530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.079618660650479,
+      "learning_rate": 2.5371931166212216e-06,
+      "loss": 0.8821,
+      "step": 10531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.644707411799997,
+      "learning_rate": 2.536291137852594e-06,
+      "loss": 0.8932,
+      "step": 10532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.47030619589864,
+      "learning_rate": 2.535389264955023e-06,
+      "loss": 0.7789,
+      "step": 10533
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1467908309207373,
+      "learning_rate": 2.534487497967262e-06,
+      "loss": 0.7969,
+      "step": 10534
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.237899029129065,
+      "learning_rate": 2.5335858369280674e-06,
+      "loss": 0.7349,
+      "step": 10535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1593482814339904,
+      "learning_rate": 2.5326842818761814e-06,
+      "loss": 0.9478,
+      "step": 10536
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.634220781553909,
+      "learning_rate": 2.5317828328503445e-06,
+      "loss": 0.5628,
+      "step": 10537
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2964740933610637,
+      "learning_rate": 2.5308814898892985e-06,
+      "loss": 0.8776,
+      "step": 10538
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2890706532702323,
+      "learning_rate": 2.529980253031774e-06,
+      "loss": 0.8245,
+      "step": 10539
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9268702925197276,
+      "learning_rate": 2.529079122316499e-06,
+      "loss": 0.8048,
+      "step": 10540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.649525541418398,
+      "learning_rate": 2.5281780977821967e-06,
+      "loss": 0.756,
+      "step": 10541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0204706753279478,
+      "learning_rate": 2.5272771794675866e-06,
+      "loss": 0.5728,
+      "step": 10542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.111737330204023,
+      "learning_rate": 2.526376367411384e-06,
+      "loss": 0.925,
+      "step": 10543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2110803769546512,
+      "learning_rate": 2.5254756616522953e-06,
+      "loss": 0.758,
+      "step": 10544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.081815939804654,
+      "learning_rate": 2.52457506222903e-06,
+      "loss": 0.8077,
+      "step": 10545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1908060523474995,
+      "learning_rate": 2.52367456918029e-06,
+      "loss": 0.9752,
+      "step": 10546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.213356201020907,
+      "learning_rate": 2.522774182544765e-06,
+      "loss": 0.7733,
+      "step": 10547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0291053309354057,
+      "learning_rate": 2.5218739023611523e-06,
+      "loss": 0.8647,
+      "step": 10548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2566436256646845,
+      "learning_rate": 2.5209737286681367e-06,
+      "loss": 0.6959,
+      "step": 10549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.180166760931707,
+      "learning_rate": 2.5200736615044012e-06,
+      "loss": 0.7801,
+      "step": 10550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.335985900692321,
+      "learning_rate": 2.5191737009086235e-06,
+      "loss": 0.8586,
+      "step": 10551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0798718605371438,
+      "learning_rate": 2.5182738469194757e-06,
+      "loss": 0.6493,
+      "step": 10552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7681375001932493,
+      "learning_rate": 2.5173740995756313e-06,
+      "loss": 0.6429,
+      "step": 10553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.778114279656514,
+      "learning_rate": 2.5164744589157488e-06,
+      "loss": 0.7844,
+      "step": 10554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.589660864994518,
+      "learning_rate": 2.5155749249784876e-06,
+      "loss": 0.9659,
+      "step": 10555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8450841216958862,
+      "learning_rate": 2.514675497802508e-06,
+      "loss": 0.6138,
+      "step": 10556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8491549087020505,
+      "learning_rate": 2.513776177426457e-06,
+      "loss": 0.7459,
+      "step": 10557
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2471511876788166,
+      "learning_rate": 2.51287696388898e-06,
+      "loss": 0.5934,
+      "step": 10558
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.294987870372235,
+      "learning_rate": 2.5119778572287195e-06,
+      "loss": 0.6376,
+      "step": 10559
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.725867696908553,
+      "learning_rate": 2.5110788574843113e-06,
+      "loss": 0.7515,
+      "step": 10560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2114715296603324,
+      "learning_rate": 2.510179964694388e-06,
+      "loss": 0.7085,
+      "step": 10561
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.30448884205131,
+      "learning_rate": 2.5092811788975754e-06,
+      "loss": 0.8825,
+      "step": 10562
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0273185648384506,
+      "learning_rate": 2.508382500132499e-06,
+      "loss": 0.7392,
+      "step": 10563
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1883256210965056,
+      "learning_rate": 2.5074839284377774e-06,
+      "loss": 0.7349,
+      "step": 10564
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2956545392209415,
+      "learning_rate": 2.506585463852019e-06,
+      "loss": 0.7891,
+      "step": 10565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.847711243387223,
+      "learning_rate": 2.5056871064138383e-06,
+      "loss": 0.8387,
+      "step": 10566
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.090124834719396,
+      "learning_rate": 2.504788856161838e-06,
+      "loss": 0.7068,
+      "step": 10567
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9497481589562842,
+      "learning_rate": 2.5038907131346168e-06,
+      "loss": 0.8837,
+      "step": 10568
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3287542714143306,
+      "learning_rate": 2.5029926773707713e-06,
+      "loss": 0.6682,
+      "step": 10569
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0660017682578085,
+      "learning_rate": 2.50209474890889e-06,
+      "loss": 0.9903,
+      "step": 10570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7820246542778,
+      "learning_rate": 2.5011969277875643e-06,
+      "loss": 0.8991,
+      "step": 10571
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7138459164429154,
+      "learning_rate": 2.5002992140453674e-06,
+      "loss": 0.8512,
+      "step": 10572
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.217852048698336,
+      "learning_rate": 2.499401607720883e-06,
+      "loss": 0.8032,
+      "step": 10573
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.828633423889349,
+      "learning_rate": 2.49850410885268e-06,
+      "loss": 0.9798,
+      "step": 10574
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4019424242094387,
+      "learning_rate": 2.4976067174793268e-06,
+      "loss": 0.7562,
+      "step": 10575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1387757630612403,
+      "learning_rate": 2.496709433639386e-06,
+      "loss": 0.8814,
+      "step": 10576
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1474430375714957,
+      "learning_rate": 2.495812257371416e-06,
+      "loss": 0.9609,
+      "step": 10577
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.352166863461294,
+      "learning_rate": 2.4949151887139705e-06,
+      "loss": 0.7057,
+      "step": 10578
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1222313328354958,
+      "learning_rate": 2.4940182277055987e-06,
+      "loss": 0.6382,
+      "step": 10579
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.135699230457888,
+      "learning_rate": 2.4931213743848428e-06,
+      "loss": 0.8431,
+      "step": 10580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0842813757920613,
+      "learning_rate": 2.4922246287902458e-06,
+      "loss": 0.5665,
+      "step": 10581
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2315095331009087,
+      "learning_rate": 2.491327990960342e-06,
+      "loss": 0.8206,
+      "step": 10582
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.099749161780314,
+      "learning_rate": 2.4904314609336616e-06,
+      "loss": 0.8959,
+      "step": 10583
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.882662222878393,
+      "learning_rate": 2.4895350387487304e-06,
+      "loss": 0.8649,
+      "step": 10584
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3069815534497766,
+      "learning_rate": 2.48863872444407e-06,
+      "loss": 0.7583,
+      "step": 10585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1244010011839434,
+      "learning_rate": 2.4877425180581964e-06,
+      "loss": 0.8952,
+      "step": 10586
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3599548297459014,
+      "learning_rate": 2.486846419629622e-06,
+      "loss": 0.6467,
+      "step": 10587
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.536158893078955,
+      "learning_rate": 2.4859504291968522e-06,
+      "loss": 0.8885,
+      "step": 10588
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0939617638011874,
+      "learning_rate": 2.485054546798395e-06,
+      "loss": 0.8659,
+      "step": 10589
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1730020987858847,
+      "learning_rate": 2.4841587724727416e-06,
+      "loss": 0.6782,
+      "step": 10590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9120976819654936,
+      "learning_rate": 2.4832631062583906e-06,
+      "loss": 0.6914,
+      "step": 10591
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3714364886215105,
+      "learning_rate": 2.4823675481938293e-06,
+      "loss": 1.0008,
+      "step": 10592
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5252649035018826,
+      "learning_rate": 2.4814720983175413e-06,
+      "loss": 0.7319,
+      "step": 10593
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0912670998312506,
+      "learning_rate": 2.4805767566680057e-06,
+      "loss": 0.8468,
+      "step": 10594
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.973312310173647,
+      "learning_rate": 2.4796815232836965e-06,
+      "loss": 0.7681,
+      "step": 10595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.602118929397051,
+      "learning_rate": 2.478786398203089e-06,
+      "loss": 0.6691,
+      "step": 10596
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1712351534207652,
+      "learning_rate": 2.4778913814646433e-06,
+      "loss": 0.6793,
+      "step": 10597
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.647890835153736,
+      "learning_rate": 2.47699647310682e-06,
+      "loss": 0.8448,
+      "step": 10598
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8289972454863037,
+      "learning_rate": 2.4761016731680792e-06,
+      "loss": 0.841,
+      "step": 10599
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4226208500574713,
+      "learning_rate": 2.4752069816868707e-06,
+      "loss": 0.8532,
+      "step": 10600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.185445747466083,
+      "learning_rate": 2.474312398701641e-06,
+      "loss": 0.7682,
+      "step": 10601
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2808501379716515,
+      "learning_rate": 2.473417924250833e-06,
+      "loss": 0.7143,
+      "step": 10602
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9523215994641885,
+      "learning_rate": 2.472523558372883e-06,
+      "loss": 0.8676,
+      "step": 10603
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3619299192885137,
+      "learning_rate": 2.4716293011062248e-06,
+      "loss": 0.6754,
+      "step": 10604
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9649559867071265,
+      "learning_rate": 2.470735152489287e-06,
+      "loss": 0.8069,
+      "step": 10605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.18097902114621,
+      "learning_rate": 2.4698411125604903e-06,
+      "loss": 0.8138,
+      "step": 10606
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0786773087464363,
+      "learning_rate": 2.46894718135826e-06,
+      "loss": 0.9073,
+      "step": 10607
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0015055818667737,
+      "learning_rate": 2.4680533589210028e-06,
+      "loss": 0.6146,
+      "step": 10608
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7807395237793375,
+      "learning_rate": 2.467159645287133e-06,
+      "loss": 0.8179,
+      "step": 10609
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.326104339593203,
+      "learning_rate": 2.4662660404950544e-06,
+      "loss": 0.9177,
+      "step": 10610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.314631180591328,
+      "learning_rate": 2.465372544583167e-06,
+      "loss": 0.8374,
+      "step": 10611
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9161192366595358,
+      "learning_rate": 2.4644791575898665e-06,
+      "loss": 0.6382,
+      "step": 10612
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.049137021923541,
+      "learning_rate": 2.4635858795535413e-06,
+      "loss": 0.8555,
+      "step": 10613
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.800265366730342,
+      "learning_rate": 2.4626927105125834e-06,
+      "loss": 0.9055,
+      "step": 10614
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0639305909124923,
+      "learning_rate": 2.461799650505369e-06,
+      "loss": 0.7536,
+      "step": 10615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.751198085569275,
+      "learning_rate": 2.4609066995702745e-06,
+      "loss": 0.7269,
+      "step": 10616
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1378624670533335,
+      "learning_rate": 2.460013857745676e-06,
+      "loss": 0.7777,
+      "step": 10617
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3072223192124353,
+      "learning_rate": 2.4591211250699393e-06,
+      "loss": 0.7269,
+      "step": 10618
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.955283401902372,
+      "learning_rate": 2.4582285015814263e-06,
+      "loss": 0.6922,
+      "step": 10619
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5084221873553454,
+      "learning_rate": 2.457335987318495e-06,
+      "loss": 0.8047,
+      "step": 10620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0794192906288163,
+      "learning_rate": 2.4564435823194977e-06,
+      "loss": 0.776,
+      "step": 10621
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0512646908104026,
+      "learning_rate": 2.455551286622788e-06,
+      "loss": 0.6364,
+      "step": 10622
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0550818528170827,
+      "learning_rate": 2.454659100266703e-06,
+      "loss": 0.8621,
+      "step": 10623
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0273628719564758,
+      "learning_rate": 2.4537670232895866e-06,
+      "loss": 0.5727,
+      "step": 10624
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.96667685128792,
+      "learning_rate": 2.4528750557297745e-06,
+      "loss": 0.7948,
+      "step": 10625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.339738684029364,
+      "learning_rate": 2.4519831976255892e-06,
+      "loss": 0.6801,
+      "step": 10626
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4590454771666383,
+      "learning_rate": 2.451091449015364e-06,
+      "loss": 0.6939,
+      "step": 10627
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0785808996527964,
+      "learning_rate": 2.4501998099374147e-06,
+      "loss": 0.7683,
+      "step": 10628
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.023466798063421,
+      "learning_rate": 2.4493082804300585e-06,
+      "loss": 0.9335,
+      "step": 10629
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.587321157177841,
+      "learning_rate": 2.4484168605316062e-06,
+      "loss": 0.8991,
+      "step": 10630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.203713063227471,
+      "learning_rate": 2.4475255502803624e-06,
+      "loss": 0.9016,
+      "step": 10631
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8008005292384028,
+      "learning_rate": 2.4466343497146345e-06,
+      "loss": 0.8454,
+      "step": 10632
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9943761704899623,
+      "learning_rate": 2.445743258872711e-06,
+      "loss": 0.5263,
+      "step": 10633
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.555168683726513,
+      "learning_rate": 2.4448522777928903e-06,
+      "loss": 0.9003,
+      "step": 10634
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.541001127733336,
+      "learning_rate": 2.4439614065134576e-06,
+      "loss": 0.6131,
+      "step": 10635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.493377517720112,
+      "learning_rate": 2.443070645072696e-06,
+      "loss": 0.6767,
+      "step": 10636
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2408434675938715,
+      "learning_rate": 2.4421799935088837e-06,
+      "loss": 0.7826,
+      "step": 10637
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.387882897428757,
+      "learning_rate": 2.441289451860293e-06,
+      "loss": 0.8179,
+      "step": 10638
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9831907471170689,
+      "learning_rate": 2.4403990201651915e-06,
+      "loss": 0.8188,
+      "step": 10639
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0205498699757178,
+      "learning_rate": 2.4395086984618486e-06,
+      "loss": 0.6244,
+      "step": 10640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9446742360315412,
+      "learning_rate": 2.4386184867885154e-06,
+      "loss": 0.6875,
+      "step": 10641
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1409105676284863,
+      "learning_rate": 2.437728385183453e-06,
+      "loss": 0.889,
+      "step": 10642
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5684807257548803,
+      "learning_rate": 2.436838393684907e-06,
+      "loss": 0.8394,
+      "step": 10643
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8601821220257264,
+      "learning_rate": 2.435948512331125e-06,
+      "loss": 0.889,
+      "step": 10644
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2669411806796615,
+      "learning_rate": 2.4350587411603454e-06,
+      "loss": 0.9597,
+      "step": 10645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1516533623179859,
+      "learning_rate": 2.434169080210802e-06,
+      "loss": 0.7162,
+      "step": 10646
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8175760412358004,
+      "learning_rate": 2.433279529520732e-06,
+      "loss": 0.8747,
+      "step": 10647
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8647443800738164,
+      "learning_rate": 2.432390089128354e-06,
+      "loss": 0.7261,
+      "step": 10648
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.009390322282209,
+      "learning_rate": 2.4315007590718913e-06,
+      "loss": 0.6707,
+      "step": 10649
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6142605007353894,
+      "learning_rate": 2.4306115393895645e-06,
+      "loss": 0.8227,
+      "step": 10650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9782863192658321,
+      "learning_rate": 2.4297224301195795e-06,
+      "loss": 0.5721,
+      "step": 10651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.41153932273888,
+      "learning_rate": 2.4288334313001467e-06,
+      "loss": 0.8633,
+      "step": 10652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.536224315896448,
+      "learning_rate": 2.4279445429694683e-06,
+      "loss": 0.9446,
+      "step": 10653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.407566267905626,
+      "learning_rate": 2.427055765165741e-06,
+      "loss": 0.7864,
+      "step": 10654
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2138393748344494,
+      "learning_rate": 2.4261670979271577e-06,
+      "loss": 0.8607,
+      "step": 10655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.282820799990956,
+      "learning_rate": 2.4252785412919065e-06,
+      "loss": 0.7413,
+      "step": 10656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2594482013934583,
+      "learning_rate": 2.4243900952981685e-06,
+      "loss": 0.6955,
+      "step": 10657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1996451463157891,
+      "learning_rate": 2.423501759984128e-06,
+      "loss": 0.6597,
+      "step": 10658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21060666465764,
+      "learning_rate": 2.4226135353879516e-06,
+      "loss": 0.6415,
+      "step": 10659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9201979148311408,
+      "learning_rate": 2.4217254215478137e-06,
+      "loss": 0.6661,
+      "step": 10660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8605097878134043,
+      "learning_rate": 2.420837418501876e-06,
+      "loss": 0.8243,
+      "step": 10661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9959493867749185,
+      "learning_rate": 2.4199495262882993e-06,
+      "loss": 0.9277,
+      "step": 10662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9261360329140775,
+      "learning_rate": 2.419061744945237e-06,
+      "loss": 0.6935,
+      "step": 10663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7455355157922312,
+      "learning_rate": 2.4181740745108377e-06,
+      "loss": 0.7326,
+      "step": 10664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.212572741883704,
+      "learning_rate": 2.4172865150232528e-06,
+      "loss": 0.9157,
+      "step": 10665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6249567352949983,
+      "learning_rate": 2.4163990665206153e-06,
+      "loss": 0.6371,
+      "step": 10666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0951017455689116,
+      "learning_rate": 2.415511729041063e-06,
+      "loss": 0.6511,
+      "step": 10667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.449154427315735,
+      "learning_rate": 2.414624502622731e-06,
+      "loss": 0.852,
+      "step": 10668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8601973327867363,
+      "learning_rate": 2.413737387303739e-06,
+      "loss": 0.6813,
+      "step": 10669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.545837489045014,
+      "learning_rate": 2.412850383122213e-06,
+      "loss": 0.8768,
+      "step": 10670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.111904004248509,
+      "learning_rate": 2.411963490116267e-06,
+      "loss": 0.6829,
+      "step": 10671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.309427880144203,
+      "learning_rate": 2.4110767083240145e-06,
+      "loss": 0.795,
+      "step": 10672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.376560587755536,
+      "learning_rate": 2.410190037783562e-06,
+      "loss": 0.849,
+      "step": 10673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8097112440906606,
+      "learning_rate": 2.4093034785330087e-06,
+      "loss": 0.7735,
+      "step": 10674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.081339331579189,
+      "learning_rate": 2.408417030610457e-06,
+      "loss": 0.8192,
+      "step": 10675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3097609047699343,
+      "learning_rate": 2.4075306940539988e-06,
+      "loss": 0.6716,
+      "step": 10676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.557448071337241,
+      "learning_rate": 2.4066444689017173e-06,
+      "loss": 0.7989,
+      "step": 10677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.097109441365908,
+      "learning_rate": 2.4057583551917e-06,
+      "loss": 0.7512,
+      "step": 10678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.039630042553069,
+      "learning_rate": 2.4048723529620246e-06,
+      "loss": 0.7052,
+      "step": 10679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.833056336293738,
+      "learning_rate": 2.403986462250763e-06,
+      "loss": 0.7624,
+      "step": 10680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.477178856173003,
+      "learning_rate": 2.4031006830959853e-06,
+      "loss": 0.7894,
+      "step": 10681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.348282789400161,
+      "learning_rate": 2.4022150155357526e-06,
+      "loss": 0.7994,
+      "step": 10682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9754770937418489,
+      "learning_rate": 2.40132945960813e-06,
+      "loss": 0.8675,
+      "step": 10683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.403685485809211,
+      "learning_rate": 2.4004440153511642e-06,
+      "loss": 0.76,
+      "step": 10684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8084697493903108,
+      "learning_rate": 2.399558682802911e-06,
+      "loss": 0.8491,
+      "step": 10685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3608261464839764,
+      "learning_rate": 2.3986734620014146e-06,
+      "loss": 0.8137,
+      "step": 10686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2617682757234694,
+      "learning_rate": 2.3977883529847087e-06,
+      "loss": 0.6696,
+      "step": 10687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.251402588657994,
+      "learning_rate": 2.396903355790835e-06,
+      "loss": 0.8153,
+      "step": 10688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4689479906744936,
+      "learning_rate": 2.396018470457821e-06,
+      "loss": 0.7402,
+      "step": 10689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0764225140248858,
+      "learning_rate": 2.395133697023693e-06,
+      "loss": 0.9647,
+      "step": 10690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2753411119308091,
+      "learning_rate": 2.394249035526472e-06,
+      "loss": 0.6576,
+      "step": 10691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.040667844463106,
+      "learning_rate": 2.3933644860041703e-06,
+      "loss": 0.6495,
+      "step": 10692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2843051946376898,
+      "learning_rate": 2.3924800484948045e-06,
+      "loss": 0.6803,
+      "step": 10693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1933895791439675,
+      "learning_rate": 2.3915957230363783e-06,
+      "loss": 0.6356,
+      "step": 10694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.670519560622875,
+      "learning_rate": 2.390711509666893e-06,
+      "loss": 0.8155,
+      "step": 10695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9243481273765373,
+      "learning_rate": 2.389827408424345e-06,
+      "loss": 0.6924,
+      "step": 10696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.619820162168145,
+      "learning_rate": 2.388943419346727e-06,
+      "loss": 0.7434,
+      "step": 10697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2113446157904986,
+      "learning_rate": 2.388059542472025e-06,
+      "loss": 0.6149,
+      "step": 10698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3381645330583805,
+      "learning_rate": 2.3871757778382216e-06,
+      "loss": 0.7451,
+      "step": 10699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1321606405954703,
+      "learning_rate": 2.386292125483292e-06,
+      "loss": 0.8633,
+      "step": 10700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203115862838405,
+      "learning_rate": 2.385408585445215e-06,
+      "loss": 0.7726,
+      "step": 10701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0756959485458617,
+      "learning_rate": 2.38452515776195e-06,
+      "loss": 0.5704,
+      "step": 10702
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9628463383455748,
+      "learning_rate": 2.3836418424714665e-06,
+      "loss": 0.8687,
+      "step": 10703
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7258245607336948,
+      "learning_rate": 2.3827586396117207e-06,
+      "loss": 0.7351,
+      "step": 10704
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4659804090016872,
+      "learning_rate": 2.3818755492206624e-06,
+      "loss": 0.6864,
+      "step": 10705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2715817009224804,
+      "learning_rate": 2.3809925713362437e-06,
+      "loss": 0.8988,
+      "step": 10706
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.003714363806792,
+      "learning_rate": 2.3801097059964075e-06,
+      "loss": 0.5944,
+      "step": 10707
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1484727322013213,
+      "learning_rate": 2.3792269532390923e-06,
+      "loss": 0.8087,
+      "step": 10708
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2238598439288775,
+      "learning_rate": 2.378344313102231e-06,
+      "loss": 0.8484,
+      "step": 10709
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1495157288837863,
+      "learning_rate": 2.377461785623752e-06,
+      "loss": 0.6333,
+      "step": 10710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.233129899283949,
+      "learning_rate": 2.3765793708415825e-06,
+      "loss": 0.8423,
+      "step": 10711
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9992140235459102,
+      "learning_rate": 2.3756970687936397e-06,
+      "loss": 0.9454,
+      "step": 10712
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8734347346307947,
+      "learning_rate": 2.3748148795178387e-06,
+      "loss": 0.6672,
+      "step": 10713
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6822897602217113,
+      "learning_rate": 2.373932803052089e-06,
+      "loss": 0.873,
+      "step": 10714
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6518618094833233,
+      "learning_rate": 2.3730508394342957e-06,
+      "loss": 0.8025,
+      "step": 10715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.825409286084908,
+      "learning_rate": 2.372168988702358e-06,
+      "loss": 0.7921,
+      "step": 10716
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.519846531931346,
+      "learning_rate": 2.3712872508941714e-06,
+      "loss": 0.7548,
+      "step": 10717
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9268269192085163,
+      "learning_rate": 2.3704056260476237e-06,
+      "loss": 0.8527,
+      "step": 10718
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6761935443544655,
+      "learning_rate": 2.369524114200607e-06,
+      "loss": 0.6609,
+      "step": 10719
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7088201328021255,
+      "learning_rate": 2.368642715390994e-06,
+      "loss": 0.8413,
+      "step": 10720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9042022177895515,
+      "learning_rate": 2.3677614296566654e-06,
+      "loss": 0.7757,
+      "step": 10721
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2410690163521942,
+      "learning_rate": 2.3668802570354903e-06,
+      "loss": 0.7507,
+      "step": 10722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6369978187237704,
+      "learning_rate": 2.365999197565335e-06,
+      "loss": 0.7232,
+      "step": 10723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3508835030675823,
+      "learning_rate": 2.3651182512840604e-06,
+      "loss": 0.8402,
+      "step": 10724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5576487744328062,
+      "learning_rate": 2.3642374182295207e-06,
+      "loss": 0.7247,
+      "step": 10725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.772298927239848,
+      "learning_rate": 2.363356698439573e-06,
+      "loss": 0.6171,
+      "step": 10726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3434134983930783,
+      "learning_rate": 2.362476091952059e-06,
+      "loss": 0.8644,
+      "step": 10727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.745439204680952,
+      "learning_rate": 2.361595598804819e-06,
+      "loss": 0.8449,
+      "step": 10728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5666445524205668,
+      "learning_rate": 2.360715219035694e-06,
+      "loss": 0.8476,
+      "step": 10729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4876161110615036,
+      "learning_rate": 2.3598349526825137e-06,
+      "loss": 0.6178,
+      "step": 10730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.986508938205371,
+      "learning_rate": 2.358954799783106e-06,
+      "loss": 0.7252,
+      "step": 10731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1870082081648445,
+      "learning_rate": 2.358074760375292e-06,
+      "loss": 0.7452,
+      "step": 10732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.002208013538187,
+      "learning_rate": 2.3571948344968896e-06,
+      "loss": 0.7204,
+      "step": 10733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.24227255048584,
+      "learning_rate": 2.356315022185712e-06,
+      "loss": 0.7159,
+      "step": 10734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.130539599774342,
+      "learning_rate": 2.3554353234795624e-06,
+      "loss": 0.7321,
+      "step": 10735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2172821134283967,
+      "learning_rate": 2.3545557384162494e-06,
+      "loss": 0.6231,
+      "step": 10736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0592489431874035,
+      "learning_rate": 2.3536762670335696e-06,
+      "loss": 0.8186,
+      "step": 10737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.991962665374153,
+      "learning_rate": 2.3527969093693105e-06,
+      "loss": 0.849,
+      "step": 10738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.207433675056835,
+      "learning_rate": 2.3519176654612657e-06,
+      "loss": 0.8895,
+      "step": 10739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.240468812803458,
+      "learning_rate": 2.3510385353472163e-06,
+      "loss": 0.814,
+      "step": 10740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.062252228740222,
+      "learning_rate": 2.3501595190649405e-06,
+      "loss": 0.8161,
+      "step": 10741
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3270776625774756,
+      "learning_rate": 2.3492806166522125e-06,
+      "loss": 0.8334,
+      "step": 10742
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0396489007351781,
+      "learning_rate": 2.3484018281467973e-06,
+      "loss": 0.7371,
+      "step": 10743
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1849884647138027,
+      "learning_rate": 2.3475231535864653e-06,
+      "loss": 0.7245,
+      "step": 10744
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.011641200382685,
+      "learning_rate": 2.346644593008966e-06,
+      "loss": 0.6751,
+      "step": 10745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7141452050986064,
+      "learning_rate": 2.3457661464520605e-06,
+      "loss": 0.8672,
+      "step": 10746
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.334779938233791,
+      "learning_rate": 2.3448878139534948e-06,
+      "loss": 0.8115,
+      "step": 10747
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.619475032021045,
+      "learning_rate": 2.344009595551013e-06,
+      "loss": 0.8757,
+      "step": 10748
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7553576799255803,
+      "learning_rate": 2.3431314912823543e-06,
+      "loss": 0.7924,
+      "step": 10749
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9635993410740615,
+      "learning_rate": 2.3422535011852522e-06,
+      "loss": 0.9585,
+      "step": 10750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.274498101009422,
+      "learning_rate": 2.3413756252974366e-06,
+      "loss": 0.9981,
+      "step": 10751
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9755645840883589,
+      "learning_rate": 2.3404978636566312e-06,
+      "loss": 0.8224,
+      "step": 10752
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.524004797213335,
+      "learning_rate": 2.3396202163005536e-06,
+      "loss": 0.8791,
+      "step": 10753
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9791757670136942,
+      "learning_rate": 2.338742683266923e-06,
+      "loss": 0.8301,
+      "step": 10754
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6726852457269064,
+      "learning_rate": 2.3378652645934473e-06,
+      "loss": 0.7023,
+      "step": 10755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9372646035867265,
+      "learning_rate": 2.3369879603178264e-06,
+      "loss": 0.8275,
+      "step": 10756
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.871217574832135,
+      "learning_rate": 2.336110770477766e-06,
+      "loss": 0.8108,
+      "step": 10757
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0061939230515398,
+      "learning_rate": 2.3352336951109584e-06,
+      "loss": 0.8364,
+      "step": 10758
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.441817318122058,
+      "learning_rate": 2.3343567342550933e-06,
+      "loss": 0.8715,
+      "step": 10759
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.071933359020409,
+      "learning_rate": 2.3334798879478563e-06,
+      "loss": 0.682,
+      "step": 10760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1483423002792226,
+      "learning_rate": 2.332603156226926e-06,
+      "loss": 0.6673,
+      "step": 10761
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2371094033706522,
+      "learning_rate": 2.3317265391299813e-06,
+      "loss": 0.7881,
+      "step": 10762
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1166954369860203,
+      "learning_rate": 2.330850036694687e-06,
+      "loss": 0.9388,
+      "step": 10763
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9673675233314512,
+      "learning_rate": 2.329973648958712e-06,
+      "loss": 0.787,
+      "step": 10764
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0746541195091504,
+      "learning_rate": 2.3290973759597165e-06,
+      "loss": 0.8273,
+      "step": 10765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7895542915109983,
+      "learning_rate": 2.328221217735355e-06,
+      "loss": 0.9335,
+      "step": 10766
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2191355631361196,
+      "learning_rate": 2.3273451743232777e-06,
+      "loss": 0.7235,
+      "step": 10767
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2601031265956097,
+      "learning_rate": 2.326469245761131e-06,
+      "loss": 0.6486,
+      "step": 10768
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.391740497413482,
+      "learning_rate": 2.3255934320865555e-06,
+      "loss": 0.6231,
+      "step": 10769
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1040490980234603,
+      "learning_rate": 2.324717733337186e-06,
+      "loss": 0.9129,
+      "step": 10770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.986781165921222,
+      "learning_rate": 2.3238421495506515e-06,
+      "loss": 0.928,
+      "step": 10771
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4077431913844465,
+      "learning_rate": 2.322966680764582e-06,
+      "loss": 0.9173,
+      "step": 10772
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2944249595191613,
+      "learning_rate": 2.322091327016597e-06,
+      "loss": 0.5824,
+      "step": 10773
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3096625782724916,
+      "learning_rate": 2.3212160883443107e-06,
+      "loss": 0.8136,
+      "step": 10774
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1056406284983384,
+      "learning_rate": 2.320340964785336e-06,
+      "loss": 0.8963,
+      "step": 10775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.434466645278011,
+      "learning_rate": 2.3194659563772754e-06,
+      "loss": 0.8089,
+      "step": 10776
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.770233075362702,
+      "learning_rate": 2.318591063157736e-06,
+      "loss": 1.0148,
+      "step": 10777
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0648743837457606,
+      "learning_rate": 2.317716285164309e-06,
+      "loss": 0.5968,
+      "step": 10778
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4624412527903066,
+      "learning_rate": 2.316841622434586e-06,
+      "loss": 1.0353,
+      "step": 10779
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2279835841144213,
+      "learning_rate": 2.3159670750061563e-06,
+      "loss": 0.8648,
+      "step": 10780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9321279114729233,
+      "learning_rate": 2.3150926429165967e-06,
+      "loss": 0.6773,
+      "step": 10781
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3090064849556744,
+      "learning_rate": 2.3142183262034865e-06,
+      "loss": 0.8633,
+      "step": 10782
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9680505853967762,
+      "learning_rate": 2.3133441249043973e-06,
+      "loss": 0.6776,
+      "step": 10783
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2100809340722,
+      "learning_rate": 2.3124700390568945e-06,
+      "loss": 0.8125,
+      "step": 10784
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.813013025984094,
+      "learning_rate": 2.311596068698539e-06,
+      "loss": 0.744,
+      "step": 10785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.230345005501534,
+      "learning_rate": 2.3107222138668863e-06,
+      "loss": 0.7952,
+      "step": 10786
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.853718419669522,
+      "learning_rate": 2.3098484745994933e-06,
+      "loss": 0.9189,
+      "step": 10787
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.276953256966668,
+      "learning_rate": 2.3089748509339007e-06,
+      "loss": 0.7756,
+      "step": 10788
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4761975606544198,
+      "learning_rate": 2.30810134290765e-06,
+      "loss": 0.7561,
+      "step": 10789
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3394524336980234,
+      "learning_rate": 2.307227950558281e-06,
+      "loss": 0.904,
+      "step": 10790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.990155766297233,
+      "learning_rate": 2.306354673923325e-06,
+      "loss": 1.003,
+      "step": 10791
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3199976321558546,
+      "learning_rate": 2.3054815130403073e-06,
+      "loss": 0.7365,
+      "step": 10792
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.959500219099313,
+      "learning_rate": 2.3046084679467506e-06,
+      "loss": 0.9553,
+      "step": 10793
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8642713624803604,
+      "learning_rate": 2.3037355386801683e-06,
+      "loss": 0.5945,
+      "step": 10794
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1978011310398604,
+      "learning_rate": 2.3028627252780794e-06,
+      "loss": 0.6592,
+      "step": 10795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9156580255085802,
+      "learning_rate": 2.3019900277779816e-06,
+      "loss": 0.8669,
+      "step": 10796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.505463660187892,
+      "learning_rate": 2.3011174462173836e-06,
+      "loss": 0.6953,
+      "step": 10797
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.049228855252202,
+      "learning_rate": 2.3002449806337807e-06,
+      "loss": 0.7409,
+      "step": 10798
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0208779516354896,
+      "learning_rate": 2.2993726310646603e-06,
+      "loss": 0.6004,
+      "step": 10799
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6435827153838627,
+      "learning_rate": 2.2985003975475146e-06,
+      "loss": 0.8342,
+      "step": 10800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.473187509692016,
+      "learning_rate": 2.2976282801198237e-06,
+      "loss": 0.6978,
+      "step": 10801
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.345289408878594,
+      "learning_rate": 2.2967562788190633e-06,
+      "loss": 0.8569,
+      "step": 10802
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.137103779515772,
+      "learning_rate": 2.2958843936827064e-06,
+      "loss": 0.6451,
+      "step": 10803
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1957184983786053,
+      "learning_rate": 2.2950126247482178e-06,
+      "loss": 0.6278,
+      "step": 10804
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3506876590084476,
+      "learning_rate": 2.2941409720530642e-06,
+      "loss": 0.9563,
+      "step": 10805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.036217105782044,
+      "learning_rate": 2.2932694356346986e-06,
+      "loss": 0.8232,
+      "step": 10806
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1262870508704186,
+      "learning_rate": 2.292398015530571e-06,
+      "loss": 0.6188,
+      "step": 10807
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2966770003306287,
+      "learning_rate": 2.2915267117781328e-06,
+      "loss": 0.6782,
+      "step": 10808
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.544214720282642,
+      "learning_rate": 2.2906555244148233e-06,
+      "loss": 0.8943,
+      "step": 10809
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5028747811769683,
+      "learning_rate": 2.2897844534780805e-06,
+      "loss": 0.7887,
+      "step": 10810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8062622455257205,
+      "learning_rate": 2.2889134990053356e-06,
+      "loss": 0.7956,
+      "step": 10811
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1181279827857162,
+      "learning_rate": 2.2880426610340135e-06,
+      "loss": 0.8399,
+      "step": 10812
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9746341318570264,
+      "learning_rate": 2.2871719396015423e-06,
+      "loss": 0.6502,
+      "step": 10813
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.00959367841325,
+      "learning_rate": 2.2863013347453305e-06,
+      "loss": 0.8627,
+      "step": 10814
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.138323005197888,
+      "learning_rate": 2.2854308465027963e-06,
+      "loss": 0.8092,
+      "step": 10815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9708818023938666,
+      "learning_rate": 2.2845604749113457e-06,
+      "loss": 0.7059,
+      "step": 10816
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0184398024179389,
+      "learning_rate": 2.2836902200083762e-06,
+      "loss": 0.6334,
+      "step": 10817
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.315240989359035,
+      "learning_rate": 2.2828200818312886e-06,
+      "loss": 0.7849,
+      "step": 10818
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.89410335558128,
+      "learning_rate": 2.2819500604174733e-06,
+      "loss": 0.7811,
+      "step": 10819
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6260032257542196,
+      "learning_rate": 2.2810801558043177e-06,
+      "loss": 0.8768,
+      "step": 10820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.020939859512298,
+      "learning_rate": 2.2802103680292037e-06,
+      "loss": 0.7879,
+      "step": 10821
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.633318664245112,
+      "learning_rate": 2.279340697129505e-06,
+      "loss": 0.812,
+      "step": 10822
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8993924953980907,
+      "learning_rate": 2.2784711431426e-06,
+      "loss": 0.6473,
+      "step": 10823
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0902614776045483,
+      "learning_rate": 2.277601706105847e-06,
+      "loss": 0.771,
+      "step": 10824
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.6700162649991315,
+      "learning_rate": 2.2767323860566137e-06,
+      "loss": 0.7411,
+      "step": 10825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4793369230215223,
+      "learning_rate": 2.2758631830322546e-06,
+      "loss": 0.8172,
+      "step": 10826
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9303730395040501,
+      "learning_rate": 2.274994097070122e-06,
+      "loss": 0.7134,
+      "step": 10827
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9010573443928143,
+      "learning_rate": 2.2741251282075617e-06,
+      "loss": 0.8449,
+      "step": 10828
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1218780367475898,
+      "learning_rate": 2.2732562764819157e-06,
+      "loss": 0.7246,
+      "step": 10829
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3843401103047914,
+      "learning_rate": 2.272387541930518e-06,
+      "loss": 0.7522,
+      "step": 10830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3236029869608674,
+      "learning_rate": 2.2715189245907066e-06,
+      "loss": 0.6751,
+      "step": 10831
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.851092156944753,
+      "learning_rate": 2.2706504244997997e-06,
+      "loss": 0.7903,
+      "step": 10832
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.088794238518207,
+      "learning_rate": 2.269782041695124e-06,
+      "loss": 0.9174,
+      "step": 10833
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.612955929169195,
+      "learning_rate": 2.2689137762139952e-06,
+      "loss": 0.8379,
+      "step": 10834
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2130920778106598,
+      "learning_rate": 2.2680456280937234e-06,
+      "loss": 0.7882,
+      "step": 10835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4113085067791133,
+      "learning_rate": 2.267177597371616e-06,
+      "loss": 0.8828,
+      "step": 10836
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5135562377383884,
+      "learning_rate": 2.266309684084971e-06,
+      "loss": 0.6836,
+      "step": 10837
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5902346343816403,
+      "learning_rate": 2.265441888271091e-06,
+      "loss": 0.5525,
+      "step": 10838
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7996766476085786,
+      "learning_rate": 2.264574209967262e-06,
+      "loss": 0.7593,
+      "step": 10839
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1572479300302705,
+      "learning_rate": 2.263706649210769e-06,
+      "loss": 0.5971,
+      "step": 10840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.142366523319808,
+      "learning_rate": 2.2628392060388986e-06,
+      "loss": 0.6278,
+      "step": 10841
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.126685909822024,
+      "learning_rate": 2.2619718804889203e-06,
+      "loss": 0.894,
+      "step": 10842
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.119714898166214,
+      "learning_rate": 2.26110467259811e-06,
+      "loss": 0.9856,
+      "step": 10843
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.047244886159463,
+      "learning_rate": 2.260237582403732e-06,
+      "loss": 0.7284,
+      "step": 10844
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2578792673118757,
+      "learning_rate": 2.259370609943046e-06,
+      "loss": 0.5971,
+      "step": 10845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5458647802478525,
+      "learning_rate": 2.2585037552533097e-06,
+      "loss": 0.8764,
+      "step": 10846
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.849612917574347,
+      "learning_rate": 2.2576370183717704e-06,
+      "loss": 0.8538,
+      "step": 10847
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.257823376502365,
+      "learning_rate": 2.2567703993356775e-06,
+      "loss": 0.7371,
+      "step": 10848
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.100827634084691,
+      "learning_rate": 2.2559038981822724e-06,
+      "loss": 0.9348,
+      "step": 10849
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.068318486440059,
+      "learning_rate": 2.255037514948785e-06,
+      "loss": 0.8145,
+      "step": 10850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.816715967992071,
+      "learning_rate": 2.2541712496724495e-06,
+      "loss": 0.6754,
+      "step": 10851
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2050296540022547,
+      "learning_rate": 2.2533051023904913e-06,
+      "loss": 0.9286,
+      "step": 10852
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3265080498732265,
+      "learning_rate": 2.25243907314013e-06,
+      "loss": 0.8543,
+      "step": 10853
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4130696007350916,
+      "learning_rate": 2.2515731619585814e-06,
+      "loss": 0.6952,
+      "step": 10854
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1451418203218875,
+      "learning_rate": 2.250707368883052e-06,
+      "loss": 0.7053,
+      "step": 10855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.295762058882862,
+      "learning_rate": 2.249841693950755e-06,
+      "loss": 0.9833,
+      "step": 10856
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1348903509431052,
+      "learning_rate": 2.2489761371988826e-06,
+      "loss": 0.7109,
+      "step": 10857
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.100300643802928,
+      "learning_rate": 2.2481106986646314e-06,
+      "loss": 0.846,
+      "step": 10858
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.945843431712201,
+      "learning_rate": 2.247245378385195e-06,
+      "loss": 0.7885,
+      "step": 10859
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8111114810737958,
+      "learning_rate": 2.246380176397752e-06,
+      "loss": 0.7556,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.035302733520845,
+      "learning_rate": 2.245515092739488e-06,
+      "loss": 0.9135,
+      "step": 10861
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9859526038567337,
+      "learning_rate": 2.2446501274475747e-06,
+      "loss": 0.7968,
+      "step": 10862
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9851846194870039,
+      "learning_rate": 2.2437852805591815e-06,
+      "loss": 0.6082,
+      "step": 10863
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2895974698490877,
+      "learning_rate": 2.242920552111473e-06,
+      "loss": 0.8064,
+      "step": 10864
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5891240379891634,
+      "learning_rate": 2.2420559421416075e-06,
+      "loss": 0.8701,
+      "step": 10865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1170887159038876,
+      "learning_rate": 2.241191450686743e-06,
+      "loss": 0.7707,
+      "step": 10866
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9289956285534964,
+      "learning_rate": 2.2403270777840262e-06,
+      "loss": 0.9498,
+      "step": 10867
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7681888134089332,
+      "learning_rate": 2.239462823470598e-06,
+      "loss": 0.7619,
+      "step": 10868
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.856630258969752,
+      "learning_rate": 2.238598687783603e-06,
+      "loss": 0.8555,
+      "step": 10869
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6731392628087116,
+      "learning_rate": 2.237734670760172e-06,
+      "loss": 0.6773,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.892126337220215,
+      "learning_rate": 2.236870772437433e-06,
+      "loss": 0.7647,
+      "step": 10871
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8042288757242635,
+      "learning_rate": 2.2360069928525118e-06,
+      "loss": 0.7946,
+      "step": 10872
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.325787229222094,
+      "learning_rate": 2.2351433320425237e-06,
+      "loss": 0.8464,
+      "step": 10873
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.563208765520745,
+      "learning_rate": 2.234279790044588e-06,
+      "loss": 0.9342,
+      "step": 10874
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2166974061767495,
+      "learning_rate": 2.2334163668958055e-06,
+      "loss": 0.894,
+      "step": 10875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9683000175129601,
+      "learning_rate": 2.2325530626332852e-06,
+      "loss": 0.8326,
+      "step": 10876
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.513643472117826,
+      "learning_rate": 2.231689877294123e-06,
+      "loss": 0.8916,
+      "step": 10877
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0774518724867443,
+      "learning_rate": 2.2308268109154126e-06,
+      "loss": 0.607,
+      "step": 10878
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2307489131160456,
+      "learning_rate": 2.229963863534241e-06,
+      "loss": 0.6734,
+      "step": 10879
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2079116471443188,
+      "learning_rate": 2.229101035187692e-06,
+      "loss": 0.7749,
+      "step": 10880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2664733389073417,
+      "learning_rate": 2.2282383259128416e-06,
+      "loss": 0.8177,
+      "step": 10881
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2428786421857247,
+      "learning_rate": 2.227375735746764e-06,
+      "loss": 0.8257,
+      "step": 10882
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3285457491132813,
+      "learning_rate": 2.2265132647265243e-06,
+      "loss": 0.6369,
+      "step": 10883
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7457153773203475,
+      "learning_rate": 2.225650912889188e-06,
+      "loss": 0.8718,
+      "step": 10884
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.781572718627119,
+      "learning_rate": 2.224788680271811e-06,
+      "loss": 0.7596,
+      "step": 10885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1507549606418377,
+      "learning_rate": 2.223926566911445e-06,
+      "loss": 0.7084,
+      "step": 10886
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2751701135350206,
+      "learning_rate": 2.223064572845137e-06,
+      "loss": 0.7806,
+      "step": 10887
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9519316510819535,
+      "learning_rate": 2.222202698109929e-06,
+      "loss": 0.8633,
+      "step": 10888
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.411938368820051,
+      "learning_rate": 2.221340942742858e-06,
+      "loss": 0.9403,
+      "step": 10889
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7491331851245207,
+      "learning_rate": 2.2204793067809545e-06,
+      "loss": 0.9507,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.13392521381779,
+      "learning_rate": 2.219617790261244e-06,
+      "loss": 0.727,
+      "step": 10891
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.072510102120112,
+      "learning_rate": 2.218756393220753e-06,
+      "loss": 0.794,
+      "step": 10892
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3409266269296243,
+      "learning_rate": 2.2178951156964895e-06,
+      "loss": 0.693,
+      "step": 10893
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.573633035662361,
+      "learning_rate": 2.2170339577254714e-06,
+      "loss": 0.6493,
+      "step": 10894
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2316049271894136,
+      "learning_rate": 2.2161729193447024e-06,
+      "loss": 0.8384,
+      "step": 10895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.61022820273184,
+      "learning_rate": 2.2153120005911825e-06,
+      "loss": 0.7374,
+      "step": 10896
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0874685101594164,
+      "learning_rate": 2.214451201501908e-06,
+      "loss": 0.8709,
+      "step": 10897
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.536204664972832,
+      "learning_rate": 2.213590522113867e-06,
+      "loss": 0.8914,
+      "step": 10898
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8553590801625544,
+      "learning_rate": 2.212729962464051e-06,
+      "loss": 0.8856,
+      "step": 10899
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.200655845404194,
+      "learning_rate": 2.2118695225894343e-06,
+      "loss": 0.7731,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2049683110148455,
+      "learning_rate": 2.2110092025269915e-06,
+      "loss": 0.6978,
+      "step": 10901
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8054205496048024,
+      "learning_rate": 2.210149002313697e-06,
+      "loss": 0.8408,
+      "step": 10902
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7926469600741797,
+      "learning_rate": 2.2092889219865127e-06,
+      "loss": 0.8112,
+      "step": 10903
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4129103007420865,
+      "learning_rate": 2.208428961582399e-06,
+      "loss": 1.0997,
+      "step": 10904
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5024880425526885,
+      "learning_rate": 2.2075691211383094e-06,
+      "loss": 0.9239,
+      "step": 10905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.886153402223502,
+      "learning_rate": 2.2067094006911943e-06,
+      "loss": 0.8649,
+      "step": 10906
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5998307266111063,
+      "learning_rate": 2.205849800277997e-06,
+      "loss": 0.7657,
+      "step": 10907
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2311745897929876,
+      "learning_rate": 2.204990319935657e-06,
+      "loss": 0.7546,
+      "step": 10908
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0445159984275056,
+      "learning_rate": 2.2041309597011057e-06,
+      "loss": 0.8206,
+      "step": 10909
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.623054071911052,
+      "learning_rate": 2.2032717196112776e-06,
+      "loss": 0.876,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359254561755377,
+      "learning_rate": 2.2024125997030885e-06,
+      "loss": 0.8024,
+      "step": 10911
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3057855339715436,
+      "learning_rate": 2.2015536000134624e-06,
+      "loss": 0.7512,
+      "step": 10912
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3155042594581525,
+      "learning_rate": 2.2006947205793107e-06,
+      "loss": 0.9157,
+      "step": 10913
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.218069573339502,
+      "learning_rate": 2.1998359614375412e-06,
+      "loss": 0.9114,
+      "step": 10914
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2902351389651523,
+      "learning_rate": 2.198977322625057e-06,
+      "loss": 0.6423,
+      "step": 10915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0900093429486732,
+      "learning_rate": 2.1981188041787534e-06,
+      "loss": 0.9196,
+      "step": 10916
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2407340246082241,
+      "learning_rate": 2.1972604061355285e-06,
+      "loss": 0.6419,
+      "step": 10917
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.233762741689998,
+      "learning_rate": 2.196402128532264e-06,
+      "loss": 0.8319,
+      "step": 10918
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.51968200839194,
+      "learning_rate": 2.1955439714058422e-06,
+      "loss": 0.805,
+      "step": 10919
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2897023251890154,
+      "learning_rate": 2.1946859347931442e-06,
+      "loss": 0.8985,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6815984718681194,
+      "learning_rate": 2.193828018731039e-06,
+      "loss": 0.794,
+      "step": 10921
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8700662886133013,
+      "learning_rate": 2.192970223256393e-06,
+      "loss": 0.8154,
+      "step": 10922
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1569959916082184,
+      "learning_rate": 2.1921125484060692e-06,
+      "loss": 0.7117,
+      "step": 10923
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9865353255058769,
+      "learning_rate": 2.191254994216922e-06,
+      "loss": 0.7901,
+      "step": 10924
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8781055363670873,
+      "learning_rate": 2.1903975607258025e-06,
+      "loss": 0.8198,
+      "step": 10925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.615824789710484,
+      "learning_rate": 2.1895402479695566e-06,
+      "loss": 0.8355,
+      "step": 10926
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2074801920921705,
+      "learning_rate": 2.1886830559850264e-06,
+      "loss": 0.843,
+      "step": 10927
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9792109713537565,
+      "learning_rate": 2.187825984809049e-06,
+      "loss": 0.9041,
+      "step": 10928
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.051337685309214,
+      "learning_rate": 2.186969034478448e-06,
+      "loss": 0.789,
+      "step": 10929
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.084369557789357,
+      "learning_rate": 2.186112205030054e-06,
+      "loss": 0.7939,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0189786581956766,
+      "learning_rate": 2.1852554965006854e-06,
+      "loss": 0.5946,
+      "step": 10931
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2090016306273443,
+      "learning_rate": 2.1843989089271572e-06,
+      "loss": 0.7941,
+      "step": 10932
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0073868436878426,
+      "learning_rate": 2.183542442346279e-06,
+      "loss": 0.6688,
+      "step": 10933
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8292933009062715,
+      "learning_rate": 2.182686096794852e-06,
+      "loss": 0.7189,
+      "step": 10934
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0306770331854078,
+      "learning_rate": 2.181829872309682e-06,
+      "loss": 0.6437,
+      "step": 10935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9664853289315398,
+      "learning_rate": 2.180973768927555e-06,
+      "loss": 0.6195,
+      "step": 10936
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0369381826265704,
+      "learning_rate": 2.1801177866852658e-06,
+      "loss": 0.7608,
+      "step": 10937
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0884992046505935,
+      "learning_rate": 2.179261925619595e-06,
+      "loss": 0.859,
+      "step": 10938
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.083903367160716,
+      "learning_rate": 2.1784061857673217e-06,
+      "loss": 0.7328,
+      "step": 10939
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0914820102152827,
+      "learning_rate": 2.177550567165218e-06,
+      "loss": 0.727,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2814072451211613,
+      "learning_rate": 2.176695069850053e-06,
+      "loss": 0.816,
+      "step": 10941
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.460968874620374,
+      "learning_rate": 2.1758396938585885e-06,
+      "loss": 0.8139,
+      "step": 10942
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.033492452666567,
+      "learning_rate": 2.174984439227582e-06,
+      "loss": 0.7143,
+      "step": 10943
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.564374932489983,
+      "learning_rate": 2.174129305993784e-06,
+      "loss": 0.7948,
+      "step": 10944
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1731106286087334,
+      "learning_rate": 2.173274294193945e-06,
+      "loss": 0.8515,
+      "step": 10945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1662236702716933,
+      "learning_rate": 2.1724194038648057e-06,
+      "loss": 0.709,
+      "step": 10946
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6161469981153305,
+      "learning_rate": 2.1715646350431015e-06,
+      "loss": 0.8046,
+      "step": 10947
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0674676803234133,
+      "learning_rate": 2.1707099877655634e-06,
+      "loss": 0.9332,
+      "step": 10948
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.483352425793532,
+      "learning_rate": 2.1698554620689178e-06,
+      "loss": 0.8229,
+      "step": 10949
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0813911947697115,
+      "learning_rate": 2.169001057989889e-06,
+      "loss": 0.7448,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1916408320798961,
+      "learning_rate": 2.1681467755651882e-06,
+      "loss": 0.746,
+      "step": 10951
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.086632350675895,
+      "learning_rate": 2.1672926148315247e-06,
+      "loss": 0.6201,
+      "step": 10952
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.43318536182062,
+      "learning_rate": 2.1664385758256106e-06,
+      "loss": 0.8652,
+      "step": 10953
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.778180782449322,
+      "learning_rate": 2.165584658584138e-06,
+      "loss": 0.7204,
+      "step": 10954
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.06212169521181,
+      "learning_rate": 2.1647308631438068e-06,
+      "loss": 0.8326,
+      "step": 10955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.309623243418074,
+      "learning_rate": 2.163877189541305e-06,
+      "loss": 0.8463,
+      "step": 10956
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.058767346472798,
+      "learning_rate": 2.1630236378133168e-06,
+      "loss": 0.8652,
+      "step": 10957
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3973358695017755,
+      "learning_rate": 2.1621702079965214e-06,
+      "loss": 0.9513,
+      "step": 10958
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.002934166738737,
+      "learning_rate": 2.16131690012759e-06,
+      "loss": 0.7114,
+      "step": 10959
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8349576398234022,
+      "learning_rate": 2.1604637142431978e-06,
+      "loss": 0.6379,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0291588576306423,
+      "learning_rate": 2.1596106503800013e-06,
+      "loss": 0.6674,
+      "step": 10961
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.024321342652837,
+      "learning_rate": 2.1587577085746596e-06,
+      "loss": 0.7356,
+      "step": 10962
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9824404739945085,
+      "learning_rate": 2.157904888863829e-06,
+      "loss": 0.8945,
+      "step": 10963
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9815252201999465,
+      "learning_rate": 2.157052191284154e-06,
+      "loss": 0.9456,
+      "step": 10964
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9819574269773024,
+      "learning_rate": 2.1561996158722785e-06,
+      "loss": 0.6237,
+      "step": 10965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7731752096770586,
+      "learning_rate": 2.1553471626648392e-06,
+      "loss": 0.7723,
+      "step": 10966
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8845382841396434,
+      "learning_rate": 2.1544948316984658e-06,
+      "loss": 0.773,
+      "step": 10967
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0675435759855434,
+      "learning_rate": 2.1536426230097906e-06,
+      "loss": 0.7023,
+      "step": 10968
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2273820366981674,
+      "learning_rate": 2.1527905366354292e-06,
+      "loss": 0.7218,
+      "step": 10969
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0685549992381107,
+      "learning_rate": 2.1519385726119972e-06,
+      "loss": 0.7925,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.572589661692582,
+      "learning_rate": 2.151086730976112e-06,
+      "loss": 0.8882,
+      "step": 10971
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0292245620900466,
+      "learning_rate": 2.1502350117643704e-06,
+      "loss": 0.7536,
+      "step": 10972
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.17431282753668,
+      "learning_rate": 2.1493834150133795e-06,
+      "loss": 0.798,
+      "step": 10973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.233646446624618,
+      "learning_rate": 2.1485319407597315e-06,
+      "loss": 0.7852,
+      "step": 10974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.01488174054106,
+      "learning_rate": 2.1476805890400172e-06,
+      "loss": 0.6553,
+      "step": 10975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.946557639961911,
+      "learning_rate": 2.14682935989082e-06,
+      "loss": 0.8699,
+      "step": 10976
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1164005712499443,
+      "learning_rate": 2.1459782533487176e-06,
+      "loss": 0.7423,
+      "step": 10977
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.269784322817972,
+      "learning_rate": 2.14512726945029e-06,
+      "loss": 0.7964,
+      "step": 10978
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.744931565568374,
+      "learning_rate": 2.1442764082321e-06,
+      "loss": 0.8301,
+      "step": 10979
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7974149987888155,
+      "learning_rate": 2.1434256697307103e-06,
+      "loss": 0.8446,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3616651485623588,
+      "learning_rate": 2.142575053982684e-06,
+      "loss": 0.7204,
+      "step": 10981
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9669025843798453,
+      "learning_rate": 2.1417245610245714e-06,
+      "loss": 0.7062,
+      "step": 10982
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6310394134994315,
+      "learning_rate": 2.14087419089292e-06,
+      "loss": 0.6667,
+      "step": 10983
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8875930230772933,
+      "learning_rate": 2.140023943624272e-06,
+      "loss": 0.7522,
+      "step": 10984
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4686689382057803,
+      "learning_rate": 2.139173819255163e-06,
+      "loss": 0.7223,
+      "step": 10985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.286383481814569,
+      "learning_rate": 2.13832381782213e-06,
+      "loss": 0.9452,
+      "step": 10986
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.35324706326343,
+      "learning_rate": 2.1374739393616923e-06,
+      "loss": 0.8295,
+      "step": 10987
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.697290890805567,
+      "learning_rate": 2.1366241839103764e-06,
+      "loss": 0.8213,
+      "step": 10988
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.146240765644824,
+      "learning_rate": 2.135774551504698e-06,
+      "loss": 0.7318,
+      "step": 10989
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1662410199682314,
+      "learning_rate": 2.1349250421811622e-06,
+      "loss": 0.5678,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.896791116926673,
+      "learning_rate": 2.1340756559762808e-06,
+      "loss": 0.7122,
+      "step": 10991
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.35936228900579,
+      "learning_rate": 2.1332263929265505e-06,
+      "loss": 0.8302,
+      "step": 10992
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.291775117305864,
+      "learning_rate": 2.132377253068467e-06,
+      "loss": 0.7654,
+      "step": 10993
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6000575603647222,
+      "learning_rate": 2.1315282364385197e-06,
+      "loss": 0.8634,
+      "step": 10994
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3116674438838376,
+      "learning_rate": 2.1306793430731908e-06,
+      "loss": 0.5808,
+      "step": 10995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4230216176709654,
+      "learning_rate": 2.1298305730089626e-06,
+      "loss": 0.7164,
+      "step": 10996
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2177489504436718,
+      "learning_rate": 2.1289819262823065e-06,
+      "loss": 0.6621,
+      "step": 10997
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1606679828345428,
+      "learning_rate": 2.1281334029296913e-06,
+      "loss": 0.6789,
+      "step": 10998
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.413642514874227,
+      "learning_rate": 2.1272850029875802e-06,
+      "loss": 0.7633,
+      "step": 10999
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.004326136175185,
+      "learning_rate": 2.1264367264924295e-06,
+      "loss": 0.6896,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2127374676573015,
+      "learning_rate": 2.125588573480693e-06,
+      "loss": 0.8182,
+      "step": 11001
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9995289262440816,
+      "learning_rate": 2.1247405439888173e-06,
+      "loss": 0.8466,
+      "step": 11002
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0070124165910355,
+      "learning_rate": 2.1238926380532423e-06,
+      "loss": 0.7171,
+      "step": 11003
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0997401470483292,
+      "learning_rate": 2.1230448557104087e-06,
+      "loss": 0.744,
+      "step": 11004
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9581162507988743,
+      "learning_rate": 2.1221971969967424e-06,
+      "loss": 0.8038,
+      "step": 11005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.181138612001788,
+      "learning_rate": 2.121349661948673e-06,
+      "loss": 1.0606,
+      "step": 11006
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0297417898880803,
+      "learning_rate": 2.12050225060262e-06,
+      "loss": 0.6141,
+      "step": 11007
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.512219167637152,
+      "learning_rate": 2.119654962994998e-06,
+      "loss": 0.8638,
+      "step": 11008
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.192730608789291,
+      "learning_rate": 2.1188077991622174e-06,
+      "loss": 0.7073,
+      "step": 11009
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2210515171029068,
+      "learning_rate": 2.1179607591406807e-06,
+      "loss": 0.6795,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2469279556531276,
+      "learning_rate": 2.117113842966792e-06,
+      "loss": 0.8226,
+      "step": 11011
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2385872023703972,
+      "learning_rate": 2.1162670506769408e-06,
+      "loss": 0.6333,
+      "step": 11012
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4839952577251223,
+      "learning_rate": 2.1154203823075146e-06,
+      "loss": 0.6825,
+      "step": 11013
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0827413741144796,
+      "learning_rate": 2.1145738378949004e-06,
+      "loss": 0.601,
+      "step": 11014
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3688061054506475,
+      "learning_rate": 2.113727417475475e-06,
+      "loss": 0.8278,
+      "step": 11015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2091452005676884,
+      "learning_rate": 2.112881121085611e-06,
+      "loss": 0.6666,
+      "step": 11016
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1495669055844795,
+      "learning_rate": 2.112034948761674e-06,
+      "loss": 0.7379,
+      "step": 11017
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.389700629848623,
+      "learning_rate": 2.111188900540028e-06,
+      "loss": 0.6493,
+      "step": 11018
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1010742795379838,
+      "learning_rate": 2.110342976457029e-06,
+      "loss": 0.7494,
+      "step": 11019
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.102990428711384,
+      "learning_rate": 2.1094971765490266e-06,
+      "loss": 0.7758,
+      "step": 11020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6751397891377433,
+      "learning_rate": 2.108651500852367e-06,
+      "loss": 0.7698,
+      "step": 11021
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4316147571836617,
+      "learning_rate": 2.107805949403396e-06,
+      "loss": 0.7764,
+      "step": 11022
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9606831259672037,
+      "learning_rate": 2.10696052223844e-06,
+      "loss": 0.718,
+      "step": 11023
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.487142205933848,
+      "learning_rate": 2.1061152193938355e-06,
+      "loss": 0.8319,
+      "step": 11024
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7122897635258496,
+      "learning_rate": 2.1052700409059057e-06,
+      "loss": 0.7379,
+      "step": 11025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.262729746215259,
+      "learning_rate": 2.1044249868109683e-06,
+      "loss": 0.8558,
+      "step": 11026
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.074440646593186,
+      "learning_rate": 2.1035800571453386e-06,
+      "loss": 0.8759,
+      "step": 11027
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.343028185463489,
+      "learning_rate": 2.1027352519453225e-06,
+      "loss": 0.6414,
+      "step": 11028
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0715678206387658,
+      "learning_rate": 2.1018905712472285e-06,
+      "loss": 0.7882,
+      "step": 11029
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4008715072236138,
+      "learning_rate": 2.1010460150873495e-06,
+      "loss": 0.7899,
+      "step": 11030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6828605619763857,
+      "learning_rate": 2.1002015835019777e-06,
+      "loss": 0.8165,
+      "step": 11031
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.623345457582111,
+      "learning_rate": 2.0993572765274044e-06,
+      "loss": 0.8926,
+      "step": 11032
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4666051819699306,
+      "learning_rate": 2.098513094199909e-06,
+      "loss": 0.8413,
+      "step": 11033
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9408638354446213,
+      "learning_rate": 2.0976690365557673e-06,
+      "loss": 0.6715,
+      "step": 11034
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8304776311083875,
+      "learning_rate": 2.096825103631251e-06,
+      "loss": 0.8121,
+      "step": 11035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0433760953878837,
+      "learning_rate": 2.0959812954626256e-06,
+      "loss": 0.6999,
+      "step": 11036
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2060671389451763,
+      "learning_rate": 2.0951376120861523e-06,
+      "loss": 0.6601,
+      "step": 11037
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.106206437353681,
+      "learning_rate": 2.0942940535380823e-06,
+      "loss": 0.6843,
+      "step": 11038
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5673605606222476,
+      "learning_rate": 2.093450619854671e-06,
+      "loss": 0.8717,
+      "step": 11039
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1241164719225347,
+      "learning_rate": 2.0926073110721613e-06,
+      "loss": 0.6317,
+      "step": 11040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1083745308895114,
+      "learning_rate": 2.0917641272267867e-06,
+      "loss": 0.7745,
+      "step": 11041
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.218289412876038,
+      "learning_rate": 2.0909210683547865e-06,
+      "loss": 0.6077,
+      "step": 11042
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9295581277836187,
+      "learning_rate": 2.0900781344923872e-06,
+      "loss": 0.778,
+      "step": 11043
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.760649540377793,
+      "learning_rate": 2.0892353256758107e-06,
+      "loss": 0.8313,
+      "step": 11044
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.279327126507535,
+      "learning_rate": 2.0883926419412748e-06,
+      "loss": 0.6726,
+      "step": 11045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2437992563701323,
+      "learning_rate": 2.08755008332499e-06,
+      "loss": 0.864,
+      "step": 11046
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.248394203203067,
+      "learning_rate": 2.0867076498631685e-06,
+      "loss": 0.6861,
+      "step": 11047
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7701017374753367,
+      "learning_rate": 2.0858653415920034e-06,
+      "loss": 0.6895,
+      "step": 11048
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7313959050722847,
+      "learning_rate": 2.0850231585476965e-06,
+      "loss": 0.8354,
+      "step": 11049
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5105720990204685,
+      "learning_rate": 2.084181100766437e-06,
+      "loss": 0.6767,
+      "step": 11050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.419254464643559,
+      "learning_rate": 2.083339168284409e-06,
+      "loss": 0.7806,
+      "step": 11051
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.352323310760641,
+      "learning_rate": 2.0824973611377924e-06,
+      "loss": 0.6351,
+      "step": 11052
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.394797347359975,
+      "learning_rate": 2.0816556793627624e-06,
+      "loss": 0.8264,
+      "step": 11053
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4343683727062517,
+      "learning_rate": 2.0808141229954876e-06,
+      "loss": 0.6805,
+      "step": 11054
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.172030045613175,
+      "learning_rate": 2.079972692072131e-06,
+      "loss": 0.9278,
+      "step": 11055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9368781283785546,
+      "learning_rate": 2.0791313866288485e-06,
+      "loss": 0.973,
+      "step": 11056
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9280009051443925,
+      "learning_rate": 2.078290206701798e-06,
+      "loss": 0.9088,
+      "step": 11057
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.202828708858704,
+      "learning_rate": 2.0774491523271243e-06,
+      "loss": 0.977,
+      "step": 11058
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.799114230654783,
+      "learning_rate": 2.0766082235409695e-06,
+      "loss": 0.8045,
+      "step": 11059
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.228600448367717,
+      "learning_rate": 2.0757674203794696e-06,
+      "loss": 0.7092,
+      "step": 11060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.093745260004982,
+      "learning_rate": 2.074926742878757e-06,
+      "loss": 0.7605,
+      "step": 11061
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4921463042000167,
+      "learning_rate": 2.074086191074956e-06,
+      "loss": 0.5115,
+      "step": 11062
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1234580726445544,
+      "learning_rate": 2.0732457650041884e-06,
+      "loss": 0.7346,
+      "step": 11063
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2515490660227395,
+      "learning_rate": 2.072405464702566e-06,
+      "loss": 0.7466,
+      "step": 11064
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1202436158954843,
+      "learning_rate": 2.071565290206205e-06,
+      "loss": 0.822,
+      "step": 11065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6487110064008097,
+      "learning_rate": 2.0707252415512015e-06,
+      "loss": 0.8664,
+      "step": 11066
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9350433743970323,
+      "learning_rate": 2.06988531877366e-06,
+      "loss": 0.8997,
+      "step": 11067
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.19661436727268,
+      "learning_rate": 2.069045521909672e-06,
+      "loss": 0.7718,
+      "step": 11068
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3476314578164974,
+      "learning_rate": 2.068205850995326e-06,
+      "loss": 0.8019,
+      "step": 11069
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.856017816733232,
+      "learning_rate": 2.067366306066704e-06,
+      "loss": 0.8704,
+      "step": 11070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.672281148974109,
+      "learning_rate": 2.0665268871598832e-06,
+      "loss": 0.885,
+      "step": 11071
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1203371263657975,
+      "learning_rate": 2.065687594310935e-06,
+      "loss": 0.5639,
+      "step": 11072
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4968280935813962,
+      "learning_rate": 2.064848427555926e-06,
+      "loss": 0.6829,
+      "step": 11073
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1387246571028955,
+      "learning_rate": 2.064009386930915e-06,
+      "loss": 0.8184,
+      "step": 11074
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2888050047537094,
+      "learning_rate": 2.0631704724719613e-06,
+      "loss": 0.9693,
+      "step": 11075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5117502652147055,
+      "learning_rate": 2.0623316842151132e-06,
+      "loss": 0.6472,
+      "step": 11076
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4315290904276763,
+      "learning_rate": 2.061493022196415e-06,
+      "loss": 0.8506,
+      "step": 11077
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8877946647461243,
+      "learning_rate": 2.0606544864519064e-06,
+      "loss": 0.657,
+      "step": 11078
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9199554134132295,
+      "learning_rate": 2.0598160770176208e-06,
+      "loss": 0.9944,
+      "step": 11079
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0765662655111077,
+      "learning_rate": 2.0589777939295855e-06,
+      "loss": 0.582,
+      "step": 11080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4339440206014675,
+      "learning_rate": 2.0581396372238254e-06,
+      "loss": 1.0055,
+      "step": 11081
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.91361476535108,
+      "learning_rate": 2.0573016069363554e-06,
+      "loss": 0.9114,
+      "step": 11082
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6647894326153905,
+      "learning_rate": 2.056463703103192e-06,
+      "loss": 0.8605,
+      "step": 11083
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2463637902340836,
+      "learning_rate": 2.0556259257603355e-06,
+      "loss": 0.7139,
+      "step": 11084
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.103148611859987,
+      "learning_rate": 2.0547882749437923e-06,
+      "loss": 0.8997,
+      "step": 11085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6645453409481097,
+      "learning_rate": 2.053950750689556e-06,
+      "loss": 0.9925,
+      "step": 11086
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1979825661478405,
+      "learning_rate": 2.053113353033618e-06,
+      "loss": 0.7368,
+      "step": 11087
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8434063625887787,
+      "learning_rate": 2.0522760820119615e-06,
+      "loss": 0.8822,
+      "step": 11088
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.149571480607861,
+      "learning_rate": 2.0514389376605646e-06,
+      "loss": 0.6354,
+      "step": 11089
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.023031662226296,
+      "learning_rate": 2.0506019200154076e-06,
+      "loss": 0.6891,
+      "step": 11090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.488455768821655,
+      "learning_rate": 2.049765029112453e-06,
+      "loss": 0.6735,
+      "step": 11091
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.004247739973904,
+      "learning_rate": 2.0489282649876633e-06,
+      "loss": 0.7927,
+      "step": 11092
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.850149795103478,
+      "learning_rate": 2.0480916276770005e-06,
+      "loss": 0.9494,
+      "step": 11093
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9937371537114972,
+      "learning_rate": 2.0472551172164152e-06,
+      "loss": 0.7699,
+      "step": 11094
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4426110713438187,
+      "learning_rate": 2.046418733641853e-06,
+      "loss": 0.8912,
+      "step": 11095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0486806854619175,
+      "learning_rate": 2.0455824769892556e-06,
+      "loss": 0.7755,
+      "step": 11096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.047676233594313,
+      "learning_rate": 2.0447463472945593e-06,
+      "loss": 0.9305,
+      "step": 11097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.775837859695019,
+      "learning_rate": 2.0439103445936946e-06,
+      "loss": 0.7547,
+      "step": 11098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.312000334697641,
+      "learning_rate": 2.0430744689225833e-06,
+      "loss": 0.9226,
+      "step": 11099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3194888458463554,
+      "learning_rate": 2.0422387203171503e-06,
+      "loss": 0.7559,
+      "step": 11100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1479681429868658,
+      "learning_rate": 2.0414030988133076e-06,
+      "loss": 0.6693,
+      "step": 11101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.467753164289959,
+      "learning_rate": 2.04056760444696e-06,
+      "loss": 0.809,
+      "step": 11102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9811398016156792,
+      "learning_rate": 2.039732237254015e-06,
+      "loss": 0.7989,
+      "step": 11103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0828733872158227,
+      "learning_rate": 2.0388969972703688e-06,
+      "loss": 0.8728,
+      "step": 11104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2716478973384187,
+      "learning_rate": 2.0380618845319132e-06,
+      "loss": 0.7067,
+      "step": 11105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0891875477036894,
+      "learning_rate": 2.0372268990745347e-06,
+      "loss": 0.6384,
+      "step": 11106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8233374294362457,
+      "learning_rate": 2.036392040934114e-06,
+      "loss": 0.9521,
+      "step": 11107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8028730149208334,
+      "learning_rate": 2.03555731014653e-06,
+      "loss": 0.7111,
+      "step": 11108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3202256204205005,
+      "learning_rate": 2.0347227067476478e-06,
+      "loss": 0.7056,
+      "step": 11109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.238829029014054,
+      "learning_rate": 2.033888230773336e-06,
+      "loss": 0.6343,
+      "step": 11110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.070647053624805,
+      "learning_rate": 2.0330538822594532e-06,
+      "loss": 0.6619,
+      "step": 11111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.46153557156174,
+      "learning_rate": 2.0322196612418527e-06,
+      "loss": 0.7016,
+      "step": 11112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0297844769175444,
+      "learning_rate": 2.031385567756383e-06,
+      "loss": 0.9513,
+      "step": 11113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.442468632238764,
+      "learning_rate": 2.030551601838887e-06,
+      "loss": 0.8653,
+      "step": 11114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.092450790155816,
+      "learning_rate": 2.029717763525202e-06,
+      "loss": 0.667,
+      "step": 11115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1501064185487615,
+      "learning_rate": 2.02888405285116e-06,
+      "loss": 0.7874,
+      "step": 11116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8753788488625314,
+      "learning_rate": 2.0280504698525856e-06,
+      "loss": 0.9595,
+      "step": 11117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9109979982382952,
+      "learning_rate": 2.0272170145653025e-06,
+      "loss": 0.7753,
+      "step": 11118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.132186147297626,
+      "learning_rate": 2.0263836870251277e-06,
+      "loss": 0.6806,
+      "step": 11119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9190463662312407,
+      "learning_rate": 2.025550487267864e-06,
+      "loss": 0.8023,
+      "step": 11120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.083697376273296,
+      "learning_rate": 2.0247174153293226e-06,
+      "loss": 0.9773,
+      "step": 11121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.275129470552053,
+      "learning_rate": 2.0238844712453e-06,
+      "loss": 0.8363,
+      "step": 11122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8003983395098184,
+      "learning_rate": 2.02305165505159e-06,
+      "loss": 0.768,
+      "step": 11123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5005482089909745,
+      "learning_rate": 2.0222189667839805e-06,
+      "loss": 0.8425,
+      "step": 11124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0432436411249806,
+      "learning_rate": 2.0213864064782516e-06,
+      "loss": 0.7494,
+      "step": 11125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1277372723769445,
+      "learning_rate": 2.020553974170186e-06,
+      "loss": 0.666,
+      "step": 11126
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.012062985352448,
+      "learning_rate": 2.019721669895549e-06,
+      "loss": 0.8392,
+      "step": 11127
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.000137817976671,
+      "learning_rate": 2.018889493690111e-06,
+      "loss": 0.719,
+      "step": 11128
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2589122798812222,
+      "learning_rate": 2.01805744558963e-06,
+      "loss": 0.8491,
+      "step": 11129
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0639176826059824,
+      "learning_rate": 2.0172255256298623e-06,
+      "loss": 0.6138,
+      "step": 11130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0672556133168416,
+      "learning_rate": 2.016393733846557e-06,
+      "loss": 0.9713,
+      "step": 11131
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9116244452372868,
+      "learning_rate": 2.015562070275457e-06,
+      "loss": 0.7582,
+      "step": 11132
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.050596280477637,
+      "learning_rate": 2.0147305349523e-06,
+      "loss": 0.7005,
+      "step": 11133
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0248387416229554,
+      "learning_rate": 2.013899127912824e-06,
+      "loss": 0.7489,
+      "step": 11134
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4086784679864603,
+      "learning_rate": 2.0130678491927495e-06,
+      "loss": 0.5693,
+      "step": 11135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.628152288947109,
+      "learning_rate": 2.0122366988278035e-06,
+      "loss": 1.0138,
+      "step": 11136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.023157497141381,
+      "learning_rate": 2.0114056768537005e-06,
+      "loss": 0.7154,
+      "step": 11137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.339120717706341,
+      "learning_rate": 2.0105747833061516e-06,
+      "loss": 0.867,
+      "step": 11138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.908689824201288,
+      "learning_rate": 2.009744018220863e-06,
+      "loss": 0.7082,
+      "step": 11139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.105015960048751,
+      "learning_rate": 2.0089133816335303e-06,
+      "loss": 0.7207,
+      "step": 11140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.320437276618764,
+      "learning_rate": 2.0080828735798554e-06,
+      "loss": 0.7903,
+      "step": 11141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.060312948335172,
+      "learning_rate": 2.0072524940955213e-06,
+      "loss": 0.7279,
+      "step": 11142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4008618120872587,
+      "learning_rate": 2.00642224321621e-06,
+      "loss": 0.8559,
+      "step": 11143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.43124878512629,
+      "learning_rate": 2.005592120977606e-06,
+      "loss": 0.8929,
+      "step": 11144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2804786354602062,
+      "learning_rate": 2.0047621274153746e-06,
+      "loss": 0.9123,
+      "step": 11145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4970845751165838,
+      "learning_rate": 2.0039322625651865e-06,
+      "loss": 0.8607,
+      "step": 11146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1607986667997128,
+      "learning_rate": 2.0031025264627015e-06,
+      "loss": 0.7266,
+      "step": 11147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.112611708322419,
+      "learning_rate": 2.002272919143576e-06,
+      "loss": 0.8076,
+      "step": 11148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0723509110753937,
+      "learning_rate": 2.0014434406434584e-06,
+      "loss": 0.7303,
+      "step": 11149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.073241825388482,
+      "learning_rate": 2.0006140909979937e-06,
+      "loss": 0.8759,
+      "step": 11150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1083289492012336,
+      "learning_rate": 1.9997848702428226e-06,
+      "loss": 0.6682,
+      "step": 11151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1973651242184578,
+      "learning_rate": 1.9989557784135792e-06,
+      "loss": 0.8609,
+      "step": 11152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015160866366085,
+      "learning_rate": 1.9981268155458864e-06,
+      "loss": 0.776,
+      "step": 11153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.32745616435002,
+      "learning_rate": 1.9972979816753717e-06,
+      "loss": 0.8917,
+      "step": 11154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3240113821997084,
+      "learning_rate": 1.9964692768376497e-06,
+      "loss": 0.7975,
+      "step": 11155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8876548245258173,
+      "learning_rate": 1.9956407010683324e-06,
+      "loss": 0.7784,
+      "step": 11156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9162141279629903,
+      "learning_rate": 1.994812254403025e-06,
+      "loss": 0.6916,
+      "step": 11157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.345649387544351,
+      "learning_rate": 1.9939839368773267e-06,
+      "loss": 0.6484,
+      "step": 11158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.627196404191098,
+      "learning_rate": 1.9931557485268365e-06,
+      "loss": 0.7274,
+      "step": 11159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.883217479507767,
+      "learning_rate": 1.992327689387137e-06,
+      "loss": 0.6822,
+      "step": 11160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.066978327687895,
+      "learning_rate": 1.991499759493817e-06,
+      "loss": 0.8722,
+      "step": 11161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09149664595904,
+      "learning_rate": 1.990671958882454e-06,
+      "loss": 0.8458,
+      "step": 11162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9833863458683507,
+      "learning_rate": 1.9898442875886157e-06,
+      "loss": 0.7223,
+      "step": 11163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9206856922305764,
+      "learning_rate": 1.9890167456478748e-06,
+      "loss": 0.8696,
+      "step": 11164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.883339154597117,
+      "learning_rate": 1.9881893330957893e-06,
+      "loss": 0.6302,
+      "step": 11165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4067632361660616,
+      "learning_rate": 1.9873620499679157e-06,
+      "loss": 0.933,
+      "step": 11166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.386387029192385,
+      "learning_rate": 1.986534896299805e-06,
+      "loss": 0.7291,
+      "step": 11167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.075809293398219,
+      "learning_rate": 1.985707872126999e-06,
+      "loss": 0.9959,
+      "step": 11168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7680171781230325,
+      "learning_rate": 1.984880977485041e-06,
+      "loss": 0.827,
+      "step": 11169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.584497276678841,
+      "learning_rate": 1.9840542124094635e-06,
+      "loss": 0.6348,
+      "step": 11170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.101050115399943,
+      "learning_rate": 1.9832275769357906e-06,
+      "loss": 0.7643,
+      "step": 11171
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9305334530777167,
+      "learning_rate": 1.982401071099549e-06,
+      "loss": 0.6417,
+      "step": 11172
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.38781218974834,
+      "learning_rate": 1.9815746949362536e-06,
+      "loss": 0.7523,
+      "step": 11173
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5914617902831123,
+      "learning_rate": 1.980748448481416e-06,
+      "loss": 0.846,
+      "step": 11174
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9424201588745955,
+      "learning_rate": 1.9799223317705426e-06,
+      "loss": 0.7097,
+      "step": 11175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.86526982700524,
+      "learning_rate": 1.97909634483913e-06,
+      "loss": 0.9301,
+      "step": 11176
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1129612643483546,
+      "learning_rate": 1.9782704877226796e-06,
+      "loss": 0.9401,
+      "step": 11177
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.424010076986045,
+      "learning_rate": 1.9774447604566715e-06,
+      "loss": 0.6845,
+      "step": 11178
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.427128928055661,
+      "learning_rate": 1.9766191630765964e-06,
+      "loss": 0.8851,
+      "step": 11179
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.102216662343594,
+      "learning_rate": 1.9757936956179308e-06,
+      "loss": 0.8074,
+      "step": 11180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.032542659862858,
+      "learning_rate": 1.9749683581161417e-06,
+      "loss": 0.8162,
+      "step": 11181
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1712617124510503,
+      "learning_rate": 1.9741431506067012e-06,
+      "loss": 0.7417,
+      "step": 11182
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.758405944469822,
+      "learning_rate": 1.9733180731250683e-06,
+      "loss": 0.8901,
+      "step": 11183
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7306148821250626,
+      "learning_rate": 1.9724931257066988e-06,
+      "loss": 0.7037,
+      "step": 11184
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8983165296740296,
+      "learning_rate": 1.971668308387042e-06,
+      "loss": 0.8503,
+      "step": 11185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7781641993707666,
+      "learning_rate": 1.970843621201541e-06,
+      "loss": 0.9025,
+      "step": 11186
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5835589541995114,
+      "learning_rate": 1.9700190641856378e-06,
+      "loss": 0.7875,
+      "step": 11187
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7694364945563117,
+      "learning_rate": 1.9691946373747626e-06,
+      "loss": 0.5411,
+      "step": 11188
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3101392984805798,
+      "learning_rate": 1.9683703408043447e-06,
+      "loss": 0.9062,
+      "step": 11189
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.926684587361578,
+      "learning_rate": 1.967546174509805e-06,
+      "loss": 0.9652,
+      "step": 11190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1632179973491197,
+      "learning_rate": 1.9667221385265595e-06,
+      "loss": 0.7936,
+      "step": 11191
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.799308073530266,
+      "learning_rate": 1.9658982328900195e-06,
+      "loss": 0.7809,
+      "step": 11192
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2406324107440163,
+      "learning_rate": 1.9650744576355894e-06,
+      "loss": 0.8229,
+      "step": 11193
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.830276324153215,
+      "learning_rate": 1.9642508127986676e-06,
+      "loss": 0.9361,
+      "step": 11194
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7445782182644454,
+      "learning_rate": 1.963427298414653e-06,
+      "loss": 0.6332,
+      "step": 11195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8048823610610916,
+      "learning_rate": 1.9626039145189263e-06,
+      "loss": 0.762,
+      "step": 11196
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.393395820805945,
+      "learning_rate": 1.9617806611468756e-06,
+      "loss": 0.8003,
+      "step": 11197
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8428356041733989,
+      "learning_rate": 1.960957538333877e-06,
+      "loss": 0.8604,
+      "step": 11198
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.239309795431111,
+      "learning_rate": 1.9601345461153005e-06,
+      "loss": 0.7506,
+      "step": 11199
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9338181489132557,
+      "learning_rate": 1.959311684526513e-06,
+      "loss": 0.6192,
+      "step": 11200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.244721802323892,
+      "learning_rate": 1.9584889536028728e-06,
+      "loss": 0.7828,
+      "step": 11201
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.05509689518412,
+      "learning_rate": 1.95766635337974e-06,
+      "loss": 0.7655,
+      "step": 11202
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.774666378318386,
+      "learning_rate": 1.956843883892457e-06,
+      "loss": 0.758,
+      "step": 11203
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.581076400308516,
+      "learning_rate": 1.9560215451763685e-06,
+      "loss": 0.9271,
+      "step": 11204
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1384806429837373,
+      "learning_rate": 1.9551993372668153e-06,
+      "loss": 0.7649,
+      "step": 11205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.11559585753827,
+      "learning_rate": 1.9543772601991277e-06,
+      "loss": 0.7046,
+      "step": 11206
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8150792739704078,
+      "learning_rate": 1.9535553140086322e-06,
+      "loss": 0.7836,
+      "step": 11207
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1282662114726882,
+      "learning_rate": 1.9527334987306495e-06,
+      "loss": 0.6549,
+      "step": 11208
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.877803267705164,
+      "learning_rate": 1.951911814400495e-06,
+      "loss": 0.8462,
+      "step": 11209
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.233760479627893,
+      "learning_rate": 1.9510902610534782e-06,
+      "loss": 0.9857,
+      "step": 11210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.146715648256149,
+      "learning_rate": 1.950268838724902e-06,
+      "loss": 0.8668,
+      "step": 11211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0885459962472492,
+      "learning_rate": 1.949447547450067e-06,
+      "loss": 0.5353,
+      "step": 11212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9799607954646388,
+      "learning_rate": 1.948626387264267e-06,
+      "loss": 0.9012,
+      "step": 11213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.683722481240191,
+      "learning_rate": 1.9478053582027826e-06,
+      "loss": 0.8232,
+      "step": 11214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5865320385837913,
+      "learning_rate": 1.9469844603009026e-06,
+      "loss": 0.7494,
+      "step": 11215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.143969102201056,
+      "learning_rate": 1.946163693593899e-06,
+      "loss": 0.9033,
+      "step": 11216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3921508429296376,
+      "learning_rate": 1.945343058117043e-06,
+      "loss": 0.6693,
+      "step": 11217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.397345161416357,
+      "learning_rate": 1.9445225539055982e-06,
+      "loss": 0.9595,
+      "step": 11218
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.81497991266333,
+      "learning_rate": 1.9437021809948232e-06,
+      "loss": 0.7018,
+      "step": 11219
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.280793274363507,
+      "learning_rate": 1.9428819394199753e-06,
+      "loss": 0.9287,
+      "step": 11220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1165367758348963,
+      "learning_rate": 1.9420618292162974e-06,
+      "loss": 0.826,
+      "step": 11221
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9548856797814151,
+      "learning_rate": 1.9412418504190316e-06,
+      "loss": 0.7802,
+      "step": 11222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2130837575464586,
+      "learning_rate": 1.9404220030634173e-06,
+      "loss": 1.0059,
+      "step": 11223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.242513749676882,
+      "learning_rate": 1.9396022871846836e-06,
+      "loss": 0.7698,
+      "step": 11224
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5002746288313977,
+      "learning_rate": 1.9387827028180556e-06,
+      "loss": 1.0686,
+      "step": 11225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.071389276385505,
+      "learning_rate": 1.937963249998752e-06,
+      "loss": 0.7213,
+      "step": 11226
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8971433810111789,
+      "learning_rate": 1.9371439287619876e-06,
+      "loss": 0.9148,
+      "step": 11227
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.058336146251433,
+      "learning_rate": 1.9363247391429695e-06,
+      "loss": 0.9043,
+      "step": 11228
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5521188468687095,
+      "learning_rate": 1.935505681176899e-06,
+      "loss": 0.7536,
+      "step": 11229
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9961406680819596,
+      "learning_rate": 1.9346867548989767e-06,
+      "loss": 0.6137,
+      "step": 11230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3287526167102106,
+      "learning_rate": 1.9338679603443922e-06,
+      "loss": 0.7268,
+      "step": 11231
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.885172744143351,
+      "learning_rate": 1.933049297548327e-06,
+      "loss": 0.8939,
+      "step": 11232
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3973766833442514,
+      "learning_rate": 1.932230766545966e-06,
+      "loss": 0.8858,
+      "step": 11233
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.060503264131041,
+      "learning_rate": 1.9314123673724805e-06,
+      "loss": 0.7976,
+      "step": 11234
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0292651198118485,
+      "learning_rate": 1.93059410006304e-06,
+      "loss": 0.6405,
+      "step": 11235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7433406024012914,
+      "learning_rate": 1.9297759646528076e-06,
+      "loss": 0.822,
+      "step": 11236
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2478001518276185,
+      "learning_rate": 1.928957961176937e-06,
+      "loss": 0.6252,
+      "step": 11237
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0074070024794155,
+      "learning_rate": 1.9281400896705865e-06,
+      "loss": 1.0216,
+      "step": 11238
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.00127757903928,
+      "learning_rate": 1.9273223501688943e-06,
+      "loss": 0.907,
+      "step": 11239
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2072611267993456,
+      "learning_rate": 1.9265047427070054e-06,
+      "loss": 0.7643,
+      "step": 11240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.872172754423958,
+      "learning_rate": 1.925687267320053e-06,
+      "loss": 0.6963,
+      "step": 11241
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0309509193319855,
+      "learning_rate": 1.924869924043165e-06,
+      "loss": 0.6201,
+      "step": 11242
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.264909144046112,
+      "learning_rate": 1.9240527129114654e-06,
+      "loss": 0.8862,
+      "step": 11243
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0998693963477613,
+      "learning_rate": 1.9232356339600717e-06,
+      "loss": 0.7587,
+      "step": 11244
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8656023295715762,
+      "learning_rate": 1.9224186872240946e-06,
+      "loss": 0.8269,
+      "step": 11245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0906162549576792,
+      "learning_rate": 1.921601872738641e-06,
+      "loss": 0.651,
+      "step": 11246
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0999810808882575,
+      "learning_rate": 1.920785190538809e-06,
+      "loss": 0.9119,
+      "step": 11247
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9399740713107665,
+      "learning_rate": 1.9199686406596968e-06,
+      "loss": 0.7019,
+      "step": 11248
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2688745408715607,
+      "learning_rate": 1.919152223136391e-06,
+      "loss": 0.7997,
+      "step": 11249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2109370241862845,
+      "learning_rate": 1.918335938003977e-06,
+      "loss": 0.8957,
+      "step": 11250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1354763945433857,
+      "learning_rate": 1.9175197852975304e-06,
+      "loss": 0.7642,
+      "step": 11251
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.587301332689423,
+      "learning_rate": 1.9167037650521215e-06,
+      "loss": 0.9197,
+      "step": 11252
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.354840286459649,
+      "learning_rate": 1.915887877302823e-06,
+      "loss": 0.9286,
+      "step": 11253
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.41535765471683,
+      "learning_rate": 1.9150721220846884e-06,
+      "loss": 0.8186,
+      "step": 11254
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3324689554813793,
+      "learning_rate": 1.914256499432774e-06,
+      "loss": 0.9051,
+      "step": 11255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4918242811780944,
+      "learning_rate": 1.913441009382133e-06,
+      "loss": 0.8814,
+      "step": 11256
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8810069020890343,
+      "learning_rate": 1.9126256519678034e-06,
+      "loss": 0.7174,
+      "step": 11257
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9990378565451787,
+      "learning_rate": 1.9118104272248266e-06,
+      "loss": 0.6647,
+      "step": 11258
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3588044997107764,
+      "learning_rate": 1.910995335188234e-06,
+      "loss": 0.8217,
+      "step": 11259
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.120350876414718,
+      "learning_rate": 1.9101803758930514e-06,
+      "loss": 0.7448,
+      "step": 11260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.280917222181941,
+      "learning_rate": 1.909365549374299e-06,
+      "loss": 0.8714,
+      "step": 11261
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.776913933654663,
+      "learning_rate": 1.9085508556669905e-06,
+      "loss": 0.6576,
+      "step": 11262
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.992728161739486,
+      "learning_rate": 1.9077362948061404e-06,
+      "loss": 0.8701,
+      "step": 11263
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.319965885880352,
+      "learning_rate": 1.906921866826747e-06,
+      "loss": 0.6769,
+      "step": 11264
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.557396783757622,
+      "learning_rate": 1.9061075717638078e-06,
+      "loss": 0.8761,
+      "step": 11265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3512685312014354,
+      "learning_rate": 1.9052934096523184e-06,
+      "loss": 0.8234,
+      "step": 11266
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.467837321277957,
+      "learning_rate": 1.9044793805272633e-06,
+      "loss": 0.9242,
+      "step": 11267
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.277419006562993,
+      "learning_rate": 1.9036654844236234e-06,
+      "loss": 0.8428,
+      "step": 11268
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.899382007536757,
+      "learning_rate": 1.9028517213763737e-06,
+      "loss": 0.5905,
+      "step": 11269
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.024447281837936,
+      "learning_rate": 1.902038091420481e-06,
+      "loss": 0.8599,
+      "step": 11270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9583200218944343,
+      "learning_rate": 1.9012245945909153e-06,
+      "loss": 0.7503,
+      "step": 11271
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.505296379279449,
+      "learning_rate": 1.900411230922628e-06,
+      "loss": 0.7302,
+      "step": 11272
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4739772460957186,
+      "learning_rate": 1.8995980004505715e-06,
+      "loss": 0.6293,
+      "step": 11273
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.26199041017295,
+      "learning_rate": 1.8987849032096973e-06,
+      "loss": 0.8854,
+      "step": 11274
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4242703475570013,
+      "learning_rate": 1.8979719392349388e-06,
+      "loss": 0.9165,
+      "step": 11275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9294858351909268,
+      "learning_rate": 1.8971591085612367e-06,
+      "loss": 0.7833,
+      "step": 11276
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.538975948509586,
+      "learning_rate": 1.8963464112235185e-06,
+      "loss": 0.8324,
+      "step": 11277
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6057326637602367,
+      "learning_rate": 1.895533847256707e-06,
+      "loss": 0.7887,
+      "step": 11278
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9758352333790574,
+      "learning_rate": 1.89472141669572e-06,
+      "loss": 0.8127,
+      "step": 11279
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8343809020841442,
+      "learning_rate": 1.8939091195754678e-06,
+      "loss": 0.6587,
+      "step": 11280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9435802677757086,
+      "learning_rate": 1.8930969559308615e-06,
+      "loss": 0.953,
+      "step": 11281
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3442421242183444,
+      "learning_rate": 1.8922849257967972e-06,
+      "loss": 0.8918,
+      "step": 11282
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.201730419425246,
+      "learning_rate": 1.891473029208169e-06,
+      "loss": 1.0201,
+      "step": 11283
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9077072892239972,
+      "learning_rate": 1.8906612661998698e-06,
+      "loss": 0.6866,
+      "step": 11284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.173829962353591,
+      "learning_rate": 1.8898496368067804e-06,
+      "loss": 0.8421,
+      "step": 11285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1418303513939265,
+      "learning_rate": 1.8890381410637786e-06,
+      "loss": 0.6485,
+      "step": 11286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0946189466008045,
+      "learning_rate": 1.8882267790057374e-06,
+      "loss": 0.7716,
+      "step": 11287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.676623963708801,
+      "learning_rate": 1.887415550667519e-06,
+      "loss": 0.7542,
+      "step": 11288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.382157568637439,
+      "learning_rate": 1.8866044560839902e-06,
+      "loss": 0.7346,
+      "step": 11289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1851370704590236,
+      "learning_rate": 1.8857934952899987e-06,
+      "loss": 0.6631,
+      "step": 11290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2752891208653425,
+      "learning_rate": 1.884982668320398e-06,
+      "loss": 0.7028,
+      "step": 11291
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9043482868408028,
+      "learning_rate": 1.8841719752100313e-06,
+      "loss": 0.9218,
+      "step": 11292
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9766542627614103,
+      "learning_rate": 1.8833614159937307e-06,
+      "loss": 0.576,
+      "step": 11293
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2931742899128964,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.8697,
+      "step": 11294
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0178556498222178,
+      "learning_rate": 1.881740699382662e-06,
+      "loss": 0.926,
+      "step": 11295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.039542770768269,
+      "learning_rate": 1.880930542057538e-06,
+      "loss": 0.9138,
+      "step": 11296
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.251034670908419,
+      "learning_rate": 1.880120518765775e-06,
+      "loss": 0.7942,
+      "step": 11297
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1847609546452005,
+      "learning_rate": 1.8793106295421797e-06,
+      "loss": 0.7402,
+      "step": 11298
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4816027686365363,
+      "learning_rate": 1.8785008744215606e-06,
+      "loss": 0.6791,
+      "step": 11299
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1688527540723572,
+      "learning_rate": 1.8776912534387076e-06,
+      "loss": 0.8674,
+      "step": 11300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1285994354957403,
+      "learning_rate": 1.8768817666284166e-06,
+      "loss": 0.976,
+      "step": 11301
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7568992202543616,
+      "learning_rate": 1.8760724140254716e-06,
+      "loss": 0.8868,
+      "step": 11302
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5854425488431154,
+      "learning_rate": 1.8752631956646522e-06,
+      "loss": 0.6516,
+      "step": 11303
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9007584745788244,
+      "learning_rate": 1.874454111580733e-06,
+      "loss": 0.6331,
+      "step": 11304
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0776183161566526,
+      "learning_rate": 1.873645161808481e-06,
+      "loss": 0.5854,
+      "step": 11305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2058497477816537,
+      "learning_rate": 1.8728363463826577e-06,
+      "loss": 0.9059,
+      "step": 11306
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.997417880926202,
+      "learning_rate": 1.872027665338025e-06,
+      "loss": 0.6768,
+      "step": 11307
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.58683176606728,
+      "learning_rate": 1.871219118709326e-06,
+      "loss": 0.7801,
+      "step": 11308
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3172043115563583,
+      "learning_rate": 1.8704107065313116e-06,
+      "loss": 0.7592,
+      "step": 11309
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0922079533478053,
+      "learning_rate": 1.8696024288387189e-06,
+      "loss": 0.8511,
+      "step": 11310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.047709356372317,
+      "learning_rate": 1.8687942856662815e-06,
+      "loss": 0.7276,
+      "step": 11311
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0321399790537544,
+      "learning_rate": 1.8679862770487273e-06,
+      "loss": 0.7242,
+      "step": 11312
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9832762023096895,
+      "learning_rate": 1.8671784030207762e-06,
+      "loss": 0.8956,
+      "step": 11313
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8754054004566831,
+      "learning_rate": 1.8663706636171503e-06,
+      "loss": 0.8444,
+      "step": 11314
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1690886931927664,
+      "learning_rate": 1.8655630588725536e-06,
+      "loss": 0.6424,
+      "step": 11315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9636902191112942,
+      "learning_rate": 1.8647555888216918e-06,
+      "loss": 0.7781,
+      "step": 11316
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9679242928168728,
+      "learning_rate": 1.8639482534992682e-06,
+      "loss": 0.7244,
+      "step": 11317
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.147498253939957,
+      "learning_rate": 1.863141052939969e-06,
+      "loss": 0.7284,
+      "step": 11318
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3497727131838135,
+      "learning_rate": 1.8623339871784869e-06,
+      "loss": 0.8451,
+      "step": 11319
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.841530483755602,
+      "learning_rate": 1.8615270562495014e-06,
+      "loss": 0.8394,
+      "step": 11320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7809221813874845,
+      "learning_rate": 1.8607202601876873e-06,
+      "loss": 0.7486,
+      "step": 11321
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9845898313983685,
+      "learning_rate": 1.8599135990277156e-06,
+      "loss": 0.8683,
+      "step": 11322
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0939843152801252,
+      "learning_rate": 1.8591070728042503e-06,
+      "loss": 0.6484,
+      "step": 11323
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.436150886005619,
+      "learning_rate": 1.8583006815519473e-06,
+      "loss": 0.6543,
+      "step": 11324
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4208048778278384,
+      "learning_rate": 1.8574944253054638e-06,
+      "loss": 0.8276,
+      "step": 11325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.483267901885839,
+      "learning_rate": 1.8566883040994411e-06,
+      "loss": 0.8931,
+      "step": 11326
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8074326848258586,
+      "learning_rate": 1.855882317968523e-06,
+      "loss": 0.7359,
+      "step": 11327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8167069739686694,
+      "learning_rate": 1.8550764669473454e-06,
+      "loss": 0.9328,
+      "step": 11328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.056046280703215,
+      "learning_rate": 1.8542707510705355e-06,
+      "loss": 0.7995,
+      "step": 11329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2338056898006227,
+      "learning_rate": 1.853465170372717e-06,
+      "loss": 0.9379,
+      "step": 11330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3534665786335083,
+      "learning_rate": 1.8526597248885065e-06,
+      "loss": 0.7602,
+      "step": 11331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7432490694725904,
+      "learning_rate": 1.8518544146525203e-06,
+      "loss": 0.5349,
+      "step": 11332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.214288668962746,
+      "learning_rate": 1.8510492396993595e-06,
+      "loss": 0.8393,
+      "step": 11333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3750453281786417,
+      "learning_rate": 1.8502442000636246e-06,
+      "loss": 0.9523,
+      "step": 11334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8370670219777525,
+      "learning_rate": 1.8494392957799146e-06,
+      "loss": 0.9111,
+      "step": 11335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4897764278988097,
+      "learning_rate": 1.8486345268828115e-06,
+      "loss": 0.7288,
+      "step": 11336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.31666368894735,
+      "learning_rate": 1.8478298934069028e-06,
+      "loss": 0.8083,
+      "step": 11337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.427255711421496,
+      "learning_rate": 1.8470253953867635e-06,
+      "loss": 0.7742,
+      "step": 11338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5420638886196905,
+      "learning_rate": 1.846221032856965e-06,
+      "loss": 0.8558,
+      "step": 11339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0692023859144135,
+      "learning_rate": 1.8454168058520732e-06,
+      "loss": 0.9072,
+      "step": 11340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1135910239081626,
+      "learning_rate": 1.8446127144066445e-06,
+      "loss": 0.708,
+      "step": 11341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1968062337211949,
+      "learning_rate": 1.8438087585552371e-06,
+      "loss": 0.7232,
+      "step": 11342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.784672659833036,
+      "learning_rate": 1.8430049383323983e-06,
+      "loss": 0.6256,
+      "step": 11343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0368346747814803,
+      "learning_rate": 1.8422012537726646e-06,
+      "loss": 0.781,
+      "step": 11344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2583213774299729,
+      "learning_rate": 1.8413977049105774e-06,
+      "loss": 0.6468,
+      "step": 11345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.205842925775989,
+      "learning_rate": 1.8405942917806657e-06,
+      "loss": 0.6842,
+      "step": 11346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9059008117637282,
+      "learning_rate": 1.8397910144174536e-06,
+      "loss": 0.6867,
+      "step": 11347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078623624754994,
+      "learning_rate": 1.8389878728554594e-06,
+      "loss": 0.8231,
+      "step": 11348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3288293385724974,
+      "learning_rate": 1.8381848671291953e-06,
+      "loss": 0.7679,
+      "step": 11349
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1200159651503978,
+      "learning_rate": 1.8373819972731721e-06,
+      "loss": 0.6557,
+      "step": 11350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4628262201638074,
+      "learning_rate": 1.8365792633218844e-06,
+      "loss": 0.7684,
+      "step": 11351
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.079872284074732,
+      "learning_rate": 1.835776665309833e-06,
+      "loss": 0.7149,
+      "step": 11352
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.059637260612399,
+      "learning_rate": 1.8349742032715072e-06,
+      "loss": 0.6433,
+      "step": 11353
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8031253709166821,
+      "learning_rate": 1.8341718772413852e-06,
+      "loss": 0.5984,
+      "step": 11354
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0730063208402314,
+      "learning_rate": 1.8333696872539502e-06,
+      "loss": 0.6929,
+      "step": 11355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.004218784585229,
+      "learning_rate": 1.8325676333436727e-06,
+      "loss": 0.8353,
+      "step": 11356
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.20245662667832,
+      "learning_rate": 1.8317657155450174e-06,
+      "loss": 0.6936,
+      "step": 11357
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5020502790619124,
+      "learning_rate": 1.8309639338924456e-06,
+      "loss": 0.8041,
+      "step": 11358
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1717441005600426,
+      "learning_rate": 1.8301622884204096e-06,
+      "loss": 0.6786,
+      "step": 11359
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0383605823673157,
+      "learning_rate": 1.8293607791633617e-06,
+      "loss": 0.7409,
+      "step": 11360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6887867103894267,
+      "learning_rate": 1.8285594061557421e-06,
+      "loss": 0.8658,
+      "step": 11361
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1390848171867127,
+      "learning_rate": 1.8277581694319884e-06,
+      "loss": 0.6609,
+      "step": 11362
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.900336225029946,
+      "learning_rate": 1.8269570690265308e-06,
+      "loss": 0.6218,
+      "step": 11363
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.144450403820145,
+      "learning_rate": 1.8261561049737946e-06,
+      "loss": 0.7456,
+      "step": 11364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.768660435644833,
+      "learning_rate": 1.8253552773081995e-06,
+      "loss": 0.9101,
+      "step": 11365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6304714328112113,
+      "learning_rate": 1.8245545860641584e-06,
+      "loss": 0.792,
+      "step": 11366
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9673982122482825,
+      "learning_rate": 1.8237540312760764e-06,
+      "loss": 0.7477,
+      "step": 11367
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.048806307647256,
+      "learning_rate": 1.822953612978362e-06,
+      "loss": 0.9039,
+      "step": 11368
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1212340053609537,
+      "learning_rate": 1.8221533312054024e-06,
+      "loss": 0.7202,
+      "step": 11369
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.050840286499657,
+      "learning_rate": 1.8213531859915939e-06,
+      "loss": 0.9829,
+      "step": 11370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3538309089559246,
+      "learning_rate": 1.8205531773713175e-06,
+      "loss": 1.0359,
+      "step": 11371
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8147874154193866,
+      "learning_rate": 1.8197533053789524e-06,
+      "loss": 0.7266,
+      "step": 11372
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2436405232094074,
+      "learning_rate": 1.8189535700488708e-06,
+      "loss": 0.8857,
+      "step": 11373
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0779028703623283,
+      "learning_rate": 1.818153971415439e-06,
+      "loss": 0.7375,
+      "step": 11374
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0492086509574494,
+      "learning_rate": 1.817354509513017e-06,
+      "loss": 0.7288,
+      "step": 11375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9796407037420316,
+      "learning_rate": 1.81655518437596e-06,
+      "loss": 0.8502,
+      "step": 11376
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9740323009965421,
+      "learning_rate": 1.8157559960386146e-06,
+      "loss": 0.8604,
+      "step": 11377
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.310158952369822,
+      "learning_rate": 1.8149569445353277e-06,
+      "loss": 0.9277,
+      "step": 11378
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2689000773541905,
+      "learning_rate": 1.8141580299004342e-06,
+      "loss": 0.6735,
+      "step": 11379
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1317677982289005,
+      "learning_rate": 1.8133592521682657e-06,
+      "loss": 0.7935,
+      "step": 11380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.598034571928999,
+      "learning_rate": 1.812560611373147e-06,
+      "loss": 0.8949,
+      "step": 11381
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3292088737258045,
+      "learning_rate": 1.8117621075493979e-06,
+      "loss": 0.688,
+      "step": 11382
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1208592790913934,
+      "learning_rate": 1.8109637407313313e-06,
+      "loss": 0.739,
+      "step": 11383
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8299731073499217,
+      "learning_rate": 1.8101655109532552e-06,
+      "loss": 0.7265,
+      "step": 11384
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.18237764511542,
+      "learning_rate": 1.8093674182494698e-06,
+      "loss": 0.7733,
+      "step": 11385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.806883412528859,
+      "learning_rate": 1.8085694626542755e-06,
+      "loss": 0.9495,
+      "step": 11386
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5129791770711383,
+      "learning_rate": 1.8077716442019554e-06,
+      "loss": 1.008,
+      "step": 11387
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5844402562391866,
+      "learning_rate": 1.8069739629267991e-06,
+      "loss": 0.8033,
+      "step": 11388
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0101833117977306,
+      "learning_rate": 1.8061764188630831e-06,
+      "loss": 0.6351,
+      "step": 11389
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9724169855217677,
+      "learning_rate": 1.8053790120450793e-06,
+      "loss": 0.8339,
+      "step": 11390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1386817785535115,
+      "learning_rate": 1.8045817425070544e-06,
+      "loss": 1.0094,
+      "step": 11391
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.786941948662358,
+      "learning_rate": 1.8037846102832667e-06,
+      "loss": 0.6987,
+      "step": 11392
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.562710207189171,
+      "learning_rate": 1.8029876154079755e-06,
+      "loss": 0.7114,
+      "step": 11393
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.137293177998229,
+      "learning_rate": 1.8021907579154257e-06,
+      "loss": 0.8662,
+      "step": 11394
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1454687260370355,
+      "learning_rate": 1.8013940378398593e-06,
+      "loss": 0.7825,
+      "step": 11395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4962086646803807,
+      "learning_rate": 1.8005974552155158e-06,
+      "loss": 0.7436,
+      "step": 11396
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0333533554285426,
+      "learning_rate": 1.7998010100766255e-06,
+      "loss": 0.7493,
+      "step": 11397
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9416997188028144,
+      "learning_rate": 1.799004702457413e-06,
+      "loss": 0.7562,
+      "step": 11398
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.98195787292031,
+      "learning_rate": 1.7982085323920973e-06,
+      "loss": 0.9395,
+      "step": 11399
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.416008233666806,
+      "learning_rate": 1.7974124999148918e-06,
+      "loss": 0.7337,
+      "step": 11400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.656449567520698,
+      "learning_rate": 1.7966166050600037e-06,
+      "loss": 0.8032,
+      "step": 11401
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6950323811310892,
+      "learning_rate": 1.795820847861633e-06,
+      "loss": 0.8402,
+      "step": 11402
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2086675257317685,
+      "learning_rate": 1.7950252283539776e-06,
+      "loss": 0.6307,
+      "step": 11403
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1641047267624254,
+      "learning_rate": 1.7942297465712282e-06,
+      "loss": 0.7734,
+      "step": 11404
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.039327265070134,
+      "learning_rate": 1.7934344025475625e-06,
+      "loss": 0.7339,
+      "step": 11405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.343006938702314,
+      "learning_rate": 1.7926391963171642e-06,
+      "loss": 0.8106,
+      "step": 11406
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6442308133381203,
+      "learning_rate": 1.7918441279142025e-06,
+      "loss": 0.9647,
+      "step": 11407
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2456916706474783,
+      "learning_rate": 1.7910491973728433e-06,
+      "loss": 0.7475,
+      "step": 11408
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.028844803317216,
+      "learning_rate": 1.7902544047272468e-06,
+      "loss": 0.7075,
+      "step": 11409
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.228920755749624,
+      "learning_rate": 1.7894597500115657e-06,
+      "loss": 0.9408,
+      "step": 11410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.656615108896132,
+      "learning_rate": 1.788665233259953e-06,
+      "loss": 0.7247,
+      "step": 11411
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9196892780015025,
+      "learning_rate": 1.7878708545065433e-06,
+      "loss": 0.7977,
+      "step": 11412
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.32698596295369,
+      "learning_rate": 1.7870766137854784e-06,
+      "loss": 0.8626,
+      "step": 11413
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6266381707709305,
+      "learning_rate": 1.7862825111308873e-06,
+      "loss": 0.7586,
+      "step": 11414
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.491016118157462,
+      "learning_rate": 1.7854885465768939e-06,
+      "loss": 0.8074,
+      "step": 11415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0976679319047915,
+      "learning_rate": 1.7846947201576171e-06,
+      "loss": 0.6119,
+      "step": 11416
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3475138938612896,
+      "learning_rate": 1.7839010319071687e-06,
+      "loss": 0.8068,
+      "step": 11417
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2522127164408796,
+      "learning_rate": 1.7831074818596555e-06,
+      "loss": 0.7691,
+      "step": 11418
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3413977006809663,
+      "learning_rate": 1.7823140700491786e-06,
+      "loss": 0.9432,
+      "step": 11419
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1398266985952024,
+      "learning_rate": 1.7815207965098303e-06,
+      "loss": 0.6235,
+      "step": 11420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6576420441774813,
+      "learning_rate": 1.780727661275703e-06,
+      "loss": 0.6347,
+      "step": 11421
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8135338425923218,
+      "learning_rate": 1.7799346643808796e-06,
+      "loss": 0.8493,
+      "step": 11422
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6852224083954184,
+      "learning_rate": 1.7791418058594318e-06,
+      "loss": 0.6772,
+      "step": 11423
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0795864263687704,
+      "learning_rate": 1.7783490857454354e-06,
+      "loss": 0.8334,
+      "step": 11424
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.844989548785981,
+      "learning_rate": 1.7775565040729536e-06,
+      "loss": 0.6239,
+      "step": 11425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.069556424649365,
+      "learning_rate": 1.776764060876046e-06,
+      "loss": 0.6524,
+      "step": 11426
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2578234822206538,
+      "learning_rate": 1.7759717561887652e-06,
+      "loss": 0.7606,
+      "step": 11427
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5271179329724056,
+      "learning_rate": 1.7751795900451562e-06,
+      "loss": 0.8544,
+      "step": 11428
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.158089177261372,
+      "learning_rate": 1.7743875624792662e-06,
+      "loss": 0.8825,
+      "step": 11429
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0209618513332845,
+      "learning_rate": 1.7735956735251224e-06,
+      "loss": 0.8949,
+      "step": 11430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3520327398784766,
+      "learning_rate": 1.7728039232167603e-06,
+      "loss": 0.6834,
+      "step": 11431
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.040397591871451,
+      "learning_rate": 1.7720123115882009e-06,
+      "loss": 0.835,
+      "step": 11432
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9766826201163731,
+      "learning_rate": 1.771220838673462e-06,
+      "loss": 0.7228,
+      "step": 11433
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9759695287917525,
+      "learning_rate": 1.770429504506554e-06,
+      "loss": 0.7009,
+      "step": 11434
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6978713447945077,
+      "learning_rate": 1.769638309121483e-06,
+      "loss": 0.787,
+      "step": 11435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2009131736540186,
+      "learning_rate": 1.7688472525522481e-06,
+      "loss": 0.7655,
+      "step": 11436
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1172074343423626,
+      "learning_rate": 1.7680563348328433e-06,
+      "loss": 0.7091,
+      "step": 11437
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.081262403722863,
+      "learning_rate": 1.7672655559972535e-06,
+      "loss": 0.9095,
+      "step": 11438
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6179112960972284,
+      "learning_rate": 1.7664749160794642e-06,
+      "loss": 0.754,
+      "step": 11439
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.153167357451841,
+      "learning_rate": 1.7656844151134495e-06,
+      "loss": 0.7073,
+      "step": 11440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5115593631109197,
+      "learning_rate": 1.7648940531331787e-06,
+      "loss": 0.8155,
+      "step": 11441
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1299265218822567,
+      "learning_rate": 1.7641038301726155e-06,
+      "loss": 0.878,
+      "step": 11442
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.974894421952031,
+      "learning_rate": 1.763313746265715e-06,
+      "loss": 0.9263,
+      "step": 11443
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7176625858118446,
+      "learning_rate": 1.7625238014464358e-06,
+      "loss": 0.7669,
+      "step": 11444
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.459237187405715,
+      "learning_rate": 1.7617339957487167e-06,
+      "loss": 0.8759,
+      "step": 11445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1650606913735353,
+      "learning_rate": 1.760944329206498e-06,
+      "loss": 0.7059,
+      "step": 11446
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.853403502340185,
+      "learning_rate": 1.760154801853719e-06,
+      "loss": 0.873,
+      "step": 11447
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8913179540692227,
+      "learning_rate": 1.7593654137243005e-06,
+      "loss": 0.9153,
+      "step": 11448
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0717241931721104,
+      "learning_rate": 1.7585761648521688e-06,
+      "loss": 0.7493,
+      "step": 11449
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3528522205368883,
+      "learning_rate": 1.7577870552712378e-06,
+      "loss": 0.6591,
+      "step": 11450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.111086820745274,
+      "learning_rate": 1.756998085015419e-06,
+      "loss": 0.7693,
+      "step": 11451
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.041872283906586,
+      "learning_rate": 1.7562092541186144e-06,
+      "loss": 0.8944,
+      "step": 11452
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9724765634031125,
+      "learning_rate": 1.7554205626147202e-06,
+      "loss": 0.956,
+      "step": 11453
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7286139713699713,
+      "learning_rate": 1.7546320105376346e-06,
+      "loss": 0.8392,
+      "step": 11454
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.082453275635426,
+      "learning_rate": 1.7538435979212376e-06,
+      "loss": 0.6549,
+      "step": 11455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.908907230306434,
+      "learning_rate": 1.7530553247994093e-06,
+      "loss": 0.7594,
+      "step": 11456
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.216656781471003,
+      "learning_rate": 1.7522671912060262e-06,
+      "loss": 0.8617,
+      "step": 11457
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.065423039372403,
+      "learning_rate": 1.751479197174955e-06,
+      "loss": 0.6962,
+      "step": 11458
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9893446587674077,
+      "learning_rate": 1.750691342740058e-06,
+      "loss": 0.6698,
+      "step": 11459
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0888487695905753,
+      "learning_rate": 1.7499036279351905e-06,
+      "loss": 0.7988,
+      "step": 11460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8313993002689246,
+      "learning_rate": 1.7491160527942002e-06,
+      "loss": 0.7259,
+      "step": 11461
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.943247745347911,
+      "learning_rate": 1.7483286173509372e-06,
+      "loss": 0.7316,
+      "step": 11462
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.313760755171004,
+      "learning_rate": 1.7475413216392317e-06,
+      "loss": 0.8466,
+      "step": 11463
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.010894568299653,
+      "learning_rate": 1.746754165692921e-06,
+      "loss": 0.8661,
+      "step": 11464
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1187476965263317,
+      "learning_rate": 1.7459671495458308e-06,
+      "loss": 0.8849,
+      "step": 11465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.353988609257844,
+      "learning_rate": 1.7451802732317763e-06,
+      "loss": 0.695,
+      "step": 11466
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.108346503258252,
+      "learning_rate": 1.7443935367845755e-06,
+      "loss": 0.6729,
+      "step": 11467
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1027682422506382,
+      "learning_rate": 1.7436069402380352e-06,
+      "loss": 0.7573,
+      "step": 11468
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8635537037354728,
+      "learning_rate": 1.742820483625957e-06,
+      "loss": 0.6441,
+      "step": 11469
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1471050886412686,
+      "learning_rate": 1.7420341669821373e-06,
+      "loss": 0.7484,
+      "step": 11470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.617573287982733,
+      "learning_rate": 1.7412479903403628e-06,
+      "loss": 0.7972,
+      "step": 11471
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9556487839383208,
+      "learning_rate": 1.7404619537344236e-06,
+      "loss": 0.6977,
+      "step": 11472
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.533938831619165,
+      "learning_rate": 1.7396760571980902e-06,
+      "loss": 0.8229,
+      "step": 11473
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6582476885109259,
+      "learning_rate": 1.7388903007651398e-06,
+      "loss": 0.6698,
+      "step": 11474
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.573143140971036,
+      "learning_rate": 1.7381046844693356e-06,
+      "loss": 0.6625,
+      "step": 11475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.224877265510444,
+      "learning_rate": 1.737319208344438e-06,
+      "loss": 0.8834,
+      "step": 11476
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5797389432356495,
+      "learning_rate": 1.7365338724242003e-06,
+      "loss": 0.7388,
+      "step": 11477
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.202304799800638,
+      "learning_rate": 1.7357486767423699e-06,
+      "loss": 0.8333,
+      "step": 11478
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0366850824221507,
+      "learning_rate": 1.7349636213326876e-06,
+      "loss": 0.6875,
+      "step": 11479
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0146795399917656,
+      "learning_rate": 1.7341787062288928e-06,
+      "loss": 0.714,
+      "step": 11480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1375019196946958,
+      "learning_rate": 1.7333939314647097e-06,
+      "loss": 0.7325,
+      "step": 11481
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9398330925730032,
+      "learning_rate": 1.7326092970738656e-06,
+      "loss": 0.9206,
+      "step": 11482
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6036275126419715,
+      "learning_rate": 1.7318248030900792e-06,
+      "loss": 0.8296,
+      "step": 11483
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5024187196287864,
+      "learning_rate": 1.7310404495470557e-06,
+      "loss": 0.8479,
+      "step": 11484
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3763983042526835,
+      "learning_rate": 1.7302562364785063e-06,
+      "loss": 1.0101,
+      "step": 11485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.970935335410082,
+      "learning_rate": 1.729472163918129e-06,
+      "loss": 0.7664,
+      "step": 11486
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.159486226152642,
+      "learning_rate": 1.7286882318996162e-06,
+      "loss": 0.6278,
+      "step": 11487
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.063737223164717,
+      "learning_rate": 1.7279044404566554e-06,
+      "loss": 0.9572,
+      "step": 11488
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.213170056024607,
+      "learning_rate": 1.727120789622927e-06,
+      "loss": 0.8611,
+      "step": 11489
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.686528980168588,
+      "learning_rate": 1.7263372794321104e-06,
+      "loss": 0.9779,
+      "step": 11490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9152624404947385,
+      "learning_rate": 1.7255539099178687e-06,
+      "loss": 0.8743,
+      "step": 11491
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.130827466156152,
+      "learning_rate": 1.7247706811138692e-06,
+      "loss": 0.7988,
+      "step": 11492
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.127930341774562,
+      "learning_rate": 1.723987593053768e-06,
+      "loss": 0.7292,
+      "step": 11493
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8634836589578745,
+      "learning_rate": 1.7232046457712164e-06,
+      "loss": 0.8038,
+      "step": 11494
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.472349952040512,
+      "learning_rate": 1.7224218392998587e-06,
+      "loss": 0.7665,
+      "step": 11495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5634450181510333,
+      "learning_rate": 1.7216391736733345e-06,
+      "loss": 0.6202,
+      "step": 11496
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9865394194607751,
+      "learning_rate": 1.7208566489252742e-06,
+      "loss": 0.595,
+      "step": 11497
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.389091926978394,
+      "learning_rate": 1.7200742650893105e-06,
+      "loss": 0.703,
+      "step": 11498
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1353807247743772,
+      "learning_rate": 1.7192920221990566e-06,
+      "loss": 0.6838,
+      "step": 11499
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2552392625362985,
+      "learning_rate": 1.7185099202881328e-06,
+      "loss": 0.712,
+      "step": 11500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9789720800219204,
+      "learning_rate": 1.7177279593901463e-06,
+      "loss": 0.7742,
+      "step": 11501
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7661028268976657,
+      "learning_rate": 1.7169461395386988e-06,
+      "loss": 0.9063,
+      "step": 11502
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2123037449251255,
+      "learning_rate": 1.7161644607673884e-06,
+      "loss": 0.7231,
+      "step": 11503
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8845461357566058,
+      "learning_rate": 1.7153829231098018e-06,
+      "loss": 0.9071,
+      "step": 11504
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.586566219758026,
+      "learning_rate": 1.7146015265995302e-06,
+      "loss": 0.8429,
+      "step": 11505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9020172272770677,
+      "learning_rate": 1.7138202712701463e-06,
+      "loss": 0.9237,
+      "step": 11506
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.983753133807552,
+      "learning_rate": 1.7130391571552224e-06,
+      "loss": 0.7461,
+      "step": 11507
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.051273685115285,
+      "learning_rate": 1.712258184288328e-06,
+      "loss": 0.7362,
+      "step": 11508
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1999689236428974,
+      "learning_rate": 1.7114773527030215e-06,
+      "loss": 0.784,
+      "step": 11509
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.863983880536434,
+      "learning_rate": 1.7106966624328575e-06,
+      "loss": 0.8013,
+      "step": 11510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0615593871742477,
+      "learning_rate": 1.7099161135113828e-06,
+      "loss": 0.7652,
+      "step": 11511
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1346903808457913,
+      "learning_rate": 1.7091357059721409e-06,
+      "loss": 0.8908,
+      "step": 11512
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9964400951762644,
+      "learning_rate": 1.7083554398486663e-06,
+      "loss": 0.8013,
+      "step": 11513
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3904868346241486,
+      "learning_rate": 1.7075753151744885e-06,
+      "loss": 0.7798,
+      "step": 11514
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.27941509858628,
+      "learning_rate": 1.7067953319831327e-06,
+      "loss": 0.8113,
+      "step": 11515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.254144863842929,
+      "learning_rate": 1.7060154903081177e-06,
+      "loss": 0.7555,
+      "step": 11516
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.18284093813623,
+      "learning_rate": 1.7052357901829502e-06,
+      "loss": 0.6239,
+      "step": 11517
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9335497667318364,
+      "learning_rate": 1.70445623164114e-06,
+      "loss": 0.8665,
+      "step": 11518
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0923824956880366,
+      "learning_rate": 1.7036768147161853e-06,
+      "loss": 0.6268,
+      "step": 11519
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1507330011453205,
+      "learning_rate": 1.7028975394415791e-06,
+      "loss": 0.7042,
+      "step": 11520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.076587328281301,
+      "learning_rate": 1.702118405850809e-06,
+      "loss": 0.8582,
+      "step": 11521
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.039367653011386,
+      "learning_rate": 1.7013394139773537e-06,
+      "loss": 0.6368,
+      "step": 11522
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4567539901608964,
+      "learning_rate": 1.700560563854694e-06,
+      "loss": 0.8012,
+      "step": 11523
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8977652182496523,
+      "learning_rate": 1.6997818555162915e-06,
+      "loss": 0.8452,
+      "step": 11524
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.182724061044222,
+      "learning_rate": 1.6990032889956142e-06,
+      "loss": 0.8884,
+      "step": 11525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.823546638522307,
+      "learning_rate": 1.6982248643261173e-06,
+      "loss": 0.8207,
+      "step": 11526
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9952720157680588,
+      "learning_rate": 1.6974465815412517e-06,
+      "loss": 0.6948,
+      "step": 11527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2124331100487233,
+      "learning_rate": 1.6966684406744615e-06,
+      "loss": 0.7188,
+      "step": 11528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9766118642879202,
+      "learning_rate": 1.6958904417591853e-06,
+      "loss": 0.8796,
+      "step": 11529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.496869594459726,
+      "learning_rate": 1.6951125848288558e-06,
+      "loss": 0.6756,
+      "step": 11530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0018513620665646,
+      "learning_rate": 1.6943348699168981e-06,
+      "loss": 0.748,
+      "step": 11531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2538447891006332,
+      "learning_rate": 1.6935572970567321e-06,
+      "loss": 0.6757,
+      "step": 11532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1780072464583404,
+      "learning_rate": 1.6927798662817745e-06,
+      "loss": 0.8313,
+      "step": 11533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8958023770374641,
+      "learning_rate": 1.6920025776254334e-06,
+      "loss": 0.6547,
+      "step": 11534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9849441849590266,
+      "learning_rate": 1.6912254311211052e-06,
+      "loss": 0.7814,
+      "step": 11535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2286652948518233,
+      "learning_rate": 1.6904484268021915e-06,
+      "loss": 0.8539,
+      "step": 11536
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.398951620860734,
+      "learning_rate": 1.689671564702079e-06,
+      "loss": 0.9725,
+      "step": 11537
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0973152011364795,
+      "learning_rate": 1.688894844854152e-06,
+      "loss": 0.7628,
+      "step": 11538
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9346250219459242,
+      "learning_rate": 1.6881182672917879e-06,
+      "loss": 0.8599,
+      "step": 11539
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0166457745568167,
+      "learning_rate": 1.6873418320483565e-06,
+      "loss": 0.7397,
+      "step": 11540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.06248853506984,
+      "learning_rate": 1.6865655391572266e-06,
+      "loss": 0.7817,
+      "step": 11541
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4152859969489393,
+      "learning_rate": 1.6857893886517523e-06,
+      "loss": 0.9826,
+      "step": 11542
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1573260721352394,
+      "learning_rate": 1.6850133805652907e-06,
+      "loss": 0.7744,
+      "step": 11543
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1978564712921096,
+      "learning_rate": 1.6842375149311868e-06,
+      "loss": 0.8055,
+      "step": 11544
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8144609822738866,
+      "learning_rate": 1.6834617917827817e-06,
+      "loss": 0.9226,
+      "step": 11545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6799575748370255,
+      "learning_rate": 1.6826862111534098e-06,
+      "loss": 1.1633,
+      "step": 11546
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9982348257661988,
+      "learning_rate": 1.6819107730763996e-06,
+      "loss": 0.6653,
+      "step": 11547
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.193033765194342,
+      "learning_rate": 1.681135477585073e-06,
+      "loss": 0.7994,
+      "step": 11548
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8536349506958416,
+      "learning_rate": 1.680360324712746e-06,
+      "loss": 0.7053,
+      "step": 11549
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1092078344084888,
+      "learning_rate": 1.6795853144927282e-06,
+      "loss": 0.6014,
+      "step": 11550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8786275951284495,
+      "learning_rate": 1.6788104469583256e-06,
+      "loss": 0.862,
+      "step": 11551
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8890144148501506,
+      "learning_rate": 1.678035722142835e-06,
+      "loss": 0.7251,
+      "step": 11552
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.313844677463248,
+      "learning_rate": 1.6772611400795475e-06,
+      "loss": 0.7982,
+      "step": 11553
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8737377979105676,
+      "learning_rate": 1.6764867008017493e-06,
+      "loss": 0.8755,
+      "step": 11554
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3158199692494126,
+      "learning_rate": 1.6757124043427193e-06,
+      "loss": 0.7729,
+      "step": 11555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.149634900222453,
+      "learning_rate": 1.674938250735731e-06,
+      "loss": 0.6611,
+      "step": 11556
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3472619542579327,
+      "learning_rate": 1.6741642400140513e-06,
+      "loss": 0.923,
+      "step": 11557
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9706734007119593,
+      "learning_rate": 1.6733903722109391e-06,
+      "loss": 0.6914,
+      "step": 11558
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9649837872905664,
+      "learning_rate": 1.672616647359655e-06,
+      "loss": 0.7595,
+      "step": 11559
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.116056300056902,
+      "learning_rate": 1.6718430654934408e-06,
+      "loss": 0.6614,
+      "step": 11560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3677157426395037,
+      "learning_rate": 1.671069626645544e-06,
+      "loss": 0.8762,
+      "step": 11561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.170814637254187,
+      "learning_rate": 1.6702963308491994e-06,
+      "loss": 0.5872,
+      "step": 11562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.675236757691958,
+      "learning_rate": 1.669523178137637e-06,
+      "loss": 0.7689,
+      "step": 11563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.460149670103318,
+      "learning_rate": 1.668750168544081e-06,
+      "loss": 0.5765,
+      "step": 11564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8694012554481674,
+      "learning_rate": 1.6679773021017475e-06,
+      "loss": 0.8147,
+      "step": 11565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1674399337602588,
+      "learning_rate": 1.6672045788438534e-06,
+      "loss": 0.6412,
+      "step": 11566
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.82749816842515,
+      "learning_rate": 1.6664319988036e-06,
+      "loss": 0.766,
+      "step": 11567
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.782006480850039,
+      "learning_rate": 1.6656595620141857e-06,
+      "loss": 0.9969,
+      "step": 11568
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.917226571716162,
+      "learning_rate": 1.664887268508808e-06,
+      "loss": 0.8058,
+      "step": 11569
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.158198192713858,
+      "learning_rate": 1.6641151183206516e-06,
+      "loss": 0.5752,
+      "step": 11570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2503155688286998,
+      "learning_rate": 1.663343111482898e-06,
+      "loss": 0.8346,
+      "step": 11571
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0593787225390985,
+      "learning_rate": 1.6625712480287225e-06,
+      "loss": 0.8158,
+      "step": 11572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1781869152384967,
+      "learning_rate": 1.661799527991293e-06,
+      "loss": 0.7105,
+      "step": 11573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9822206373740312,
+      "learning_rate": 1.6610279514037725e-06,
+      "loss": 1.0292,
+      "step": 11574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6440330831286527,
+      "learning_rate": 1.6602565182993152e-06,
+      "loss": 0.8902,
+      "step": 11575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1162579916282909,
+      "learning_rate": 1.6594852287110758e-06,
+      "loss": 0.7112,
+      "step": 11576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8387605329596586,
+      "learning_rate": 1.6587140826721964e-06,
+      "loss": 0.764,
+      "step": 11577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2493367306627112,
+      "learning_rate": 1.657943080215812e-06,
+      "loss": 0.9231,
+      "step": 11578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7236758397646628,
+      "learning_rate": 1.657172221375058e-06,
+      "loss": 0.8844,
+      "step": 11579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7640976717547865,
+      "learning_rate": 1.656401506183058e-06,
+      "loss": 0.9128,
+      "step": 11580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8530267327596417,
+      "learning_rate": 1.6556309346729322e-06,
+      "loss": 0.8653,
+      "step": 11581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9364710091600166,
+      "learning_rate": 1.6548605068777935e-06,
+      "loss": 0.8207,
+      "step": 11582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.227409133307853,
+      "learning_rate": 1.6540902228307471e-06,
+      "loss": 0.6705,
+      "step": 11583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.632353132730468,
+      "learning_rate": 1.6533200825648993e-06,
+      "loss": 0.925,
+      "step": 11584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.367227969458966,
+      "learning_rate": 1.6525500861133386e-06,
+      "loss": 0.7086,
+      "step": 11585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8796913896961225,
+      "learning_rate": 1.6517802335091549e-06,
+      "loss": 0.8935,
+      "step": 11586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.161414607570636,
+      "learning_rate": 1.6510105247854325e-06,
+      "loss": 0.8138,
+      "step": 11587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3576560268130717,
+      "learning_rate": 1.6502409599752468e-06,
+      "loss": 1.0356,
+      "step": 11588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1885620402902366,
+      "learning_rate": 1.6494715391116671e-06,
+      "loss": 0.7667,
+      "step": 11589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.982704837935902,
+      "learning_rate": 1.6487022622277582e-06,
+      "loss": 0.9341,
+      "step": 11590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9770449843935244,
+      "learning_rate": 1.6479331293565758e-06,
+      "loss": 0.6206,
+      "step": 11591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2243245103197857,
+      "learning_rate": 1.6471641405311727e-06,
+      "loss": 0.8998,
+      "step": 11592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6121384610435086,
+      "learning_rate": 1.6463952957845924e-06,
+      "loss": 0.8961,
+      "step": 11593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9457251925852406,
+      "learning_rate": 1.6456265951498763e-06,
+      "loss": 0.8465,
+      "step": 11594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.010744677355854,
+      "learning_rate": 1.6448580386600576e-06,
+      "loss": 0.5901,
+      "step": 11595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9623032987856139,
+      "learning_rate": 1.6440896263481582e-06,
+      "loss": 0.652,
+      "step": 11596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2448015236298318,
+      "learning_rate": 1.6433213582472036e-06,
+      "loss": 0.8436,
+      "step": 11597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.622537210418425,
+      "learning_rate": 1.6425532343902051e-06,
+      "loss": 0.7716,
+      "step": 11598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.048627924231485,
+      "learning_rate": 1.641785254810172e-06,
+      "loss": 0.8653,
+      "step": 11599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6986037807005947,
+      "learning_rate": 1.6410174195401057e-06,
+      "loss": 0.7022,
+      "step": 11600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1340475367295806,
+      "learning_rate": 1.6402497286129998e-06,
+      "loss": 0.5719,
+      "step": 11601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1869140597572159,
+      "learning_rate": 1.6394821820618496e-06,
+      "loss": 0.6643,
+      "step": 11602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.38715578785001,
+      "learning_rate": 1.63871477991963e-06,
+      "loss": 0.8363,
+      "step": 11603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8786419818770899,
+      "learning_rate": 1.6379475222193248e-06,
+      "loss": 0.7363,
+      "step": 11604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.137478708566257,
+      "learning_rate": 1.637180408993902e-06,
+      "loss": 0.8002,
+      "step": 11605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.036289553838576,
+      "learning_rate": 1.636413440276326e-06,
+      "loss": 0.7065,
+      "step": 11606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.946888775737773,
+      "learning_rate": 1.635646616099556e-06,
+      "loss": 0.6791,
+      "step": 11607
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5981379300737246,
+      "learning_rate": 1.634879936496544e-06,
+      "loss": 0.9078,
+      "step": 11608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1659868531604176,
+      "learning_rate": 1.6341134015002352e-06,
+      "loss": 0.6611,
+      "step": 11609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6909623082613683,
+      "learning_rate": 1.63334701114357e-06,
+      "loss": 0.6045,
+      "step": 11610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.747009022908939,
+      "learning_rate": 1.63258076545948e-06,
+      "loss": 1.0334,
+      "step": 11611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.252754411612389,
+      "learning_rate": 1.6318146644808952e-06,
+      "loss": 0.7391,
+      "step": 11612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1379669355864364,
+      "learning_rate": 1.631048708240736e-06,
+      "loss": 0.7323,
+      "step": 11613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1189226391569294,
+      "learning_rate": 1.6302828967719175e-06,
+      "loss": 0.7466,
+      "step": 11614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9677370848642752,
+      "learning_rate": 1.6295172301073469e-06,
+      "loss": 0.6699,
+      "step": 11615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.288160825863522,
+      "learning_rate": 1.6287517082799265e-06,
+      "loss": 0.914,
+      "step": 11616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3684753165886134,
+      "learning_rate": 1.6279863313225559e-06,
+      "loss": 0.6892,
+      "step": 11617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4679830956215585,
+      "learning_rate": 1.6272210992681213e-06,
+      "loss": 0.6823,
+      "step": 11618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.008368485207978,
+      "learning_rate": 1.626456012149506e-06,
+      "loss": 0.823,
+      "step": 11619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3641935822886997,
+      "learning_rate": 1.6256910699995921e-06,
+      "loss": 0.8629,
+      "step": 11620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.953022755720124,
+      "learning_rate": 1.624926272851245e-06,
+      "loss": 0.6729,
+      "step": 11621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.2176961930045715,
+      "learning_rate": 1.6241616207373339e-06,
+      "loss": 0.6527,
+      "step": 11622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265874110366144,
+      "learning_rate": 1.6233971136907168e-06,
+      "loss": 0.8595,
+      "step": 11623
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.606604222120935,
+      "learning_rate": 1.6226327517442453e-06,
+      "loss": 0.9813,
+      "step": 11624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0078340048653214,
+      "learning_rate": 1.6218685349307667e-06,
+      "loss": 0.7721,
+      "step": 11625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2143381718812143,
+      "learning_rate": 1.6211044632831185e-06,
+      "loss": 0.7002,
+      "step": 11626
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9876203540699556,
+      "learning_rate": 1.620340536834139e-06,
+      "loss": 0.5707,
+      "step": 11627
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.764357532700348,
+      "learning_rate": 1.6195767556166548e-06,
+      "loss": 0.6861,
+      "step": 11628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.392857718223417,
+      "learning_rate": 1.6188131196634827e-06,
+      "loss": 0.9462,
+      "step": 11629
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5821241497093463,
+      "learning_rate": 1.6180496290074432e-06,
+      "loss": 0.7149,
+      "step": 11630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.2319502862602745,
+      "learning_rate": 1.617286283681343e-06,
+      "loss": 0.756,
+      "step": 11631
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.237499785332781,
+      "learning_rate": 1.6165230837179846e-06,
+      "loss": 0.8128,
+      "step": 11632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0520889755139877,
+      "learning_rate": 1.6157600291501646e-06,
+      "loss": 0.6022,
+      "step": 11633
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.466450337166461,
+      "learning_rate": 1.6149971200106723e-06,
+      "loss": 0.6938,
+      "step": 11634
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5081527358322644,
+      "learning_rate": 1.6142343563322955e-06,
+      "loss": 0.716,
+      "step": 11635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.860860170964973,
+      "learning_rate": 1.613471738147807e-06,
+      "loss": 0.7564,
+      "step": 11636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4130923769696966,
+      "learning_rate": 1.6127092654899784e-06,
+      "loss": 0.9215,
+      "step": 11637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.030121258815563,
+      "learning_rate": 1.6119469383915804e-06,
+      "loss": 0.6594,
+      "step": 11638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2914192885664626,
+      "learning_rate": 1.6111847568853645e-06,
+      "loss": 0.847,
+      "step": 11639
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.022543485031463,
+      "learning_rate": 1.6104227210040884e-06,
+      "loss": 0.8641,
+      "step": 11640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2018291915794577,
+      "learning_rate": 1.6096608307804973e-06,
+      "loss": 0.5661,
+      "step": 11641
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9182439820186126,
+      "learning_rate": 1.608899086247331e-06,
+      "loss": 0.6765,
+      "step": 11642
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.003007038082785,
+      "learning_rate": 1.608137487437323e-06,
+      "loss": 0.8768,
+      "step": 11643
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0622592099222523,
+      "learning_rate": 1.6073760343831996e-06,
+      "loss": 0.7239,
+      "step": 11644
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9574680519777952,
+      "learning_rate": 1.6066147271176857e-06,
+      "loss": 0.5986,
+      "step": 11645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9624156197153269,
+      "learning_rate": 1.6058535656734958e-06,
+      "loss": 0.6891,
+      "step": 11646
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.18906427317176,
+      "learning_rate": 1.6050925500833342e-06,
+      "loss": 0.7804,
+      "step": 11647
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1363384281061966,
+      "learning_rate": 1.604331680379908e-06,
+      "loss": 0.8319,
+      "step": 11648
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.159531611759536,
+      "learning_rate": 1.603570956595913e-06,
+      "loss": 0.6554,
+      "step": 11649
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09509841584444,
+      "learning_rate": 1.602810378764037e-06,
+      "loss": 0.8475,
+      "step": 11650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.085566378009153,
+      "learning_rate": 1.6020499469169659e-06,
+      "loss": 0.7902,
+      "step": 11651
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.717992928467379,
+      "learning_rate": 1.601289661087374e-06,
+      "loss": 0.835,
+      "step": 11652
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0089492390145955,
+      "learning_rate": 1.600529521307938e-06,
+      "loss": 0.6741,
+      "step": 11653
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.013613631019047,
+      "learning_rate": 1.5997695276113168e-06,
+      "loss": 0.8437,
+      "step": 11654
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1442250773587148,
+      "learning_rate": 1.599009680030173e-06,
+      "loss": 0.8131,
+      "step": 11655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.677687003425993,
+      "learning_rate": 1.598249978597159e-06,
+      "loss": 0.7902,
+      "step": 11656
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9269828990543478,
+      "learning_rate": 1.5974904233449167e-06,
+      "loss": 0.7978,
+      "step": 11657
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421414503936807,
+      "learning_rate": 1.59673101430609e-06,
+      "loss": 0.7528,
+      "step": 11658
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1061271284694054,
+      "learning_rate": 1.595971751513311e-06,
+      "loss": 0.5758,
+      "step": 11659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1991034040444295,
+      "learning_rate": 1.5952126349992076e-06,
+      "loss": 1.0267,
+      "step": 11660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.142483416826919,
+      "learning_rate": 1.5944536647963998e-06,
+      "loss": 0.8497,
+      "step": 11661
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9830775418223563,
+      "learning_rate": 1.5936948409375007e-06,
+      "loss": 0.6941,
+      "step": 11662
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0740725784265845,
+      "learning_rate": 1.5929361634551227e-06,
+      "loss": 0.9115,
+      "step": 11663
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2394617236985408,
+      "learning_rate": 1.5921776323818655e-06,
+      "loss": 0.7288,
+      "step": 11664
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4854201543805523,
+      "learning_rate": 1.5914192477503248e-06,
+      "loss": 0.8791,
+      "step": 11665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9248309519537719,
+      "learning_rate": 1.5906610095930908e-06,
+      "loss": 0.6367,
+      "step": 11666
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3306449524710975,
+      "learning_rate": 1.589902917942746e-06,
+      "loss": 0.9247,
+      "step": 11667
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3495807944498273,
+      "learning_rate": 1.5891449728318675e-06,
+      "loss": 0.8686,
+      "step": 11668
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0308596146514013,
+      "learning_rate": 1.5883871742930257e-06,
+      "loss": 0.683,
+      "step": 11669
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2230027767398044,
+      "learning_rate": 1.5876295223587835e-06,
+      "loss": 0.7009,
+      "step": 11670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1121425671220564,
+      "learning_rate": 1.586872017061704e-06,
+      "loss": 0.8247,
+      "step": 11671
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.144686894919765,
+      "learning_rate": 1.5861146584343318e-06,
+      "loss": 0.7485,
+      "step": 11672
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.106377731121101,
+      "learning_rate": 1.585357446509218e-06,
+      "loss": 0.6232,
+      "step": 11673
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.354971363885519,
+      "learning_rate": 1.5846003813188993e-06,
+      "loss": 0.7388,
+      "step": 11674
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.285205491024823,
+      "learning_rate": 1.5838434628959081e-06,
+      "loss": 0.754,
+      "step": 11675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.29824068687438,
+      "learning_rate": 1.5830866912727722e-06,
+      "loss": 0.7442,
+      "step": 11676
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0935256824869612,
+      "learning_rate": 1.582330066482009e-06,
+      "loss": 0.9385,
+      "step": 11677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2549730047420635,
+      "learning_rate": 1.5815735885561385e-06,
+      "loss": 0.7055,
+      "step": 11678
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0736959975534273,
+      "learning_rate": 1.5808172575276615e-06,
+      "loss": 0.8494,
+      "step": 11679
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1953257333642306,
+      "learning_rate": 1.5800610734290805e-06,
+      "loss": 0.8442,
+      "step": 11680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9028095077545419,
+      "learning_rate": 1.5793050362928935e-06,
+      "loss": 0.6695,
+      "step": 11681
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6366401995600364,
+      "learning_rate": 1.5785491461515878e-06,
+      "loss": 0.8388,
+      "step": 11682
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3612556488395797,
+      "learning_rate": 1.5777934030376445e-06,
+      "loss": 0.7559,
+      "step": 11683
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.871535669013075,
+      "learning_rate": 1.5770378069835412e-06,
+      "loss": 0.7614,
+      "step": 11684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0915805957140985,
+      "learning_rate": 1.5762823580217463e-06,
+      "loss": 0.8018,
+      "step": 11685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.432345143284952,
+      "learning_rate": 1.5755270561847236e-06,
+      "loss": 0.804,
+      "step": 11686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.01231794400258,
+      "learning_rate": 1.5747719015049294e-06,
+      "loss": 0.8648,
+      "step": 11687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.024273647387123,
+      "learning_rate": 1.5740168940148137e-06,
+      "loss": 0.7537,
+      "step": 11688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0261403166405503,
+      "learning_rate": 1.5732620337468258e-06,
+      "loss": 0.6638,
+      "step": 11689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.842657867694939,
+      "learning_rate": 1.5725073207333963e-06,
+      "loss": 0.8813,
+      "step": 11690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0557828764674513,
+      "learning_rate": 1.5717527550069628e-06,
+      "loss": 0.6515,
+      "step": 11691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1839633321165803,
+      "learning_rate": 1.5709983365999481e-06,
+      "loss": 0.7633,
+      "step": 11692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8900897812875916,
+      "learning_rate": 1.570244065544772e-06,
+      "loss": 0.8022,
+      "step": 11693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.766169210576413,
+      "learning_rate": 1.5694899418738462e-06,
+      "loss": 0.8055,
+      "step": 11694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1479009090789585,
+      "learning_rate": 1.568735965619577e-06,
+      "loss": 0.7811,
+      "step": 11695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.213712065571078,
+      "learning_rate": 1.5679821368143678e-06,
+      "loss": 0.9002,
+      "step": 11696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1331961050250743,
+      "learning_rate": 1.5672284554906087e-06,
+      "loss": 0.6557,
+      "step": 11697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6509926520629743,
+      "learning_rate": 1.5664749216806858e-06,
+      "loss": 0.6333,
+      "step": 11698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.409419823143706,
+      "learning_rate": 1.5657215354169841e-06,
+      "loss": 0.7882,
+      "step": 11699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7594846245842146,
+      "learning_rate": 1.564968296731877e-06,
+      "loss": 1.0087,
+      "step": 11700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0656619603140176,
+      "learning_rate": 1.564215205657732e-06,
+      "loss": 0.7787,
+      "step": 11701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0610626565317616,
+      "learning_rate": 1.563462262226912e-06,
+      "loss": 0.7304,
+      "step": 11702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5111973759960775,
+      "learning_rate": 1.5627094664717719e-06,
+      "loss": 0.7813,
+      "step": 11703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.378532652733307,
+      "learning_rate": 1.561956818424661e-06,
+      "loss": 0.9478,
+      "step": 11704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109666975005324,
+      "learning_rate": 1.561204318117921e-06,
+      "loss": 1.0176,
+      "step": 11705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.150201350757961,
+      "learning_rate": 1.5604519655838918e-06,
+      "loss": 0.737,
+      "step": 11706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.738172485558021,
+      "learning_rate": 1.5596997608549036e-06,
+      "loss": 0.7459,
+      "step": 11707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4206707544405672,
+      "learning_rate": 1.5589477039632755e-06,
+      "loss": 0.8825,
+      "step": 11708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9930531836302081,
+      "learning_rate": 1.5581957949413295e-06,
+      "loss": 0.7727,
+      "step": 11709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9918925258225302,
+      "learning_rate": 1.557444033821376e-06,
+      "loss": 0.6037,
+      "step": 11710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6530654400790272,
+      "learning_rate": 1.5566924206357187e-06,
+      "loss": 0.7106,
+      "step": 11711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7193038524251003,
+      "learning_rate": 1.5559409554166572e-06,
+      "loss": 0.8085,
+      "step": 11712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2402311046022148,
+      "learning_rate": 1.5551896381964814e-06,
+      "loss": 0.69,
+      "step": 11713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.315981871355234,
+      "learning_rate": 1.554438469007482e-06,
+      "loss": 0.7076,
+      "step": 11714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.763819630547147,
+      "learning_rate": 1.5536874478819325e-06,
+      "loss": 0.9343,
+      "step": 11715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9248333479304829,
+      "learning_rate": 1.5529365748521097e-06,
+      "loss": 0.7739,
+      "step": 11716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.314872126510811,
+      "learning_rate": 1.5521858499502794e-06,
+      "loss": 0.8333,
+      "step": 11717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28219376407491,
+      "learning_rate": 1.5514352732087024e-06,
+      "loss": 0.8002,
+      "step": 11718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0540571801699876,
+      "learning_rate": 1.5506848446596317e-06,
+      "loss": 0.8331,
+      "step": 11719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3401535956918362,
+      "learning_rate": 1.5499345643353154e-06,
+      "loss": 0.6621,
+      "step": 11720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.343041665634448,
+      "learning_rate": 1.5491844322679944e-06,
+      "loss": 0.7769,
+      "step": 11721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0599080132807248,
+      "learning_rate": 1.5484344484899038e-06,
+      "loss": 0.6448,
+      "step": 11722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078721569776104,
+      "learning_rate": 1.5476846130332702e-06,
+      "loss": 0.7145,
+      "step": 11723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.867839363024628,
+      "learning_rate": 1.546934925930319e-06,
+      "loss": 0.7492,
+      "step": 11724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4971684528324722,
+      "learning_rate": 1.5461853872132648e-06,
+      "loss": 0.7826,
+      "step": 11725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1736180027896834,
+      "learning_rate": 1.5454359969143161e-06,
+      "loss": 0.7971,
+      "step": 11726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1651113223312946,
+      "learning_rate": 1.544686755065677e-06,
+      "loss": 0.8372,
+      "step": 11727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.116386427566427,
+      "learning_rate": 1.5439376616995432e-06,
+      "loss": 0.6426,
+      "step": 11728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.650814686925478,
+      "learning_rate": 1.5431887168481051e-06,
+      "loss": 0.7878,
+      "step": 11729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4804968769531826,
+      "learning_rate": 1.5424399205435464e-06,
+      "loss": 0.8263,
+      "step": 11730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3373191601888736,
+      "learning_rate": 1.5416912728180433e-06,
+      "loss": 0.9431,
+      "step": 11731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1902996255596872,
+      "learning_rate": 1.5409427737037713e-06,
+      "loss": 0.5916,
+      "step": 11732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9420320820650778,
+      "learning_rate": 1.540194423232889e-06,
+      "loss": 0.7396,
+      "step": 11733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9421550176897697,
+      "learning_rate": 1.5394462214375593e-06,
+      "loss": 0.7602,
+      "step": 11734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1297143581922326,
+      "learning_rate": 1.5386981683499325e-06,
+      "loss": 0.8303,
+      "step": 11735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.474956279256187,
+      "learning_rate": 1.5379502640021544e-06,
+      "loss": 0.8077,
+      "step": 11736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5736108325254254,
+      "learning_rate": 1.537202508426363e-06,
+      "loss": 0.8319,
+      "step": 11737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.99466974667524,
+      "learning_rate": 1.5364549016546925e-06,
+      "loss": 0.6868,
+      "step": 11738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.429257385194195,
+      "learning_rate": 1.5357074437192688e-06,
+      "loss": 0.7382,
+      "step": 11739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.690557730542317,
+      "learning_rate": 1.5349601346522115e-06,
+      "loss": 0.9525,
+      "step": 11740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3892139959752408,
+      "learning_rate": 1.5342129744856327e-06,
+      "loss": 0.8586,
+      "step": 11741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.246417079789784,
+      "learning_rate": 1.5334659632516424e-06,
+      "loss": 0.7111,
+      "step": 11742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.04953755197335,
+      "learning_rate": 1.5327191009823405e-06,
+      "loss": 0.9248,
+      "step": 11743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0561000788993378,
+      "learning_rate": 1.5319723877098202e-06,
+      "loss": 0.6407,
+      "step": 11744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4121830465732867,
+      "learning_rate": 1.531225823466171e-06,
+      "loss": 0.861,
+      "step": 11745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.943511952710782,
+      "learning_rate": 1.5304794082834713e-06,
+      "loss": 0.7851,
+      "step": 11746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.033895002138429,
+      "learning_rate": 1.5297331421938017e-06,
+      "loss": 0.7477,
+      "step": 11747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.411275563605233,
+      "learning_rate": 1.5289870252292265e-06,
+      "loss": 0.8111,
+      "step": 11748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2200501396034522,
+      "learning_rate": 1.5282410574218072e-06,
+      "loss": 0.6858,
+      "step": 11749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2255861364160716,
+      "learning_rate": 1.5274952388036052e-06,
+      "loss": 0.838,
+      "step": 11750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0448790604560108,
+      "learning_rate": 1.5267495694066631e-06,
+      "loss": 0.9144,
+      "step": 11751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.451501943948351,
+      "learning_rate": 1.5260040492630286e-06,
+      "loss": 0.7553,
+      "step": 11752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.971732693388332,
+      "learning_rate": 1.5252586784047374e-06,
+      "loss": 0.5885,
+      "step": 11753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8793257767330749,
+      "learning_rate": 1.5245134568638197e-06,
+      "loss": 0.7663,
+      "step": 11754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1827015542298267,
+      "learning_rate": 1.523768384672299e-06,
+      "loss": 0.8335,
+      "step": 11755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.999809718085218,
+      "learning_rate": 1.5230234618621908e-06,
+      "loss": 0.6811,
+      "step": 11756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2797141436354273,
+      "learning_rate": 1.5222786884655117e-06,
+      "loss": 0.6387,
+      "step": 11757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.127294083441729,
+      "learning_rate": 1.5215340645142607e-06,
+      "loss": 0.7942,
+      "step": 11758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9122480268987005,
+      "learning_rate": 1.5207895900404363e-06,
+      "loss": 0.8419,
+      "step": 11759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.49208328055085,
+      "learning_rate": 1.520045265076034e-06,
+      "loss": 0.7412,
+      "step": 11760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1584029139726524,
+      "learning_rate": 1.5193010896530364e-06,
+      "loss": 0.6046,
+      "step": 11761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0507186824275307,
+      "learning_rate": 1.518557063803423e-06,
+      "loss": 0.7953,
+      "step": 11762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.699870078408754,
+      "learning_rate": 1.5178131875591662e-06,
+      "loss": 0.7872,
+      "step": 11763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1774852643070464,
+      "learning_rate": 1.5170694609522306e-06,
+      "loss": 0.7314,
+      "step": 11764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2348866456643242,
+      "learning_rate": 1.51632588401458e-06,
+      "loss": 0.6141,
+      "step": 11765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.796387734651648,
+      "learning_rate": 1.5155824567781618e-06,
+      "loss": 0.8628,
+      "step": 11766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8236303372563016,
+      "learning_rate": 1.5148391792749272e-06,
+      "loss": 0.8383,
+      "step": 11767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8419343292037438,
+      "learning_rate": 1.5140960515368165e-06,
+      "loss": 0.8332,
+      "step": 11768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9861032420533655,
+      "learning_rate": 1.5133530735957586e-06,
+      "loss": 0.9382,
+      "step": 11769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.820971430643831,
+      "learning_rate": 1.512610245483686e-06,
+      "loss": 0.9357,
+      "step": 11770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.24329939787031,
+      "learning_rate": 1.5118675672325184e-06,
+      "loss": 0.8016,
+      "step": 11771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4048175381101533,
+      "learning_rate": 1.5111250388741693e-06,
+      "loss": 0.7385,
+      "step": 11772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3805236917994828,
+      "learning_rate": 1.510382660440548e-06,
+      "loss": 0.8097,
+      "step": 11773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9802087296646151,
+      "learning_rate": 1.5096404319635533e-06,
+      "loss": 0.7513,
+      "step": 11774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2531205753170487,
+      "learning_rate": 1.5088983534750867e-06,
+      "loss": 0.9287,
+      "step": 11775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2207927063175203,
+      "learning_rate": 1.5081564250070291e-06,
+      "loss": 0.6001,
+      "step": 11776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6533999507537023,
+      "learning_rate": 1.5074146465912687e-06,
+      "loss": 0.6969,
+      "step": 11777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0242663816314272,
+      "learning_rate": 1.506673018259679e-06,
+      "loss": 0.6823,
+      "step": 11778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.090717687130839,
+      "learning_rate": 1.50593154004413e-06,
+      "loss": 0.669,
+      "step": 11779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9917606716300486,
+      "learning_rate": 1.505190211976484e-06,
+      "loss": 0.6555,
+      "step": 11780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2268310156849456,
+      "learning_rate": 1.5044490340885987e-06,
+      "loss": 0.8598,
+      "step": 11781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.29413589630399,
+      "learning_rate": 1.5037080064123212e-06,
+      "loss": 0.8512,
+      "step": 11782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1283059586453574,
+      "learning_rate": 1.5029671289795007e-06,
+      "loss": 0.7436,
+      "step": 11783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0880613237033918,
+      "learning_rate": 1.502226401821968e-06,
+      "loss": 0.7045,
+      "step": 11784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6396106211016908,
+      "learning_rate": 1.501485824971558e-06,
+      "loss": 0.8361,
+      "step": 11785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.279218666490566,
+      "learning_rate": 1.5007453984600956e-06,
+      "loss": 0.7882,
+      "step": 11786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2937057883823573,
+      "learning_rate": 1.5000051223193928e-06,
+      "loss": 0.9365,
+      "step": 11787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.535592938408729,
+      "learning_rate": 1.4992649965812673e-06,
+      "loss": 0.8504,
+      "step": 11788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3668824832232094,
+      "learning_rate": 1.498525021277521e-06,
+      "loss": 0.8092,
+      "step": 11789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2662460875564423,
+      "learning_rate": 1.497785196439953e-06,
+      "loss": 0.7668,
+      "step": 11790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1480825054581725,
+      "learning_rate": 1.497045522100355e-06,
+      "loss": 0.6976,
+      "step": 11791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9975831369235408,
+      "learning_rate": 1.4963059982905104e-06,
+      "loss": 0.7775,
+      "step": 11792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0808644527994078,
+      "learning_rate": 1.4955666250422035e-06,
+      "loss": 0.5732,
+      "step": 11793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5587395855211166,
+      "learning_rate": 1.4948274023872005e-06,
+      "loss": 0.905,
+      "step": 11794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4971924377898995,
+      "learning_rate": 1.4940883303572724e-06,
+      "loss": 0.6836,
+      "step": 11795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.059643970587331,
+      "learning_rate": 1.493349408984176e-06,
+      "loss": 0.6815,
+      "step": 11796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.984812638905341,
+      "learning_rate": 1.4926106382996658e-06,
+      "loss": 0.6248,
+      "step": 11797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1406174620366696,
+      "learning_rate": 1.491872018335488e-06,
+      "loss": 0.895,
+      "step": 11798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.004195460696544,
+      "learning_rate": 1.4911335491233818e-06,
+      "loss": 0.7579,
+      "step": 11799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3261734113671206,
+      "learning_rate": 1.4903952306950808e-06,
+      "loss": 0.7119,
+      "step": 11800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0981793624711478,
+      "learning_rate": 1.489657063082316e-06,
+      "loss": 0.6314,
+      "step": 11801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.822400896133419,
+      "learning_rate": 1.4889190463168019e-06,
+      "loss": 0.7695,
+      "step": 11802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.044716709468024,
+      "learning_rate": 1.4881811804302577e-06,
+      "loss": 0.6887,
+      "step": 11803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.906767094846111,
+      "learning_rate": 1.487443465454389e-06,
+      "loss": 0.7718,
+      "step": 11804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.540702129229706,
+      "learning_rate": 1.4867059014208979e-06,
+      "loss": 0.664,
+      "step": 11805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8845795468755726,
+      "learning_rate": 1.4859684883614788e-06,
+      "loss": 0.7997,
+      "step": 11806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6630561961517576,
+      "learning_rate": 1.4852312263078183e-06,
+      "loss": 0.8174,
+      "step": 11807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0872482194035258,
+      "learning_rate": 1.4844941152916026e-06,
+      "loss": 0.574,
+      "step": 11808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2758097675178584,
+      "learning_rate": 1.483757155344503e-06,
+      "loss": 0.7754,
+      "step": 11809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4738076583347954,
+      "learning_rate": 1.4830203464981879e-06,
+      "loss": 0.7589,
+      "step": 11810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9069465482549661,
+      "learning_rate": 1.4822836887843244e-06,
+      "loss": 0.8287,
+      "step": 11811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.14458756054688,
+      "learning_rate": 1.4815471822345618e-06,
+      "loss": 0.8202,
+      "step": 11812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7302832514301496,
+      "learning_rate": 1.4808108268805549e-06,
+      "loss": 0.6985,
+      "step": 11813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.229811161223776,
+      "learning_rate": 1.4800746227539437e-06,
+      "loss": 0.7065,
+      "step": 11814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2354121971987313,
+      "learning_rate": 1.479338569886366e-06,
+      "loss": 0.6868,
+      "step": 11815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5103895302292996,
+      "learning_rate": 1.47860266830945e-06,
+      "loss": 0.8755,
+      "step": 11816
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.405924765813395,
+      "learning_rate": 1.4778669180548195e-06,
+      "loss": 0.7774,
+      "step": 11817
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7987824297343473,
+      "learning_rate": 1.4771313191540927e-06,
+      "loss": 0.8691,
+      "step": 11818
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0671720371857907,
+      "learning_rate": 1.4763958716388798e-06,
+      "loss": 0.7754,
+      "step": 11819
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2659948389212237,
+      "learning_rate": 1.4756605755407816e-06,
+      "loss": 0.7351,
+      "step": 11820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0818440169517367,
+      "learning_rate": 1.474925430891399e-06,
+      "loss": 0.8285,
+      "step": 11821
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1041964941954814,
+      "learning_rate": 1.4741904377223215e-06,
+      "loss": 0.6647,
+      "step": 11822
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6854203131516894,
+      "learning_rate": 1.473455596065133e-06,
+      "loss": 0.6712,
+      "step": 11823
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.318078115372113,
+      "learning_rate": 1.4727209059514114e-06,
+      "loss": 0.8064,
+      "step": 11824
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0314617807505395,
+      "learning_rate": 1.4719863674127272e-06,
+      "loss": 0.6854,
+      "step": 11825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7984947652917385,
+      "learning_rate": 1.4712519804806485e-06,
+      "loss": 0.7979,
+      "step": 11826
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.828650418915175,
+      "learning_rate": 1.4705177451867287e-06,
+      "loss": 0.7521,
+      "step": 11827
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9231608068495039,
+      "learning_rate": 1.4697836615625233e-06,
+      "loss": 0.8246,
+      "step": 11828
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3310817416101146,
+      "learning_rate": 1.4690497296395773e-06,
+      "loss": 0.7461,
+      "step": 11829
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.420855121332244,
+      "learning_rate": 1.4683159494494259e-06,
+      "loss": 1.0475,
+      "step": 11830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0991048473765983,
+      "learning_rate": 1.4675823210236046e-06,
+      "loss": 0.7011,
+      "step": 11831
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.008765583631384,
+      "learning_rate": 1.4668488443936385e-06,
+      "loss": 0.5788,
+      "step": 11832
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5418047195256395,
+      "learning_rate": 1.466115519591046e-06,
+      "loss": 0.7757,
+      "step": 11833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1854551451556463,
+      "learning_rate": 1.46538234664734e-06,
+      "loss": 0.7567,
+      "step": 11834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.030800503577493,
+      "learning_rate": 1.4646493255940253e-06,
+      "loss": 0.862,
+      "step": 11835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9220252104277775,
+      "learning_rate": 1.4639164564626041e-06,
+      "loss": 0.789,
+      "step": 11836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8751665967559648,
+      "learning_rate": 1.4631837392845694e-06,
+      "loss": 0.7119,
+      "step": 11837
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8816673941050395,
+      "learning_rate": 1.4624511740914032e-06,
+      "loss": 0.9238,
+      "step": 11838
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.235272623691521,
+      "learning_rate": 1.4617187609145906e-06,
+      "loss": 0.5948,
+      "step": 11839
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9858457019588174,
+      "learning_rate": 1.4609864997856026e-06,
+      "loss": 0.7922,
+      "step": 11840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.477943657960904,
+      "learning_rate": 1.4602543907359068e-06,
+      "loss": 0.6658,
+      "step": 11841
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03057117710403,
+      "learning_rate": 1.4595224337969633e-06,
+      "loss": 0.965,
+      "step": 11842
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0304130509529483,
+      "learning_rate": 1.4587906290002242e-06,
+      "loss": 0.5961,
+      "step": 11843
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5101981679830456,
+      "learning_rate": 1.4580589763771413e-06,
+      "loss": 0.9123,
+      "step": 11844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9382891585975313,
+      "learning_rate": 1.4573274759591499e-06,
+      "loss": 1.0831,
+      "step": 11845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.376232495851072,
+      "learning_rate": 1.456596127777688e-06,
+      "loss": 0.8775,
+      "step": 11846
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.827938536602597,
+      "learning_rate": 1.4558649318641837e-06,
+      "loss": 0.7595,
+      "step": 11847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.852002720421721,
+      "learning_rate": 1.4551338882500538e-06,
+      "loss": 0.8924,
+      "step": 11848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.105766221044859,
+      "learning_rate": 1.4544029969667167e-06,
+      "loss": 0.9786,
+      "step": 11849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9891085556592925,
+      "learning_rate": 1.4536722580455792e-06,
+      "loss": 1.0498,
+      "step": 11850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0192126972236593,
+      "learning_rate": 1.4529416715180434e-06,
+      "loss": 0.821,
+      "step": 11851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2656723412597257,
+      "learning_rate": 1.452211237415503e-06,
+      "loss": 0.6385,
+      "step": 11852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1728528220265835,
+      "learning_rate": 1.451480955769346e-06,
+      "loss": 0.8466,
+      "step": 11853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0367242312647935,
+      "learning_rate": 1.4507508266109565e-06,
+      "loss": 0.8384,
+      "step": 11854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.636427120962915,
+      "learning_rate": 1.4500208499717088e-06,
+      "loss": 0.9121,
+      "step": 11855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0002165738433595,
+      "learning_rate": 1.4492910258829717e-06,
+      "loss": 1.0146,
+      "step": 11856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.963366491720887,
+      "learning_rate": 1.4485613543761068e-06,
+      "loss": 0.886,
+      "step": 11857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7907559884132906,
+      "learning_rate": 1.44783183548247e-06,
+      "loss": 0.7987,
+      "step": 11858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1877688671122577,
+      "learning_rate": 1.4471024692334101e-06,
+      "loss": 0.8786,
+      "step": 11859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3672824963108776,
+      "learning_rate": 1.44637325566027e-06,
+      "loss": 0.8136,
+      "step": 11860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.141095541183427,
+      "learning_rate": 1.445644194794384e-06,
+      "loss": 0.7707,
+      "step": 11861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.651446408113701,
+      "learning_rate": 1.4449152866670858e-06,
+      "loss": 0.753,
+      "step": 11862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1456741758354174,
+      "learning_rate": 1.4441865313096926e-06,
+      "loss": 0.715,
+      "step": 11863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8968741388733599,
+      "learning_rate": 1.4434579287535244e-06,
+      "loss": 0.8465,
+      "step": 11864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9771178874400814,
+      "learning_rate": 1.4427294790298902e-06,
+      "loss": 0.5079,
+      "step": 11865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9939959089341976,
+      "learning_rate": 1.4420011821700924e-06,
+      "loss": 0.653,
+      "step": 11866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.358247469544141,
+      "learning_rate": 1.4412730382054285e-06,
+      "loss": 0.7877,
+      "step": 11867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8149736642519714,
+      "learning_rate": 1.4405450471671861e-06,
+      "loss": 0.7891,
+      "step": 11868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.100104693528016,
+      "learning_rate": 1.439817209086653e-06,
+      "loss": 0.7054,
+      "step": 11869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1646906324547834,
+      "learning_rate": 1.4390895239951025e-06,
+      "loss": 0.6669,
+      "step": 11870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1893880426652954,
+      "learning_rate": 1.4383619919238035e-06,
+      "loss": 0.83,
+      "step": 11871
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.205674583787908,
+      "learning_rate": 1.4376346129040243e-06,
+      "loss": 0.8522,
+      "step": 11872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.370534949796821,
+      "learning_rate": 1.4369073869670186e-06,
+      "loss": 0.7904,
+      "step": 11873
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5689235532987245,
+      "learning_rate": 1.4361803141440384e-06,
+      "loss": 0.6157,
+      "step": 11874
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0836974490678237,
+      "learning_rate": 1.4354533944663274e-06,
+      "loss": 0.9266,
+      "step": 11875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1526562101990367,
+      "learning_rate": 1.4347266279651224e-06,
+      "loss": 0.6165,
+      "step": 11876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3358725646890663,
+      "learning_rate": 1.4340000146716544e-06,
+      "loss": 0.8713,
+      "step": 11877
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3440598895929012,
+      "learning_rate": 1.4332735546171466e-06,
+      "loss": 0.7428,
+      "step": 11878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1836420567085932,
+      "learning_rate": 1.432547247832819e-06,
+      "loss": 0.7698,
+      "step": 11879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0773860170066945,
+      "learning_rate": 1.431821094349883e-06,
+      "loss": 0.932,
+      "step": 11880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.785906430999547,
+      "learning_rate": 1.4310950941995378e-06,
+      "loss": 0.6275,
+      "step": 11881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.551077226519704,
+      "learning_rate": 1.4303692474129866e-06,
+      "loss": 0.8776,
+      "step": 11882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8018873616943383,
+      "learning_rate": 1.4296435540214181e-06,
+      "loss": 0.9635,
+      "step": 11883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8927049836962877,
+      "learning_rate": 1.4289180140560189e-06,
+      "loss": 0.7856,
+      "step": 11884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9198483494236744,
+      "learning_rate": 1.428192627547965e-06,
+      "loss": 0.6935,
+      "step": 11885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2901204744641674,
+      "learning_rate": 1.4274673945284278e-06,
+      "loss": 0.7032,
+      "step": 11886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1631250348535547,
+      "learning_rate": 1.4267423150285758e-06,
+      "loss": 0.8875,
+      "step": 11887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0647692010927643,
+      "learning_rate": 1.4260173890795631e-06,
+      "loss": 0.7885,
+      "step": 11888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1121280281453414,
+      "learning_rate": 1.4252926167125413e-06,
+      "loss": 0.6413,
+      "step": 11889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.225167028710076,
+      "learning_rate": 1.4245679979586591e-06,
+      "loss": 0.8716,
+      "step": 11890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.599848951906125,
+      "learning_rate": 1.4238435328490525e-06,
+      "loss": 0.641,
+      "step": 11891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1357974682640153,
+      "learning_rate": 1.4231192214148538e-06,
+      "loss": 0.8074,
+      "step": 11892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2671170211179095,
+      "learning_rate": 1.422395063687188e-06,
+      "loss": 0.8401,
+      "step": 11893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.080052127376581,
+      "learning_rate": 1.421671059697175e-06,
+      "loss": 0.6931,
+      "step": 11894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2705354386915664,
+      "learning_rate": 1.4209472094759252e-06,
+      "loss": 0.6901,
+      "step": 11895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.562080976717904,
+      "learning_rate": 1.4202235130545432e-06,
+      "loss": 0.961,
+      "step": 11896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1232852041809394,
+      "learning_rate": 1.4194999704641315e-06,
+      "loss": 0.6705,
+      "step": 11897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.574728534629309,
+      "learning_rate": 1.4187765817357812e-06,
+      "loss": 0.8399,
+      "step": 11898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1320839669421656,
+      "learning_rate": 1.418053346900574e-06,
+      "loss": 0.6571,
+      "step": 11899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4280875727748485,
+      "learning_rate": 1.4173302659895938e-06,
+      "loss": 0.7092,
+      "step": 11900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.065321024679428,
+      "learning_rate": 1.4166073390339113e-06,
+      "loss": 0.9594,
+      "step": 11901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1017107310782186,
+      "learning_rate": 1.4158845660645915e-06,
+      "loss": 0.7853,
+      "step": 11902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.352223033320983,
+      "learning_rate": 1.4151619471126943e-06,
+      "loss": 0.8421,
+      "step": 11903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.368560993715672,
+      "learning_rate": 1.4144394822092712e-06,
+      "loss": 0.9472,
+      "step": 11904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15239445524641,
+      "learning_rate": 1.4137171713853714e-06,
+      "loss": 0.8615,
+      "step": 11905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0781366630360205,
+      "learning_rate": 1.412995014672029e-06,
+      "loss": 0.6904,
+      "step": 11906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0541487807177357,
+      "learning_rate": 1.4122730121002808e-06,
+      "loss": 0.5993,
+      "step": 11907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.901691994857888,
+      "learning_rate": 1.4115511637011514e-06,
+      "loss": 0.8404,
+      "step": 11908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109870921130707,
+      "learning_rate": 1.4108294695056606e-06,
+      "loss": 0.7189,
+      "step": 11909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1351954732801746,
+      "learning_rate": 1.4101079295448217e-06,
+      "loss": 0.671,
+      "step": 11910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.205068077439234,
+      "learning_rate": 1.4093865438496395e-06,
+      "loss": 0.5924,
+      "step": 11911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.974736755184517,
+      "learning_rate": 1.408665312451114e-06,
+      "loss": 0.6868,
+      "step": 11912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.108490275447065,
+      "learning_rate": 1.4079442353802392e-06,
+      "loss": 0.7665,
+      "step": 11913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.427354630942883,
+      "learning_rate": 1.4072233126679985e-06,
+      "loss": 0.7411,
+      "step": 11914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2276962075994726,
+      "learning_rate": 1.406502544345375e-06,
+      "loss": 0.5558,
+      "step": 11915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.275312976454433,
+      "learning_rate": 1.40578193044334e-06,
+      "loss": 0.6542,
+      "step": 11916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.326360442041502,
+      "learning_rate": 1.4050614709928612e-06,
+      "loss": 0.7672,
+      "step": 11917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8276844180199692,
+      "learning_rate": 1.4043411660248967e-06,
+      "loss": 0.8291,
+      "step": 11918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0401950054030806,
+      "learning_rate": 1.4036210155703989e-06,
+      "loss": 0.6542,
+      "step": 11919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3412450304637833,
+      "learning_rate": 1.4029010196603182e-06,
+      "loss": 0.726,
+      "step": 11920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.583769156303167,
+      "learning_rate": 1.4021811783255912e-06,
+      "loss": 0.6131,
+      "step": 11921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.753285191737703,
+      "learning_rate": 1.4014614915971498e-06,
+      "loss": 0.7376,
+      "step": 11922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4179177305766992,
+      "learning_rate": 1.4007419595059252e-06,
+      "loss": 0.7992,
+      "step": 11923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.900721312790883,
+      "learning_rate": 1.4000225820828317e-06,
+      "loss": 0.6544,
+      "step": 11924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1837304466699807,
+      "learning_rate": 1.3993033593587874e-06,
+      "loss": 0.7358,
+      "step": 11925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7865690370461595,
+      "learning_rate": 1.3985842913646963e-06,
+      "loss": 0.6496,
+      "step": 11926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.184361629477825,
+      "learning_rate": 1.3978653781314595e-06,
+      "loss": 0.9861,
+      "step": 11927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1980001793485755,
+      "learning_rate": 1.3971466196899697e-06,
+      "loss": 0.7322,
+      "step": 11928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3045416002385157,
+      "learning_rate": 1.3964280160711119e-06,
+      "loss": 0.7112,
+      "step": 11929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.438137495772439,
+      "learning_rate": 1.3957095673057708e-06,
+      "loss": 1.0835,
+      "step": 11930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1174805871530586,
+      "learning_rate": 1.3949912734248156e-06,
+      "loss": 0.861,
+      "step": 11931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5017092098284808,
+      "learning_rate": 1.3942731344591127e-06,
+      "loss": 0.795,
+      "step": 11932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0646320608990725,
+      "learning_rate": 1.3935551504395246e-06,
+      "loss": 0.8058,
+      "step": 11933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.566522036879435,
+      "learning_rate": 1.3928373213969038e-06,
+      "loss": 0.8927,
+      "step": 11934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2337644879524214,
+      "learning_rate": 1.3921196473620975e-06,
+      "loss": 0.6867,
+      "step": 11935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8086669576425622,
+      "learning_rate": 1.3914021283659445e-06,
+      "loss": 0.7357,
+      "step": 11936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1769624559519851,
+      "learning_rate": 1.3906847644392773e-06,
+      "loss": 0.7618,
+      "step": 11937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1756535313912497,
+      "learning_rate": 1.3899675556129272e-06,
+      "loss": 0.6725,
+      "step": 11938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7101501050833012,
+      "learning_rate": 1.38925050191771e-06,
+      "loss": 0.7235,
+      "step": 11939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2126101197766537,
+      "learning_rate": 1.3885336033844388e-06,
+      "loss": 0.8096,
+      "step": 11940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.262842141599081,
+      "learning_rate": 1.3878168600439246e-06,
+      "loss": 0.7707,
+      "step": 11941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2649377087968174,
+      "learning_rate": 1.3871002719269616e-06,
+      "loss": 0.7853,
+      "step": 11942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7986622558958218,
+      "learning_rate": 1.386383839064348e-06,
+      "loss": 0.6799,
+      "step": 11943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9376638694655783,
+      "learning_rate": 1.3856675614868687e-06,
+      "loss": 0.9453,
+      "step": 11944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8552808406976893,
+      "learning_rate": 1.3849514392253038e-06,
+      "loss": 0.7036,
+      "step": 11945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8634378458065721,
+      "learning_rate": 1.3842354723104267e-06,
+      "loss": 0.6895,
+      "step": 11946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.348056084117558,
+      "learning_rate": 1.3835196607730028e-06,
+      "loss": 0.8014,
+      "step": 11947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.373780641116316,
+      "learning_rate": 1.3828040046437963e-06,
+      "loss": 0.7017,
+      "step": 11948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.129454323266344,
+      "learning_rate": 1.3820885039535564e-06,
+      "loss": 0.7656,
+      "step": 11949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1284408580116168,
+      "learning_rate": 1.3813731587330291e-06,
+      "loss": 0.7797,
+      "step": 11950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.426400927086544,
+      "learning_rate": 1.380657969012958e-06,
+      "loss": 0.8359,
+      "step": 11951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.432321488078757,
+      "learning_rate": 1.3799429348240746e-06,
+      "loss": 0.7031,
+      "step": 11952
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.171281810336406,
+      "learning_rate": 1.3792280561971056e-06,
+      "loss": 0.7722,
+      "step": 11953
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0801417175022723,
+      "learning_rate": 1.378513333162771e-06,
+      "loss": 0.6514,
+      "step": 11954
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.982527024444471,
+      "learning_rate": 1.3777987657517822e-06,
+      "loss": 0.6803,
+      "step": 11955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2095487589152927,
+      "learning_rate": 1.3770843539948508e-06,
+      "loss": 0.777,
+      "step": 11956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.142048553927815,
+      "learning_rate": 1.3763700979226701e-06,
+      "loss": 0.6609,
+      "step": 11957
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1315346429779354,
+      "learning_rate": 1.3756559975659385e-06,
+      "loss": 0.6846,
+      "step": 11958
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6289473688607934,
+      "learning_rate": 1.3749420529553414e-06,
+      "loss": 0.6941,
+      "step": 11959
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7626476133874154,
+      "learning_rate": 1.3742282641215549e-06,
+      "loss": 0.7938,
+      "step": 11960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0323044550107103,
+      "learning_rate": 1.3735146310952568e-06,
+      "loss": 0.7547,
+      "step": 11961
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5589660543640753,
+      "learning_rate": 1.3728011539071112e-06,
+      "loss": 0.7301,
+      "step": 11962
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.07167980601058,
+      "learning_rate": 1.3720878325877785e-06,
+      "loss": 0.8283,
+      "step": 11963
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0424595287069303,
+      "learning_rate": 1.3713746671679112e-06,
+      "loss": 0.7283,
+      "step": 11964
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.181487357600582,
+      "learning_rate": 1.3706616576781545e-06,
+      "loss": 0.8329,
+      "step": 11965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2416233029246984,
+      "learning_rate": 1.3699488041491526e-06,
+      "loss": 0.7077,
+      "step": 11966
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.303487104571446,
+      "learning_rate": 1.3692361066115318e-06,
+      "loss": 0.85,
+      "step": 11967
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7798744937241542,
+      "learning_rate": 1.3685235650959228e-06,
+      "loss": 0.565,
+      "step": 11968
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7548797265461398,
+      "learning_rate": 1.3678111796329446e-06,
+      "loss": 0.781,
+      "step": 11969
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1264062993256703,
+      "learning_rate": 1.3670989502532089e-06,
+      "loss": 0.7681,
+      "step": 11970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5880905458279795,
+      "learning_rate": 1.3663868769873223e-06,
+      "loss": 0.6678,
+      "step": 11971
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.59447491078861,
+      "learning_rate": 1.365674959865884e-06,
+      "loss": 0.7365,
+      "step": 11972
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3898112336841653,
+      "learning_rate": 1.3649631989194856e-06,
+      "loss": 0.9878,
+      "step": 11973
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.949068620485174,
+      "learning_rate": 1.3642515941787171e-06,
+      "loss": 0.7498,
+      "step": 11974
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9832983201446222,
+      "learning_rate": 1.363540145674152e-06,
+      "loss": 0.7534,
+      "step": 11975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.036958743291282,
+      "learning_rate": 1.3628288534363664e-06,
+      "loss": 0.9698,
+      "step": 11976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.279836735454935,
+      "learning_rate": 1.362117717495926e-06,
+      "loss": 0.8075,
+      "step": 11977
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9693741464013046,
+      "learning_rate": 1.3614067378833895e-06,
+      "loss": 0.7689,
+      "step": 11978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9848240607364034,
+      "learning_rate": 1.3606959146293086e-06,
+      "loss": 0.6488,
+      "step": 11979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1996084827442106,
+      "learning_rate": 1.3599852477642283e-06,
+      "loss": 0.7152,
+      "step": 11980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.471412685512495,
+      "learning_rate": 1.3592747373186915e-06,
+      "loss": 0.903,
+      "step": 11981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1083755682591199,
+      "learning_rate": 1.358564383323226e-06,
+      "loss": 0.6171,
+      "step": 11982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.730386551593291,
+      "learning_rate": 1.357854185808357e-06,
+      "loss": 0.9715,
+      "step": 11983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0568876079156786,
+      "learning_rate": 1.3571441448046086e-06,
+      "loss": 0.6473,
+      "step": 11984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.937945348550544,
+      "learning_rate": 1.3564342603424864e-06,
+      "loss": 0.8143,
+      "step": 11985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8781181461305687,
+      "learning_rate": 1.3557245324524999e-06,
+      "loss": 0.7355,
+      "step": 11986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2088332869171854,
+      "learning_rate": 1.3550149611651463e-06,
+      "loss": 0.6931,
+      "step": 11987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5194545294898436,
+      "learning_rate": 1.3543055465109183e-06,
+      "loss": 0.9301,
+      "step": 11988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6904146353734086,
+      "learning_rate": 1.3535962885202997e-06,
+      "loss": 0.6228,
+      "step": 11989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5210945152823405,
+      "learning_rate": 1.3528871872237687e-06,
+      "loss": 0.7996,
+      "step": 11990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9548424559959905,
+      "learning_rate": 1.3521782426517988e-06,
+      "loss": 0.9044,
+      "step": 11991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1522409553138555,
+      "learning_rate": 1.3514694548348556e-06,
+      "loss": 0.7013,
+      "step": 11992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.914861498613794,
+      "learning_rate": 1.3507608238033927e-06,
+      "loss": 0.6658,
+      "step": 11993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0479068735157262,
+      "learning_rate": 1.350052349587866e-06,
+      "loss": 0.6175,
+      "step": 11994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9379447766072277,
+      "learning_rate": 1.3493440322187185e-06,
+      "loss": 0.6911,
+      "step": 11995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3765094228978696,
+      "learning_rate": 1.3486358717263887e-06,
+      "loss": 0.8299,
+      "step": 11996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6548401833891573,
+      "learning_rate": 1.3479278681413072e-06,
+      "loss": 0.6734,
+      "step": 11997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1918385299559393,
+      "learning_rate": 1.3472200214938974e-06,
+      "loss": 0.7936,
+      "step": 11998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0886673924462618,
+      "learning_rate": 1.3465123318145817e-06,
+      "loss": 0.7909,
+      "step": 11999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9446999006268693,
+      "learning_rate": 1.3458047991337664e-06,
+      "loss": 0.7433,
+      "step": 12000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9827670184858806,
+      "learning_rate": 1.3450974234818554e-06,
+      "loss": 0.7257,
+      "step": 12001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2220462248197812,
+      "learning_rate": 1.34439020488925e-06,
+      "loss": 0.6671,
+      "step": 12002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4408501733046335,
+      "learning_rate": 1.3436831433863384e-06,
+      "loss": 0.8841,
+      "step": 12003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4175693990917466,
+      "learning_rate": 1.342976239003505e-06,
+      "loss": 0.7773,
+      "step": 12004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9613844393650721,
+      "learning_rate": 1.3422694917711276e-06,
+      "loss": 0.7051,
+      "step": 12005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.356027269628845,
+      "learning_rate": 1.3415629017195763e-06,
+      "loss": 0.8641,
+      "step": 12006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.143348509817376,
+      "learning_rate": 1.3408564688792153e-06,
+      "loss": 0.6219,
+      "step": 12007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2629000576384035,
+      "learning_rate": 1.340150193280399e-06,
+      "loss": 0.6488,
+      "step": 12008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1369386827960177,
+      "learning_rate": 1.339444074953482e-06,
+      "loss": 0.749,
+      "step": 12009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4325051582292376,
+      "learning_rate": 1.3387381139288064e-06,
+      "loss": 0.7479,
+      "step": 12010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.059417435500947,
+      "learning_rate": 1.3380323102367055e-06,
+      "loss": 0.7328,
+      "step": 12011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.971797726423136,
+      "learning_rate": 1.3373266639075134e-06,
+      "loss": 0.6444,
+      "step": 12012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2604733034697664,
+      "learning_rate": 1.3366211749715519e-06,
+      "loss": 0.8835,
+      "step": 12013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.99106049083587,
+      "learning_rate": 1.335915843459137e-06,
+      "loss": 0.8357,
+      "step": 12014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9213729504622865,
+      "learning_rate": 1.335210669400579e-06,
+      "loss": 0.9026,
+      "step": 12015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5845501823070802,
+      "learning_rate": 1.3345056528261784e-06,
+      "loss": 0.9561,
+      "step": 12016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8555049326852675,
+      "learning_rate": 1.3338007937662367e-06,
+      "loss": 0.8331,
+      "step": 12017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5722574324760874,
+      "learning_rate": 1.3330960922510373e-06,
+      "loss": 0.8617,
+      "step": 12018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3035588271119225,
+      "learning_rate": 1.3323915483108662e-06,
+      "loss": 0.7455,
+      "step": 12019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4680996917285465,
+      "learning_rate": 1.3316871619759992e-06,
+      "loss": 0.7829,
+      "step": 12020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.595450386794252,
+      "learning_rate": 1.3309829332767037e-06,
+      "loss": 0.8883,
+      "step": 12021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9371303258587322,
+      "learning_rate": 1.330278862243244e-06,
+      "loss": 0.6846,
+      "step": 12022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9380789945294399,
+      "learning_rate": 1.3295749489058741e-06,
+      "loss": 0.6094,
+      "step": 12023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3251559911876405,
+      "learning_rate": 1.3288711932948427e-06,
+      "loss": 0.8354,
+      "step": 12024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8235425769615377,
+      "learning_rate": 1.3281675954403923e-06,
+      "loss": 0.9511,
+      "step": 12025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.01492939995334,
+      "learning_rate": 1.3274641553727568e-06,
+      "loss": 0.8851,
+      "step": 12026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.586485494531846,
+      "learning_rate": 1.3267608731221671e-06,
+      "loss": 0.6417,
+      "step": 12027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.236732605162962,
+      "learning_rate": 1.3260577487188431e-06,
+      "loss": 0.8959,
+      "step": 12028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5110351497886905,
+      "learning_rate": 1.3253547821930002e-06,
+      "loss": 0.7348,
+      "step": 12029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1090331917162513,
+      "learning_rate": 1.3246519735748465e-06,
+      "loss": 0.7634,
+      "step": 12030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6439175280532763,
+      "learning_rate": 1.3239493228945821e-06,
+      "loss": 0.7293,
+      "step": 12031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7133990195520474,
+      "learning_rate": 1.323246830182403e-06,
+      "loss": 0.7886,
+      "step": 12032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.073446895176272,
+      "learning_rate": 1.3225444954684962e-06,
+      "loss": 0.8779,
+      "step": 12033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.408753577654232,
+      "learning_rate": 1.3218423187830409e-06,
+      "loss": 0.8414,
+      "step": 12034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9070291335934901,
+      "learning_rate": 1.3211403001562155e-06,
+      "loss": 0.7588,
+      "step": 12035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4833310290044637,
+      "learning_rate": 1.3204384396181819e-06,
+      "loss": 0.6607,
+      "step": 12036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.080933266707274,
+      "learning_rate": 1.3197367371991048e-06,
+      "loss": 0.8663,
+      "step": 12037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.754627037206303,
+      "learning_rate": 1.3190351929291367e-06,
+      "loss": 0.8044,
+      "step": 12038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.896155185486855,
+      "learning_rate": 1.3183338068384243e-06,
+      "loss": 0.8826,
+      "step": 12039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6422797906249578,
+      "learning_rate": 1.3176325789571075e-06,
+      "loss": 0.8195,
+      "step": 12040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8002163071487662,
+      "learning_rate": 1.316931509315319e-06,
+      "loss": 0.7774,
+      "step": 12041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.853056966562872,
+      "learning_rate": 1.3162305979431883e-06,
+      "loss": 0.7697,
+      "step": 12042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.222043851518665,
+      "learning_rate": 1.315529844870832e-06,
+      "loss": 0.8198,
+      "step": 12043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6691901656438706,
+      "learning_rate": 1.3148292501283627e-06,
+      "loss": 0.7582,
+      "step": 12044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5860851556334428,
+      "learning_rate": 1.314128813745889e-06,
+      "loss": 0.842,
+      "step": 12045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4306089051084734,
+      "learning_rate": 1.3134285357535093e-06,
+      "loss": 0.6743,
+      "step": 12046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2730164272603772,
+      "learning_rate": 1.3127284161813153e-06,
+      "loss": 0.6488,
+      "step": 12047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2847964461624266,
+      "learning_rate": 1.3120284550593936e-06,
+      "loss": 0.8332,
+      "step": 12048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8735051620786194,
+      "learning_rate": 1.3113286524178232e-06,
+      "loss": 0.8445,
+      "step": 12049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3393711974384743,
+      "learning_rate": 1.310629008286675e-06,
+      "loss": 0.766,
+      "step": 12050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4650383173344306,
+      "learning_rate": 1.3099295226960151e-06,
+      "loss": 0.5811,
+      "step": 12051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4561002235038987,
+      "learning_rate": 1.3092301956759008e-06,
+      "loss": 0.6636,
+      "step": 12052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8738642243831443,
+      "learning_rate": 1.3085310272563872e-06,
+      "loss": 0.8194,
+      "step": 12053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1078102470162734,
+      "learning_rate": 1.3078320174675141e-06,
+      "loss": 0.9259,
+      "step": 12054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.518832626446708,
+      "learning_rate": 1.3071331663393233e-06,
+      "loss": 0.8109,
+      "step": 12055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8931006448564922,
+      "learning_rate": 1.3064344739018452e-06,
+      "loss": 0.8014,
+      "step": 12056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2595461942447144,
+      "learning_rate": 1.3057359401851034e-06,
+      "loss": 0.6322,
+      "step": 12057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.525425907007293,
+      "learning_rate": 1.305037565219116e-06,
+      "loss": 0.9202,
+      "step": 12058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.477841989065804,
+      "learning_rate": 1.3043393490338918e-06,
+      "loss": 0.8317,
+      "step": 12059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0572715631232747,
+      "learning_rate": 1.3036412916594398e-06,
+      "loss": 0.583,
+      "step": 12060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.043810197474128,
+      "learning_rate": 1.3029433931257524e-06,
+      "loss": 0.9199,
+      "step": 12061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8485957637465067,
+      "learning_rate": 1.3022456534628191e-06,
+      "loss": 0.7511,
+      "step": 12062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0347310557493787,
+      "learning_rate": 1.3015480727006274e-06,
+      "loss": 0.8623,
+      "step": 12063
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.464923793380239,
+      "learning_rate": 1.3008506508691516e-06,
+      "loss": 0.6639,
+      "step": 12064
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.114201980160169,
+      "learning_rate": 1.3001533879983618e-06,
+      "loss": 0.7012,
+      "step": 12065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.831341085092814,
+      "learning_rate": 1.2994562841182213e-06,
+      "loss": 0.631,
+      "step": 12066
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7652475907095286,
+      "learning_rate": 1.298759339258685e-06,
+      "loss": 0.691,
+      "step": 12067
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.692170895179148,
+      "learning_rate": 1.2980625534497037e-06,
+      "loss": 0.8831,
+      "step": 12068
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.091552664161291,
+      "learning_rate": 1.2973659267212173e-06,
+      "loss": 1.0674,
+      "step": 12069
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8784176041331617,
+      "learning_rate": 1.2966694591031653e-06,
+      "loss": 0.7009,
+      "step": 12070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0651896804564318,
+      "learning_rate": 1.2959731506254751e-06,
+      "loss": 0.7304,
+      "step": 12071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.170575262189018,
+      "learning_rate": 1.2952770013180655e-06,
+      "loss": 0.791,
+      "step": 12072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2599616037036956,
+      "learning_rate": 1.2945810112108554e-06,
+      "loss": 0.9515,
+      "step": 12073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.249955915267691,
+      "learning_rate": 1.2938851803337516e-06,
+      "loss": 0.8948,
+      "step": 12074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8226991796921483,
+      "learning_rate": 1.2931895087166551e-06,
+      "loss": 0.7834,
+      "step": 12075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9543377673053988,
+      "learning_rate": 1.2924939963894617e-06,
+      "loss": 0.835,
+      "step": 12076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.991642648877325,
+      "learning_rate": 1.2917986433820567e-06,
+      "loss": 0.8899,
+      "step": 12077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3191663891861567,
+      "learning_rate": 1.2911034497243251e-06,
+      "loss": 0.8036,
+      "step": 12078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0471545169808705,
+      "learning_rate": 1.290408415446136e-06,
+      "loss": 0.7284,
+      "step": 12079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.374169227107538,
+      "learning_rate": 1.2897135405773602e-06,
+      "loss": 0.9757,
+      "step": 12080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.068326544276005,
+      "learning_rate": 1.2890188251478569e-06,
+      "loss": 0.7685,
+      "step": 12081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8189854459171448,
+      "learning_rate": 1.2883242691874792e-06,
+      "loss": 0.7231,
+      "step": 12082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4782751364903453,
+      "learning_rate": 1.2876298727260744e-06,
+      "loss": 0.7706,
+      "step": 12083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0806683559785852,
+      "learning_rate": 1.2869356357934815e-06,
+      "loss": 0.6287,
+      "step": 12084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9186877972775274,
+      "learning_rate": 1.2862415584195338e-06,
+      "loss": 0.7732,
+      "step": 12085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8294550443622069,
+      "learning_rate": 1.2855476406340578e-06,
+      "loss": 0.8573,
+      "step": 12086
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1979026639106714,
+      "learning_rate": 1.28485388246687e-06,
+      "loss": 0.7107,
+      "step": 12087
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.083031148624815,
+      "learning_rate": 1.2841602839477868e-06,
+      "loss": 0.8171,
+      "step": 12088
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2026124511612921,
+      "learning_rate": 1.2834668451066118e-06,
+      "loss": 0.6352,
+      "step": 12089
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.112859644840615,
+      "learning_rate": 1.2827735659731438e-06,
+      "loss": 0.8221,
+      "step": 12090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9753904870276379,
+      "learning_rate": 1.2820804465771742e-06,
+      "loss": 0.5845,
+      "step": 12091
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2447460664518728,
+      "learning_rate": 1.2813874869484866e-06,
+      "loss": 0.6325,
+      "step": 12092
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2104941797720095,
+      "learning_rate": 1.2806946871168634e-06,
+      "loss": 0.7845,
+      "step": 12093
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.116931933621655,
+      "learning_rate": 1.2800020471120717e-06,
+      "loss": 0.9044,
+      "step": 12094
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.061350756344873,
+      "learning_rate": 1.279309566963875e-06,
+      "loss": 0.9301,
+      "step": 12095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0348967795780606,
+      "learning_rate": 1.2786172467020357e-06,
+      "loss": 0.6519,
+      "step": 12096
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8926476055062846,
+      "learning_rate": 1.2779250863562987e-06,
+      "loss": 0.7,
+      "step": 12097
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.264411842284723,
+      "learning_rate": 1.2772330859564115e-06,
+      "loss": 0.7504,
+      "step": 12098
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.24655651957693,
+      "learning_rate": 1.276541245532109e-06,
+      "loss": 0.727,
+      "step": 12099
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.376289082012982,
+      "learning_rate": 1.2758495651131226e-06,
+      "loss": 0.8269,
+      "step": 12100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4978142303325606,
+      "learning_rate": 1.275158044729174e-06,
+      "loss": 0.8087,
+      "step": 12101
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.984596186196655,
+      "learning_rate": 1.2744666844099801e-06,
+      "loss": 0.9194,
+      "step": 12102
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.309181301951535,
+      "learning_rate": 1.2737754841852501e-06,
+      "loss": 0.8579,
+      "step": 12103
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3030783742685155,
+      "learning_rate": 1.2730844440846862e-06,
+      "loss": 0.7367,
+      "step": 12104
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7632276487082075,
+      "learning_rate": 1.2723935641379826e-06,
+      "loss": 0.8572,
+      "step": 12105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3229784561306184,
+      "learning_rate": 1.2717028443748303e-06,
+      "loss": 0.6366,
+      "step": 12106
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.23922631228338,
+      "learning_rate": 1.2710122848249106e-06,
+      "loss": 0.7457,
+      "step": 12107
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.885069043648792,
+      "learning_rate": 1.2703218855178984e-06,
+      "loss": 0.6857,
+      "step": 12108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1801305936480366,
+      "learning_rate": 1.2696316464834607e-06,
+      "loss": 0.6823,
+      "step": 12109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.310197015576705,
+      "learning_rate": 1.2689415677512574e-06,
+      "loss": 0.8397,
+      "step": 12110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2037905725752287,
+      "learning_rate": 1.2682516493509478e-06,
+      "loss": 0.7959,
+      "step": 12111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9607696253243787,
+      "learning_rate": 1.267561891312174e-06,
+      "loss": 0.9673,
+      "step": 12112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.030415338012892,
+      "learning_rate": 1.2668722936645773e-06,
+      "loss": 0.7038,
+      "step": 12113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.570584263931921,
+      "learning_rate": 1.2661828564377948e-06,
+      "loss": 0.6539,
+      "step": 12114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.12458773454487,
+      "learning_rate": 1.2654935796614481e-06,
+      "loss": 0.6537,
+      "step": 12115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.958910175265349,
+      "learning_rate": 1.2648044633651607e-06,
+      "loss": 0.8174,
+      "step": 12116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.970879908595904,
+      "learning_rate": 1.2641155075785444e-06,
+      "loss": 0.8749,
+      "step": 12117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.12341785178391,
+      "learning_rate": 1.2634267123312049e-06,
+      "loss": 1.0056,
+      "step": 12118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9692995860631838,
+      "learning_rate": 1.2627380776527415e-06,
+      "loss": 0.6277,
+      "step": 12119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8493543109336033,
+      "learning_rate": 1.2620496035727448e-06,
+      "loss": 0.792,
+      "step": 12120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4961292279392087,
+      "learning_rate": 1.2613612901208022e-06,
+      "loss": 0.6929,
+      "step": 12121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2951268259444317,
+      "learning_rate": 1.2606731373264936e-06,
+      "loss": 0.8013,
+      "step": 12122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.114394272837131,
+      "learning_rate": 1.2599851452193851e-06,
+      "loss": 0.682,
+      "step": 12123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.690198570598748,
+      "learning_rate": 1.259297313829046e-06,
+      "loss": 0.8856,
+      "step": 12124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.286969375219936,
+      "learning_rate": 1.258609643185032e-06,
+      "loss": 0.587,
+      "step": 12125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.539135761042485,
+      "learning_rate": 1.2579221333168945e-06,
+      "loss": 0.6395,
+      "step": 12126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.621024153698915,
+      "learning_rate": 1.2572347842541766e-06,
+      "loss": 0.8158,
+      "step": 12127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7429439903381931,
+      "learning_rate": 1.2565475960264145e-06,
+      "loss": 0.7235,
+      "step": 12128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.009634935568603,
+      "learning_rate": 1.255860568663142e-06,
+      "loss": 0.6887,
+      "step": 12129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2761338482130102,
+      "learning_rate": 1.2551737021938765e-06,
+      "loss": 0.8612,
+      "step": 12130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7730469816888064,
+      "learning_rate": 1.2544869966481389e-06,
+      "loss": 0.7913,
+      "step": 12131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9936324182451377,
+      "learning_rate": 1.253800452055438e-06,
+      "loss": 0.7833,
+      "step": 12132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1454062166312213,
+      "learning_rate": 1.2531140684452724e-06,
+      "loss": 0.6319,
+      "step": 12133
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.499479353353938,
+      "learning_rate": 1.2524278458471411e-06,
+      "loss": 0.6248,
+      "step": 12134
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0831954757606783,
+      "learning_rate": 1.2517417842905316e-06,
+      "loss": 1.0478,
+      "step": 12135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.09791154176729,
+      "learning_rate": 1.2510558838049258e-06,
+      "loss": 0.8213,
+      "step": 12136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9824792894030518,
+      "learning_rate": 1.2503701444197974e-06,
+      "loss": 0.7296,
+      "step": 12137
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9121787693164598,
+      "learning_rate": 1.249684566164614e-06,
+      "loss": 0.8747,
+      "step": 12138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9749189520059522,
+      "learning_rate": 1.248999149068838e-06,
+      "loss": 0.6571,
+      "step": 12139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4854683719566517,
+      "learning_rate": 1.2483138931619227e-06,
+      "loss": 0.9931,
+      "step": 12140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1323822796602212,
+      "learning_rate": 1.2476287984733154e-06,
+      "loss": 0.6943,
+      "step": 12141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0442498107066798,
+      "learning_rate": 1.246943865032455e-06,
+      "loss": 0.6294,
+      "step": 12142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2733846718995783,
+      "learning_rate": 1.246259092868775e-06,
+      "loss": 0.6851,
+      "step": 12143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4836026522486088,
+      "learning_rate": 1.2455744820117028e-06,
+      "loss": 1.0897,
+      "step": 12144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9002712426190462,
+      "learning_rate": 1.2448900324906559e-06,
+      "loss": 0.752,
+      "step": 12145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0311356009070536,
+      "learning_rate": 1.2442057443350458e-06,
+      "loss": 0.6281,
+      "step": 12146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2640071715509893,
+      "learning_rate": 1.2435216175742826e-06,
+      "loss": 0.8584,
+      "step": 12147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6932180432683672,
+      "learning_rate": 1.242837652237759e-06,
+      "loss": 0.7847,
+      "step": 12148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2180725099678678,
+      "learning_rate": 1.2421538483548706e-06,
+      "loss": 0.8389,
+      "step": 12149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.0156618524168355,
+      "learning_rate": 1.2414702059550016e-06,
+      "loss": 0.6468,
+      "step": 12150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.714065836928448,
+      "learning_rate": 1.2407867250675253e-06,
+      "loss": 0.872,
+      "step": 12151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.039677008559975,
+      "learning_rate": 1.2401034057218181e-06,
+      "loss": 0.5259,
+      "step": 12152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2849952059580096,
+      "learning_rate": 1.2394202479472406e-06,
+      "loss": 1.0573,
+      "step": 12153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0209988996551516,
+      "learning_rate": 1.2387372517731505e-06,
+      "loss": 0.6192,
+      "step": 12154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.984625791246723,
+      "learning_rate": 1.238054417228897e-06,
+      "loss": 0.6749,
+      "step": 12155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0183825977532783,
+      "learning_rate": 1.2373717443438227e-06,
+      "loss": 0.8338,
+      "step": 12156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8754650517489144,
+      "learning_rate": 1.2366892331472657e-06,
+      "loss": 0.7854,
+      "step": 12157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.33237632420631,
+      "learning_rate": 1.2360068836685536e-06,
+      "loss": 0.6149,
+      "step": 12158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1228756522166328,
+      "learning_rate": 1.2353246959370086e-06,
+      "loss": 0.7416,
+      "step": 12159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9251411709872002,
+      "learning_rate": 1.234642669981946e-06,
+      "loss": 0.8277,
+      "step": 12160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9147029990516031,
+      "learning_rate": 1.2339608058326736e-06,
+      "loss": 0.5665,
+      "step": 12161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9789103707618332,
+      "learning_rate": 1.2332791035184926e-06,
+      "loss": 0.6678,
+      "step": 12162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1702845328839984,
+      "learning_rate": 1.2325975630686976e-06,
+      "loss": 0.7551,
+      "step": 12163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.102793743320949,
+      "learning_rate": 1.2319161845125744e-06,
+      "loss": 0.7695,
+      "step": 12164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.412137235539118,
+      "learning_rate": 1.2312349678794072e-06,
+      "loss": 0.8846,
+      "step": 12165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.359057417420907,
+      "learning_rate": 1.2305539131984646e-06,
+      "loss": 0.6761,
+      "step": 12166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8817979431263208,
+      "learning_rate": 1.2298730204990162e-06,
+      "loss": 0.8018,
+      "step": 12167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.128519451645575,
+      "learning_rate": 1.229192289810321e-06,
+      "loss": 0.7217,
+      "step": 12168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1906455806158185,
+      "learning_rate": 1.228511721161631e-06,
+      "loss": 0.8922,
+      "step": 12169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9788095228957552,
+      "learning_rate": 1.2278313145821924e-06,
+      "loss": 0.6678,
+      "step": 12170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3611489553398446,
+      "learning_rate": 1.2271510701012412e-06,
+      "loss": 0.7963,
+      "step": 12171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.818438856004634,
+      "learning_rate": 1.2264709877480147e-06,
+      "loss": 0.8278,
+      "step": 12172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.138049766218091,
+      "learning_rate": 1.2257910675517315e-06,
+      "loss": 0.8299,
+      "step": 12173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0400192287536947,
+      "learning_rate": 1.2251113095416113e-06,
+      "loss": 0.7466,
+      "step": 12174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.764162533866647,
+      "learning_rate": 1.2244317137468658e-06,
+      "loss": 0.9692,
+      "step": 12175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0628757458997073,
+      "learning_rate": 1.223752280196699e-06,
+      "loss": 0.9384,
+      "step": 12176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9582527984686546,
+      "learning_rate": 1.2230730089203063e-06,
+      "loss": 0.6684,
+      "step": 12177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8922067965375498,
+      "learning_rate": 1.2223938999468788e-06,
+      "loss": 0.729,
+      "step": 12178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9570211453447297,
+      "learning_rate": 1.2217149533055976e-06,
+      "loss": 0.8452,
+      "step": 12179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3070781341811457,
+      "learning_rate": 1.22103616902564e-06,
+      "loss": 0.6488,
+      "step": 12180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.884232973490556,
+      "learning_rate": 1.2203575471361733e-06,
+      "loss": 1.0261,
+      "step": 12181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.138584630173745,
+      "learning_rate": 1.2196790876663617e-06,
+      "loss": 0.881,
+      "step": 12182
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9908059343107355,
+      "learning_rate": 1.2190007906453605e-06,
+      "loss": 0.6541,
+      "step": 12183
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.675293998807207,
+      "learning_rate": 1.2183226561023132e-06,
+      "loss": 0.7035,
+      "step": 12184
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0194538056055897,
+      "learning_rate": 1.2176446840663647e-06,
+      "loss": 0.6817,
+      "step": 12185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.211699012524824,
+      "learning_rate": 1.216966874566648e-06,
+      "loss": 0.8564,
+      "step": 12186
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2723622305748632,
+      "learning_rate": 1.21628922763229e-06,
+      "loss": 0.7089,
+      "step": 12187
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0242154701931776,
+      "learning_rate": 1.2156117432924108e-06,
+      "loss": 0.8012,
+      "step": 12188
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0815293522735896,
+      "learning_rate": 1.2149344215761216e-06,
+      "loss": 0.8073,
+      "step": 12189
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8787445314865536,
+      "learning_rate": 1.2142572625125332e-06,
+      "loss": 0.9439,
+      "step": 12190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.18850408537039,
+      "learning_rate": 1.2135802661307388e-06,
+      "loss": 0.8234,
+      "step": 12191
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.227183998301444,
+      "learning_rate": 1.2129034324598343e-06,
+      "loss": 0.6581,
+      "step": 12192
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4700586633423667,
+      "learning_rate": 1.212226761528904e-06,
+      "loss": 0.7162,
+      "step": 12193
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2442599675835875,
+      "learning_rate": 1.2115502533670253e-06,
+      "loss": 0.6788,
+      "step": 12194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9385970873572278,
+      "learning_rate": 1.2108739080032695e-06,
+      "loss": 0.7945,
+      "step": 12195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.347821912458622,
+      "learning_rate": 1.2101977254667007e-06,
+      "loss": 0.8351,
+      "step": 12196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6618256958555147,
+      "learning_rate": 1.2095217057863762e-06,
+      "loss": 0.7946,
+      "step": 12197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.626449735919008,
+      "learning_rate": 1.2088458489913458e-06,
+      "loss": 0.7994,
+      "step": 12198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1175353484251525,
+      "learning_rate": 1.2081701551106506e-06,
+      "loss": 0.7626,
+      "step": 12199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8971038726440534,
+      "learning_rate": 1.207494624173331e-06,
+      "loss": 0.8131,
+      "step": 12200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9240964281179742,
+      "learning_rate": 1.2068192562084146e-06,
+      "loss": 0.9151,
+      "step": 12201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.377151094641768,
+      "learning_rate": 1.2061440512449196e-06,
+      "loss": 0.8232,
+      "step": 12202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9914490451530469,
+      "learning_rate": 1.205469009311866e-06,
+      "loss": 0.7048,
+      "step": 12203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1125204837921867,
+      "learning_rate": 1.20479413043826e-06,
+      "loss": 0.6437,
+      "step": 12204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3869154557716246,
+      "learning_rate": 1.2041194146531026e-06,
+      "loss": 0.8438,
+      "step": 12205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4311032596013087,
+      "learning_rate": 1.2034448619853873e-06,
+      "loss": 0.8301,
+      "step": 12206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8592215795953146,
+      "learning_rate": 1.202770472464101e-06,
+      "loss": 0.8661,
+      "step": 12207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.850242799282281,
+      "learning_rate": 1.2020962461182268e-06,
+      "loss": 0.7636,
+      "step": 12208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0403525213632896,
+      "learning_rate": 1.201422182976732e-06,
+      "loss": 0.6282,
+      "step": 12209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9444034610523633,
+      "learning_rate": 1.200748283068588e-06,
+      "loss": 0.8235,
+      "step": 12210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.155345155755423,
+      "learning_rate": 1.2000745464227515e-06,
+      "loss": 0.8465,
+      "step": 12211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0743401198167362,
+      "learning_rate": 1.1994009730681744e-06,
+      "loss": 0.5941,
+      "step": 12212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1038511457378355,
+      "learning_rate": 1.198727563033802e-06,
+      "loss": 0.7539,
+      "step": 12213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.140375967753728,
+      "learning_rate": 1.1980543163485726e-06,
+      "loss": 0.83,
+      "step": 12214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.149940906023738,
+      "learning_rate": 1.1973812330414159e-06,
+      "loss": 0.781,
+      "step": 12215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8068762621637324,
+      "learning_rate": 1.1967083131412566e-06,
+      "loss": 0.7158,
+      "step": 12216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1874839981592604,
+      "learning_rate": 1.1960355566770104e-06,
+      "loss": 0.7598,
+      "step": 12217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1432676646915365,
+      "learning_rate": 1.1953629636775893e-06,
+      "loss": 0.5976,
+      "step": 12218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.568289141813901,
+      "learning_rate": 1.1946905341718951e-06,
+      "loss": 0.7682,
+      "step": 12219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6258149532805826,
+      "learning_rate": 1.1940182681888235e-06,
+      "loss": 1.087,
+      "step": 12220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3221709139886695,
+      "learning_rate": 1.1933461657572637e-06,
+      "loss": 0.8265,
+      "step": 12221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6558799298440863,
+      "learning_rate": 1.1926742269060965e-06,
+      "loss": 0.8464,
+      "step": 12222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.21457487997546,
+      "learning_rate": 1.1920024516641976e-06,
+      "loss": 0.8196,
+      "step": 12223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3131998776756286,
+      "learning_rate": 1.1913308400604339e-06,
+      "loss": 0.7384,
+      "step": 12224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2568421139405768,
+      "learning_rate": 1.1906593921236649e-06,
+      "loss": 0.8051,
+      "step": 12225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.128722203371468,
+      "learning_rate": 1.1899881078827485e-06,
+      "loss": 0.7848,
+      "step": 12226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1163569461501652,
+      "learning_rate": 1.189316987366526e-06,
+      "loss": 0.635,
+      "step": 12227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0280888164502544,
+      "learning_rate": 1.1886460306038405e-06,
+      "loss": 0.7111,
+      "step": 12228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.462157187428903,
+      "learning_rate": 1.1879752376235231e-06,
+      "loss": 0.8575,
+      "step": 12229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.309916961242706,
+      "learning_rate": 1.1873046084544004e-06,
+      "loss": 0.9067,
+      "step": 12230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.860921400191117,
+      "learning_rate": 1.1866341431252892e-06,
+      "loss": 0.8206,
+      "step": 12231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.605319663210265,
+      "learning_rate": 1.185963841665001e-06,
+      "loss": 0.7755,
+      "step": 12232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1868359207739645,
+      "learning_rate": 1.1852937041023433e-06,
+      "loss": 0.6553,
+      "step": 12233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.024800992684041,
+      "learning_rate": 1.1846237304661095e-06,
+      "loss": 0.7682,
+      "step": 12234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.946066947574505,
+      "learning_rate": 1.1839539207850898e-06,
+      "loss": 0.7075,
+      "step": 12235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.728670810534061,
+      "learning_rate": 1.1832842750880702e-06,
+      "loss": 0.6908,
+      "step": 12236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0823412131024341,
+      "learning_rate": 1.1826147934038262e-06,
+      "loss": 0.6438,
+      "step": 12237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.679942269494007,
+      "learning_rate": 1.181945475761126e-06,
+      "loss": 0.7602,
+      "step": 12238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.140211564465153,
+      "learning_rate": 1.181276322188732e-06,
+      "loss": 0.6886,
+      "step": 12239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7701887852922267,
+      "learning_rate": 1.1806073327153988e-06,
+      "loss": 0.7877,
+      "step": 12240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1663938001343697,
+      "learning_rate": 1.1799385073698749e-06,
+      "loss": 0.8162,
+      "step": 12241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9424234255396216,
+      "learning_rate": 1.1792698461809e-06,
+      "loss": 0.6255,
+      "step": 12242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2219088746115236,
+      "learning_rate": 1.1786013491772103e-06,
+      "loss": 0.7169,
+      "step": 12243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0606649538635478,
+      "learning_rate": 1.1779330163875325e-06,
+      "loss": 0.5729,
+      "step": 12244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1561092488112301,
+      "learning_rate": 1.1772648478405824e-06,
+      "loss": 0.6389,
+      "step": 12245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8858238746844327,
+      "learning_rate": 1.176596843565077e-06,
+      "loss": 0.9891,
+      "step": 12246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3070441835103175,
+      "learning_rate": 1.1759290035897208e-06,
+      "loss": 0.83,
+      "step": 12247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.551307726897898,
+      "learning_rate": 1.1752613279432112e-06,
+      "loss": 0.4644,
+      "step": 12248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.433634247312405,
+      "learning_rate": 1.1745938166542414e-06,
+      "loss": 0.7629,
+      "step": 12249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.563041985006181,
+      "learning_rate": 1.173926469751493e-06,
+      "loss": 0.8336,
+      "step": 12250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4977091675251186,
+      "learning_rate": 1.1732592872636478e-06,
+      "loss": 0.8041,
+      "step": 12251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7886721946122661,
+      "learning_rate": 1.172592269219373e-06,
+      "loss": 0.6418,
+      "step": 12252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1627248088790902,
+      "learning_rate": 1.1719254156473302e-06,
+      "loss": 0.5595,
+      "step": 12253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8144341421769865,
+      "learning_rate": 1.1712587265761799e-06,
+      "loss": 0.8806,
+      "step": 12254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2084402996820347,
+      "learning_rate": 1.1705922020345688e-06,
+      "loss": 0.8062,
+      "step": 12255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1565804936935455,
+      "learning_rate": 1.16992584205114e-06,
+      "loss": 0.7987,
+      "step": 12256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.153493095430804,
+      "learning_rate": 1.1692596466545275e-06,
+      "loss": 0.6276,
+      "step": 12257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4749014467342856,
+      "learning_rate": 1.168593615873358e-06,
+      "loss": 0.8718,
+      "step": 12258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5650203910133818,
+      "learning_rate": 1.1679277497362563e-06,
+      "loss": 0.8363,
+      "step": 12259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.760697156786566,
+      "learning_rate": 1.167262048271831e-06,
+      "loss": 0.9279,
+      "step": 12260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6082240326065422,
+      "learning_rate": 1.166596511508693e-06,
+      "loss": 0.8718,
+      "step": 12261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.406223727655752,
+      "learning_rate": 1.1659311394754418e-06,
+      "loss": 0.9115,
+      "step": 12262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.186724899266382,
+      "learning_rate": 1.1652659322006659e-06,
+      "loss": 0.9961,
+      "step": 12263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.08863794378199,
+      "learning_rate": 1.1646008897129546e-06,
+      "loss": 0.6929,
+      "step": 12264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9479088771710988,
+      "learning_rate": 1.1639360120408843e-06,
+      "loss": 0.6476,
+      "step": 12265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.336187728974778,
+      "learning_rate": 1.1632712992130278e-06,
+      "loss": 0.8033,
+      "step": 12266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.176900942920089,
+      "learning_rate": 1.1626067512579481e-06,
+      "loss": 0.8741,
+      "step": 12267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.140575283303607,
+      "learning_rate": 1.1619423682042008e-06,
+      "loss": 0.6856,
+      "step": 12268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3186029028732227,
+      "learning_rate": 1.161278150080341e-06,
+      "loss": 0.8029,
+      "step": 12269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4295948490905412,
+      "learning_rate": 1.160614096914906e-06,
+      "loss": 0.7729,
+      "step": 12270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1228542227981577,
+      "learning_rate": 1.1599502087364345e-06,
+      "loss": 0.7339,
+      "step": 12271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1268270867732957,
+      "learning_rate": 1.1592864855734553e-06,
+      "loss": 0.9561,
+      "step": 12272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.087378985437019,
+      "learning_rate": 1.1586229274544898e-06,
+      "loss": 0.7022,
+      "step": 12273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.083319512058147,
+      "learning_rate": 1.157959534408052e-06,
+      "loss": 0.8905,
+      "step": 12274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.479003493934112,
+      "learning_rate": 1.1572963064626497e-06,
+      "loss": 0.8818,
+      "step": 12275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.361169959693564,
+      "learning_rate": 1.1566332436467815e-06,
+      "loss": 0.714,
+      "step": 12276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.241152325531059,
+      "learning_rate": 1.155970345988946e-06,
+      "loss": 0.793,
+      "step": 12277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0552720980547208,
+      "learning_rate": 1.1553076135176222e-06,
+      "loss": 0.6742,
+      "step": 12278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.07028136496277,
+      "learning_rate": 1.1546450462612951e-06,
+      "loss": 0.8782,
+      "step": 12279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.804700109166351,
+      "learning_rate": 1.1539826442484337e-06,
+      "loss": 0.6485,
+      "step": 12280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.252105670542925,
+      "learning_rate": 1.1533204075075039e-06,
+      "loss": 0.9672,
+      "step": 12281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9933042406166497,
+      "learning_rate": 1.1526583360669635e-06,
+      "loss": 0.7545,
+      "step": 12282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.022084637967622,
+      "learning_rate": 1.1519964299552611e-06,
+      "loss": 1.0005,
+      "step": 12283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8119301688317961,
+      "learning_rate": 1.151334689200845e-06,
+      "loss": 0.6663,
+      "step": 12284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7581930778385035,
+      "learning_rate": 1.1506731138321474e-06,
+      "loss": 0.5703,
+      "step": 12285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1336738014620873,
+      "learning_rate": 1.1500117038775977e-06,
+      "loss": 0.8034,
+      "step": 12286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5278976295900155,
+      "learning_rate": 1.149350459365622e-06,
+      "loss": 0.7422,
+      "step": 12287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4872452611436735,
+      "learning_rate": 1.1486893803246302e-06,
+      "loss": 0.7665,
+      "step": 12288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.98419525506122,
+      "learning_rate": 1.1480284667830343e-06,
+      "loss": 0.7382,
+      "step": 12289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.105753885197203,
+      "learning_rate": 1.1473677187692345e-06,
+      "loss": 0.6976,
+      "step": 12290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2632396105504333,
+      "learning_rate": 1.1467071363116234e-06,
+      "loss": 0.8309,
+      "step": 12291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.023880062471959,
+      "learning_rate": 1.1460467194385889e-06,
+      "loss": 0.6277,
+      "step": 12292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0267650797075296,
+      "learning_rate": 1.1453864681785087e-06,
+      "loss": 0.9084,
+      "step": 12293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.900430572735289,
+      "learning_rate": 1.1447263825597577e-06,
+      "loss": 0.829,
+      "step": 12294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.521193328539079,
+      "learning_rate": 1.1440664626107017e-06,
+      "loss": 0.6981,
+      "step": 12295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.307872723334043,
+      "learning_rate": 1.143406708359695e-06,
+      "loss": 0.7036,
+      "step": 12296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3256212828788048,
+      "learning_rate": 1.142747119835092e-06,
+      "loss": 0.8493,
+      "step": 12297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4295664493404634,
+      "learning_rate": 1.1420876970652361e-06,
+      "loss": 0.9208,
+      "step": 12298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0696484465710296,
+      "learning_rate": 1.1414284400784643e-06,
+      "loss": 0.5935,
+      "step": 12299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.503596312482146,
+      "learning_rate": 1.1407693489031052e-06,
+      "loss": 0.6455,
+      "step": 12300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8478586171859477,
+      "learning_rate": 1.1401104235674808e-06,
+      "loss": 0.9042,
+      "step": 12301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.686756521745646,
+      "learning_rate": 1.1394516640999114e-06,
+      "loss": 0.7763,
+      "step": 12302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.790399759596297,
+      "learning_rate": 1.1387930705286998e-06,
+      "loss": 0.835,
+      "step": 12303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.917151345144757,
+      "learning_rate": 1.1381346428821482e-06,
+      "loss": 0.8093,
+      "step": 12304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8797193313097857,
+      "learning_rate": 1.1374763811885547e-06,
+      "loss": 0.6934,
+      "step": 12305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9918389886266832,
+      "learning_rate": 1.1368182854762005e-06,
+      "loss": 0.9302,
+      "step": 12306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2001626075344447,
+      "learning_rate": 1.1361603557733698e-06,
+      "loss": 0.7314,
+      "step": 12307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1933674190723442,
+      "learning_rate": 1.135502592108334e-06,
+      "loss": 0.8265,
+      "step": 12308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.097249727148508,
+      "learning_rate": 1.134844994509358e-06,
+      "loss": 0.8463,
+      "step": 12309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9072120028994815,
+      "learning_rate": 1.1341875630047012e-06,
+      "loss": 0.8783,
+      "step": 12310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.243047117844331,
+      "learning_rate": 1.1335302976226132e-06,
+      "loss": 0.8166,
+      "step": 12311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.009747752665315,
+      "learning_rate": 1.1328731983913404e-06,
+      "loss": 0.8986,
+      "step": 12312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5456238634072275,
+      "learning_rate": 1.13221626533912e-06,
+      "loss": 0.9416,
+      "step": 12313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.208898480613451,
+      "learning_rate": 1.1315594984941786e-06,
+      "loss": 0.7663,
+      "step": 12314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0986799518724333,
+      "learning_rate": 1.1309028978847419e-06,
+      "loss": 0.6872,
+      "step": 12315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.617260289414935,
+      "learning_rate": 1.1302464635390247e-06,
+      "loss": 0.8997,
+      "step": 12316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1801430283368606,
+      "learning_rate": 1.1295901954852356e-06,
+      "loss": 0.6986,
+      "step": 12317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2808926246943,
+      "learning_rate": 1.1289340937515753e-06,
+      "loss": 0.8078,
+      "step": 12318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.792051826686675,
+      "learning_rate": 1.1282781583662372e-06,
+      "loss": 0.9562,
+      "step": 12319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.325116155733676,
+      "learning_rate": 1.1276223893574123e-06,
+      "loss": 0.8006,
+      "step": 12320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.104747207704158,
+      "learning_rate": 1.1269667867532746e-06,
+      "loss": 0.9796,
+      "step": 12321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.047754719814011,
+      "learning_rate": 1.1263113505820012e-06,
+      "loss": 0.6274,
+      "step": 12322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6944525642241026,
+      "learning_rate": 1.1256560808717576e-06,
+      "loss": 0.9078,
+      "step": 12323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6927295610577615,
+      "learning_rate": 1.1250009776506982e-06,
+      "loss": 0.8093,
+      "step": 12324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2671383633222413,
+      "learning_rate": 1.124346040946978e-06,
+      "loss": 0.6367,
+      "step": 12325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.115268608209578,
+      "learning_rate": 1.1236912707887404e-06,
+      "loss": 0.8214,
+      "step": 12326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6699912295830104,
+      "learning_rate": 1.1230366672041216e-06,
+      "loss": 0.5394,
+      "step": 12327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.433628988445356,
+      "learning_rate": 1.1223822302212517e-06,
+      "loss": 0.9318,
+      "step": 12328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5545957721569095,
+      "learning_rate": 1.1217279598682518e-06,
+      "loss": 0.7332,
+      "step": 12329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2950388996581945,
+      "learning_rate": 1.1210738561732405e-06,
+      "loss": 0.8855,
+      "step": 12330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9062427058344844,
+      "learning_rate": 1.120419919164324e-06,
+      "loss": 0.81,
+      "step": 12331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8703328024862345,
+      "learning_rate": 1.1197661488696043e-06,
+      "loss": 0.841,
+      "step": 12332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1202857122391556,
+      "learning_rate": 1.1191125453171748e-06,
+      "loss": 0.7656,
+      "step": 12333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8059434833903434,
+      "learning_rate": 1.118459108535122e-06,
+      "loss": 0.831,
+      "step": 12334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2312315940891243,
+      "learning_rate": 1.1178058385515268e-06,
+      "loss": 0.9142,
+      "step": 12335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.151190707052542,
+      "learning_rate": 1.1171527353944601e-06,
+      "loss": 0.6642,
+      "step": 12336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.549796013469801,
+      "learning_rate": 1.1164997990919868e-06,
+      "loss": 0.6791,
+      "step": 12337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5870145007743077,
+      "learning_rate": 1.1158470296721691e-06,
+      "loss": 0.8623,
+      "step": 12338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3463860495302074,
+      "learning_rate": 1.1151944271630517e-06,
+      "loss": 0.7054,
+      "step": 12339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3121560909757513,
+      "learning_rate": 1.1145419915926836e-06,
+      "loss": 0.7227,
+      "step": 12340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0240088102698515,
+      "learning_rate": 1.1138897229890995e-06,
+      "loss": 0.8406,
+      "step": 12341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9612479999886103,
+      "learning_rate": 1.1132376213803286e-06,
+      "loss": 0.8193,
+      "step": 12342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4567638879501295,
+      "learning_rate": 1.1125856867943936e-06,
+      "loss": 0.7242,
+      "step": 12343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0770556332324497,
+      "learning_rate": 1.1119339192593077e-06,
+      "loss": 0.6879,
+      "step": 12344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8796483772626833,
+      "learning_rate": 1.111282318803083e-06,
+      "loss": 0.9999,
+      "step": 12345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0655781615481599,
+      "learning_rate": 1.1106308854537167e-06,
+      "loss": 0.6439,
+      "step": 12346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9738598207664577,
+      "learning_rate": 1.1099796192392021e-06,
+      "loss": 0.7321,
+      "step": 12347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5546271073159166,
+      "learning_rate": 1.109328520187528e-06,
+      "loss": 0.7115,
+      "step": 12348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.903665714220147,
+      "learning_rate": 1.1086775883266725e-06,
+      "loss": 0.7653,
+      "step": 12349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9884552800504378,
+      "learning_rate": 1.1080268236846076e-06,
+      "loss": 0.6285,
+      "step": 12350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.891336956323212,
+      "learning_rate": 1.1073762262892979e-06,
+      "loss": 0.6676,
+      "step": 12351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.258716887104413,
+      "learning_rate": 1.1067257961687005e-06,
+      "loss": 1.0678,
+      "step": 12352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.182967909162535,
+      "learning_rate": 1.1060755333507673e-06,
+      "loss": 0.8799,
+      "step": 12353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2918067913025566,
+      "learning_rate": 1.1054254378634399e-06,
+      "loss": 0.8888,
+      "step": 12354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0450868165586535,
+      "learning_rate": 1.1047755097346541e-06,
+      "loss": 0.7133,
+      "step": 12355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1026363424862566,
+      "learning_rate": 1.1041257489923429e-06,
+      "loss": 0.8746,
+      "step": 12356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9458746930644732,
+      "learning_rate": 1.1034761556644214e-06,
+      "loss": 1.028,
+      "step": 12357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0525161604393585,
+      "learning_rate": 1.1028267297788092e-06,
+      "loss": 0.7481,
+      "step": 12358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.929281540798562,
+      "learning_rate": 1.102177471363412e-06,
+      "loss": 0.7247,
+      "step": 12359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3969931835629987,
+      "learning_rate": 1.1015283804461297e-06,
+      "loss": 0.6922,
+      "step": 12360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3359197560201035,
+      "learning_rate": 1.1008794570548554e-06,
+      "loss": 0.7884,
+      "step": 12361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1318689381828455,
+      "learning_rate": 1.100230701217473e-06,
+      "loss": 1.0074,
+      "step": 12362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2205308052372932,
+      "learning_rate": 1.0995821129618656e-06,
+      "loss": 0.6799,
+      "step": 12363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1525747052608586,
+      "learning_rate": 1.0989336923158999e-06,
+      "loss": 0.7914,
+      "step": 12364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8974985306064232,
+      "learning_rate": 1.0982854393074399e-06,
+      "loss": 0.8553,
+      "step": 12365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1459735926677195,
+      "learning_rate": 1.0976373539643465e-06,
+      "loss": 0.6998,
+      "step": 12366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1225035610813214,
+      "learning_rate": 1.0969894363144661e-06,
+      "loss": 0.5244,
+      "step": 12367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.562272467329814,
+      "learning_rate": 1.0963416863856425e-06,
+      "loss": 0.9712,
+      "step": 12368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1079036515805787,
+      "learning_rate": 1.0956941042057106e-06,
+      "loss": 0.738,
+      "step": 12369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1189361227641044,
+      "learning_rate": 1.095046689802498e-06,
+      "loss": 0.7696,
+      "step": 12370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0078043742320884,
+      "learning_rate": 1.0943994432038262e-06,
+      "loss": 0.6263,
+      "step": 12371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5399916539119842,
+      "learning_rate": 1.0937523644375076e-06,
+      "loss": 0.8833,
+      "step": 12372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.991994964099076,
+      "learning_rate": 1.0931054535313507e-06,
+      "loss": 0.8169,
+      "step": 12373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9464397828973774,
+      "learning_rate": 1.0924587105131546e-06,
+      "loss": 0.7065,
+      "step": 12374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.22600056181035,
+      "learning_rate": 1.0918121354107076e-06,
+      "loss": 0.7379,
+      "step": 12375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9773928118053445,
+      "learning_rate": 1.091165728251799e-06,
+      "loss": 0.7089,
+      "step": 12376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3244554121227496,
+      "learning_rate": 1.090519489064204e-06,
+      "loss": 0.8421,
+      "step": 12377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2184161424271764,
+      "learning_rate": 1.0898734178756937e-06,
+      "loss": 0.733,
+      "step": 12378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.259069926328294,
+      "learning_rate": 1.0892275147140307e-06,
+      "loss": 0.6418,
+      "step": 12379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7091514106877135,
+      "learning_rate": 1.0885817796069697e-06,
+      "loss": 0.7523,
+      "step": 12380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.619091507096699,
+      "learning_rate": 1.087936212582264e-06,
+      "loss": 0.8354,
+      "step": 12381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0792176009984624,
+      "learning_rate": 1.0872908136676491e-06,
+      "loss": 0.6686,
+      "step": 12382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1365321373279853,
+      "learning_rate": 1.0866455828908634e-06,
+      "loss": 0.8697,
+      "step": 12383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.059235819048606,
+      "learning_rate": 1.086000520279632e-06,
+      "loss": 0.7918,
+      "step": 12384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1443178219324848,
+      "learning_rate": 1.0853556258616755e-06,
+      "loss": 0.7394,
+      "step": 12385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4406094534531206,
+      "learning_rate": 1.0847108996647066e-06,
+      "loss": 0.6951,
+      "step": 12386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.803388193918579,
+      "learning_rate": 1.0840663417164293e-06,
+      "loss": 0.535,
+      "step": 12387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1978786353095816,
+      "learning_rate": 1.0834219520445432e-06,
+      "loss": 0.8495,
+      "step": 12388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1105250860861124,
+      "learning_rate": 1.0827777306767384e-06,
+      "loss": 0.691,
+      "step": 12389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.272223706874251,
+      "learning_rate": 1.082133677640697e-06,
+      "loss": 0.7863,
+      "step": 12390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.164722451548012,
+      "learning_rate": 1.0814897929640984e-06,
+      "loss": 0.6384,
+      "step": 12391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0678006495612116,
+      "learning_rate": 1.0808460766746104e-06,
+      "loss": 0.6537,
+      "step": 12392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2072076447742948,
+      "learning_rate": 1.0802025287998953e-06,
+      "loss": 0.7809,
+      "step": 12393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.354343186377664,
+      "learning_rate": 1.0795591493676072e-06,
+      "loss": 0.8142,
+      "step": 12394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9598726416896892,
+      "learning_rate": 1.078915938405392e-06,
+      "loss": 0.6032,
+      "step": 12395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8440617996951638,
+      "learning_rate": 1.078272895940895e-06,
+      "loss": 0.7438,
+      "step": 12396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7287341954810278,
+      "learning_rate": 1.0776300220017437e-06,
+      "loss": 0.5989,
+      "step": 12397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5359516696131905,
+      "learning_rate": 1.0769873166155653e-06,
+      "loss": 0.8386,
+      "step": 12398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9235416965801138,
+      "learning_rate": 1.0763447798099813e-06,
+      "loss": 0.6411,
+      "step": 12399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8493772202126446,
+      "learning_rate": 1.075702411612598e-06,
+      "loss": 0.6918,
+      "step": 12400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1671979620776938,
+      "learning_rate": 1.0750602120510233e-06,
+      "loss": 0.6517,
+      "step": 12401
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.287965752548131,
+      "learning_rate": 1.0744181811528526e-06,
+      "loss": 0.7726,
+      "step": 12402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1610217473883435,
+      "learning_rate": 1.073776318945675e-06,
+      "loss": 0.7949,
+      "step": 12403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0484175552889545,
+      "learning_rate": 1.0731346254570735e-06,
+      "loss": 0.8245,
+      "step": 12404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9580609671605582,
+      "learning_rate": 1.0724931007146227e-06,
+      "loss": 0.6847,
+      "step": 12405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3760044779042446,
+      "learning_rate": 1.0718517447458904e-06,
+      "loss": 0.7348,
+      "step": 12406
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0207532100300134,
+      "learning_rate": 1.071210557578437e-06,
+      "loss": 0.7582,
+      "step": 12407
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3140189495016035,
+      "learning_rate": 1.0705695392398146e-06,
+      "loss": 0.6481,
+      "step": 12408
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0440587087887994,
+      "learning_rate": 1.0699286897575718e-06,
+      "loss": 0.7727,
+      "step": 12409
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.478951171226021,
+      "learning_rate": 1.0692880091592457e-06,
+      "loss": 0.8459,
+      "step": 12410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.646605442760219,
+      "learning_rate": 1.068647497472368e-06,
+      "loss": 0.6229,
+      "step": 12411
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3750431788588857,
+      "learning_rate": 1.0680071547244636e-06,
+      "loss": 0.8309,
+      "step": 12412
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3578931756217267,
+      "learning_rate": 1.0673669809430465e-06,
+      "loss": 0.8918,
+      "step": 12413
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.135878224590167,
+      "learning_rate": 1.066726976155632e-06,
+      "loss": 0.8377,
+      "step": 12414
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0346332631512007,
+      "learning_rate": 1.0660871403897177e-06,
+      "loss": 0.8869,
+      "step": 12415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4660337754402124,
+      "learning_rate": 1.0654474736727988e-06,
+      "loss": 0.6898,
+      "step": 12416
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9658765773965325,
+      "learning_rate": 1.0648079760323675e-06,
+      "loss": 0.6584,
+      "step": 12417
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4064297581871172,
+      "learning_rate": 1.064168647495899e-06,
+      "loss": 0.7617,
+      "step": 12418
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3578367035244368,
+      "learning_rate": 1.0635294880908702e-06,
+      "loss": 0.735,
+      "step": 12419
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3500463364834223,
+      "learning_rate": 1.0628904978447463e-06,
+      "loss": 0.9113,
+      "step": 12420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0206866180682903,
+      "learning_rate": 1.0622516767849867e-06,
+      "loss": 0.8521,
+      "step": 12421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1230265004469724,
+      "learning_rate": 1.0616130249390423e-06,
+      "loss": 0.6692,
+      "step": 12422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0915331278499103,
+      "learning_rate": 1.0609745423343553e-06,
+      "loss": 0.7824,
+      "step": 12423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1752890569382655,
+      "learning_rate": 1.0603362289983687e-06,
+      "loss": 0.9268,
+      "step": 12424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3206806573274985,
+      "learning_rate": 1.0596980849585065e-06,
+      "loss": 0.8204,
+      "step": 12425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9765828603415957,
+      "learning_rate": 1.0590601102421916e-06,
+      "loss": 0.708,
+      "step": 12426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3373515274913057,
+      "learning_rate": 1.0584223048768427e-06,
+      "loss": 0.7246,
+      "step": 12427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9584184830555036,
+      "learning_rate": 1.0577846688898652e-06,
+      "loss": 0.7855,
+      "step": 12428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.193455768934837,
+      "learning_rate": 1.0571472023086604e-06,
+      "loss": 0.6079,
+      "step": 12429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.527021142943165,
+      "learning_rate": 1.0565099051606214e-06,
+      "loss": 0.7062,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9650074624865321,
+      "learning_rate": 1.055872777473133e-06,
+      "loss": 0.9924,
+      "step": 12431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1476218102573164,
+      "learning_rate": 1.0552358192735784e-06,
+      "loss": 0.8805,
+      "step": 12432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9524608796236225,
+      "learning_rate": 1.0545990305893233e-06,
+      "loss": 0.7652,
+      "step": 12433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2544326369929717,
+      "learning_rate": 1.053962411447736e-06,
+      "loss": 0.6636,
+      "step": 12434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8598679843928687,
+      "learning_rate": 1.0533259618761738e-06,
+      "loss": 0.6679,
+      "step": 12435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2166957545851473,
+      "learning_rate": 1.0526896819019817e-06,
+      "loss": 0.7406,
+      "step": 12436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.425325684000458,
+      "learning_rate": 1.0520535715525066e-06,
+      "loss": 0.7043,
+      "step": 12437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9760595006569455,
+      "learning_rate": 1.0514176308550817e-06,
+      "loss": 0.7837,
+      "step": 12438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.550718765514682,
+      "learning_rate": 1.0507818598370355e-06,
+      "loss": 0.7652,
+      "step": 12439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9829041524998934,
+      "learning_rate": 1.0501462585256883e-06,
+      "loss": 0.8222,
+      "step": 12440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.307690092991363,
+      "learning_rate": 1.0495108269483518e-06,
+      "loss": 1.007,
+      "step": 12441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.123558388748662,
+      "learning_rate": 1.0488755651323358e-06,
+      "loss": 0.8258,
+      "step": 12442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.940507488179387,
+      "learning_rate": 1.048240473104934e-06,
+      "loss": 0.6458,
+      "step": 12443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1947713216790494,
+      "learning_rate": 1.0476055508934408e-06,
+      "loss": 0.9841,
+      "step": 12444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9553911764965668,
+      "learning_rate": 1.04697079852514e-06,
+      "loss": 0.8177,
+      "step": 12445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.082543893747841,
+      "learning_rate": 1.0463362160273076e-06,
+      "loss": 0.6547,
+      "step": 12446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.107093391956746,
+      "learning_rate": 1.0457018034272136e-06,
+      "loss": 0.7016,
+      "step": 12447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7146816773593203,
+      "learning_rate": 1.04506756075212e-06,
+      "loss": 0.8405,
+      "step": 12448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2372062068393093,
+      "learning_rate": 1.0444334880292794e-06,
+      "loss": 0.7577,
+      "step": 12449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6096786961904852,
+      "learning_rate": 1.0437995852859445e-06,
+      "loss": 0.7337,
+      "step": 12450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9881357921074314,
+      "learning_rate": 1.0431658525493498e-06,
+      "loss": 0.8788,
+      "step": 12451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2322198752516806,
+      "learning_rate": 1.0425322898467315e-06,
+      "loss": 0.6329,
+      "step": 12452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0911337732396336,
+      "learning_rate": 1.0418988972053162e-06,
+      "loss": 0.7258,
+      "step": 12453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.381683850333478,
+      "learning_rate": 1.0412656746523182e-06,
+      "loss": 0.8003,
+      "step": 12454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.122866012897615,
+      "learning_rate": 1.0406326222149516e-06,
+      "loss": 0.792,
+      "step": 12455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1311729351773065,
+      "learning_rate": 1.0399997399204186e-06,
+      "loss": 0.7448,
+      "step": 12456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.171575180826219,
+      "learning_rate": 1.0393670277959174e-06,
+      "loss": 0.7678,
+      "step": 12457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.899938441741303,
+      "learning_rate": 1.038734485868635e-06,
+      "loss": 0.7552,
+      "step": 12458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0256196704611025,
+      "learning_rate": 1.0381021141657526e-06,
+      "loss": 0.77,
+      "step": 12459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1471633176083027,
+      "learning_rate": 1.037469912714449e-06,
+      "loss": 0.8886,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4250809769631654,
+      "learning_rate": 1.0368378815418856e-06,
+      "loss": 0.7036,
+      "step": 12461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1268387035360017,
+      "learning_rate": 1.036206020675226e-06,
+      "loss": 0.9476,
+      "step": 12462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0849243548786665,
+      "learning_rate": 1.0355743301416215e-06,
+      "loss": 0.7947,
+      "step": 12463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.119626952869065,
+      "learning_rate": 1.0349428099682173e-06,
+      "loss": 0.8753,
+      "step": 12464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2254963234411398,
+      "learning_rate": 1.0343114601821513e-06,
+      "loss": 0.7154,
+      "step": 12465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.25564951055632,
+      "learning_rate": 1.033680280810554e-06,
+      "loss": 0.7559,
+      "step": 12466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.217789505453632,
+      "learning_rate": 1.0330492718805469e-06,
+      "loss": 0.61,
+      "step": 12467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4121677225549245,
+      "learning_rate": 1.0324184334192505e-06,
+      "loss": 0.8185,
+      "step": 12468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0410845688973036,
+      "learning_rate": 1.0317877654537672e-06,
+      "loss": 0.7176,
+      "step": 12469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.387537595863393,
+      "learning_rate": 1.031157268011203e-06,
+      "loss": 0.9148,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6662226052520797,
+      "learning_rate": 1.0305269411186502e-06,
+      "loss": 0.868,
+      "step": 12471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.422681133114633,
+      "learning_rate": 1.0298967848031948e-06,
+      "loss": 0.82,
+      "step": 12472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2989612891568716,
+      "learning_rate": 1.0292667990919164e-06,
+      "loss": 0.8269,
+      "step": 12473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1342640899075915,
+      "learning_rate": 1.0286369840118859e-06,
+      "loss": 0.9979,
+      "step": 12474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.406477069801803,
+      "learning_rate": 1.0280073395901719e-06,
+      "loss": 0.9121,
+      "step": 12475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.388330825093634,
+      "learning_rate": 1.0273778658538263e-06,
+      "loss": 0.9204,
+      "step": 12476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4773831725340116,
+      "learning_rate": 1.0267485628299007e-06,
+      "loss": 1.0332,
+      "step": 12477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5195214842940916,
+      "learning_rate": 1.026119430545441e-06,
+      "loss": 0.7654,
+      "step": 12478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8590794961378212,
+      "learning_rate": 1.025490469027477e-06,
+      "loss": 0.8331,
+      "step": 12479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.748944076765214,
+      "learning_rate": 1.0248616783030403e-06,
+      "loss": 0.9321,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1179999111332557,
+      "learning_rate": 1.0242330583991507e-06,
+      "loss": 0.956,
+      "step": 12481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.639688817715923,
+      "learning_rate": 1.0236046093428204e-06,
+      "loss": 0.6552,
+      "step": 12482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2997271636083125,
+      "learning_rate": 1.0229763311610563e-06,
+      "loss": 0.8131,
+      "step": 12483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8352640557755795,
+      "learning_rate": 1.0223482238808557e-06,
+      "loss": 0.7839,
+      "step": 12484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.300761969249262,
+      "learning_rate": 1.0217202875292115e-06,
+      "loss": 0.9561,
+      "step": 12485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.509603241149801,
+      "learning_rate": 1.021092522133108e-06,
+      "loss": 1.0099,
+      "step": 12486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.309424633431387,
+      "learning_rate": 1.0204649277195178e-06,
+      "loss": 0.8194,
+      "step": 12487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.17543177175535,
+      "learning_rate": 1.0198375043154142e-06,
+      "loss": 0.7091,
+      "step": 12488
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1737116553807017,
+      "learning_rate": 1.0192102519477565e-06,
+      "loss": 0.8356,
+      "step": 12489
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.915774826658427,
+      "learning_rate": 1.0185831706435007e-06,
+      "loss": 1.077,
+      "step": 12490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4024895920976137,
+      "learning_rate": 1.0179562604295929e-06,
+      "loss": 0.9241,
+      "step": 12491
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.808240660416366,
+      "learning_rate": 1.0173295213329714e-06,
+      "loss": 0.8722,
+      "step": 12492
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.556333907573232,
+      "learning_rate": 1.0167029533805733e-06,
+      "loss": 0.8755,
+      "step": 12493
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2070696260942455,
+      "learning_rate": 1.016076556599318e-06,
+      "loss": 0.7134,
+      "step": 12494
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3635627399329127,
+      "learning_rate": 1.0154503310161269e-06,
+      "loss": 0.8154,
+      "step": 12495
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9414499220575785,
+      "learning_rate": 1.014824276657909e-06,
+      "loss": 0.8323,
+      "step": 12496
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7598217871050665,
+      "learning_rate": 1.0141983935515675e-06,
+      "loss": 0.7519,
+      "step": 12497
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.173819021177274,
+      "learning_rate": 1.0135726817239983e-06,
+      "loss": 0.8438,
+      "step": 12498
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.081060736469185,
+      "learning_rate": 1.0129471412020886e-06,
+      "loss": 0.8833,
+      "step": 12499
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9411412676125235,
+      "learning_rate": 1.0123217720127203e-06,
+      "loss": 0.7166,
+      "step": 12500
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2185388162958253,
+      "learning_rate": 1.011696574182766e-06,
+      "loss": 0.9315,
+      "step": 12501
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.985007072471876,
+      "learning_rate": 1.0110715477390915e-06,
+      "loss": 0.6076,
+      "step": 12502
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.696520603757214,
+      "learning_rate": 1.0104466927085577e-06,
+      "loss": 0.821,
+      "step": 12503
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8922787328050714,
+      "learning_rate": 1.0098220091180145e-06,
+      "loss": 0.7253,
+      "step": 12504
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4261136780446084,
+      "learning_rate": 1.0091974969943064e-06,
+      "loss": 0.812,
+      "step": 12505
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.084597348357605,
+      "learning_rate": 1.0085731563642697e-06,
+      "loss": 0.6129,
+      "step": 12506
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3888058044962794,
+      "learning_rate": 1.0079489872547338e-06,
+      "loss": 0.9691,
+      "step": 12507
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0546712464984775,
+      "learning_rate": 1.0073249896925208e-06,
+      "loss": 0.6465,
+      "step": 12508
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.20659121698784,
+      "learning_rate": 1.006701163704445e-06,
+      "loss": 0.6139,
+      "step": 12509
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8393316382660287,
+      "learning_rate": 1.0060775093173126e-06,
+      "loss": 0.7958,
+      "step": 12510
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.403943455330177,
+      "learning_rate": 1.005454026557927e-06,
+      "loss": 0.6263,
+      "step": 12511
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1948937343506643,
+      "learning_rate": 1.0048307154530763e-06,
+      "loss": 0.6981,
+      "step": 12512
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4735433045223623,
+      "learning_rate": 1.0042075760295477e-06,
+      "loss": 0.6746,
+      "step": 12513
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1673460561843756,
+      "learning_rate": 1.0035846083141193e-06,
+      "loss": 0.8072,
+      "step": 12514
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8211296618382808,
+      "learning_rate": 1.002961812333561e-06,
+      "loss": 0.7243,
+      "step": 12515
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3961377261092904,
+      "learning_rate": 1.0023391881146349e-06,
+      "loss": 0.9441,
+      "step": 12516
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.54061552666429,
+      "learning_rate": 1.0017167356840974e-06,
+      "loss": 0.7561,
+      "step": 12517
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.978940508490949,
+      "learning_rate": 1.0010944550686968e-06,
+      "loss": 0.6771,
+      "step": 12518
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9471953641532267,
+      "learning_rate": 1.0004723462951732e-06,
+      "loss": 0.815,
+      "step": 12519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.988006988010253,
+      "learning_rate": 9.99850409390259e-07,
+      "loss": 0.8119,
+      "step": 12520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.71859625373251,
+      "learning_rate": 9.992286443806825e-07,
+      "loss": 0.9904,
+      "step": 12521
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4334618753416866,
+      "learning_rate": 9.986070512931616e-07,
+      "loss": 0.941,
+      "step": 12522
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1296235358524775,
+      "learning_rate": 9.97985630154407e-07,
+      "loss": 0.8632,
+      "step": 12523
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0683111866716817,
+      "learning_rate": 9.973643809911238e-07,
+      "loss": 0.7525,
+      "step": 12524
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.102585080847973,
+      "learning_rate": 9.967433038300067e-07,
+      "loss": 0.6801,
+      "step": 12525
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.208225105317161,
+      "learning_rate": 9.96122398697746e-07,
+      "loss": 0.8236,
+      "step": 12526
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6709412418714389,
+      "learning_rate": 9.955016656210226e-07,
+      "loss": 0.8937,
+      "step": 12527
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1807412140913656,
+      "learning_rate": 9.948811046265095e-07,
+      "loss": 0.7197,
+      "step": 12528
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0597225359737112,
+      "learning_rate": 9.942607157408784e-07,
+      "loss": 0.6991,
+      "step": 12529
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7146640190018476,
+      "learning_rate": 9.936404989907828e-07,
+      "loss": 0.7492,
+      "step": 12530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.738125079653066,
+      "learning_rate": 9.930204544028787e-07,
+      "loss": 0.9037,
+      "step": 12531
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.113193963661604,
+      "learning_rate": 9.924005820038102e-07,
+      "loss": 0.8295,
+      "step": 12532
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.740131215515999,
+      "learning_rate": 9.917808818202135e-07,
+      "loss": 0.7642,
+      "step": 12533
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9153758746698593,
+      "learning_rate": 9.911613538787196e-07,
+      "loss": 0.7726,
+      "step": 12534
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1831601396592144,
+      "learning_rate": 9.905419982059488e-07,
+      "loss": 0.7341,
+      "step": 12535
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.959332604352912,
+      "learning_rate": 9.899228148285206e-07,
+      "loss": 0.7497,
+      "step": 12536
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.172686124873355,
+      "learning_rate": 9.89303803773039e-07,
+      "loss": 0.945,
+      "step": 12537
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.44162657197384,
+      "learning_rate": 9.886849650661035e-07,
+      "loss": 0.7884,
+      "step": 12538
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.155099624540843,
+      "learning_rate": 9.880662987343103e-07,
+      "loss": 0.9184,
+      "step": 12539
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.66231241098341,
+      "learning_rate": 9.87447804804243e-07,
+      "loss": 0.8674,
+      "step": 12540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.837890652183946,
+      "learning_rate": 9.868294833024805e-07,
+      "loss": 0.6856,
+      "step": 12541
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.118627558199026,
+      "learning_rate": 9.862113342555924e-07,
+      "loss": 0.6968,
+      "step": 12542
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6573287638561656,
+      "learning_rate": 9.855933576901428e-07,
+      "loss": 0.6633,
+      "step": 12543
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6311728280963216,
+      "learning_rate": 9.849755536326866e-07,
+      "loss": 0.8172,
+      "step": 12544
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.6110214409340395,
+      "learning_rate": 9.843579221097722e-07,
+      "loss": 0.6688,
+      "step": 12545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2408126961294204,
+      "learning_rate": 9.837404631479419e-07,
+      "loss": 0.9301,
+      "step": 12546
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.176084145511143,
+      "learning_rate": 9.831231767737305e-07,
+      "loss": 0.7395,
+      "step": 12547
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.331530557858496,
+      "learning_rate": 9.82506063013659e-07,
+      "loss": 0.9351,
+      "step": 12548
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.071632682244574,
+      "learning_rate": 9.818891218942511e-07,
+      "loss": 0.7388,
+      "step": 12549
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.2920678494283715,
+      "learning_rate": 9.812723534420165e-07,
+      "loss": 0.8988,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7718403591546983,
+      "learning_rate": 9.806557576834591e-07,
+      "loss": 0.7336,
+      "step": 12551
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3256788158016857,
+      "learning_rate": 9.800393346450754e-07,
+      "loss": 0.7134,
+      "step": 12552
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4489767246721232,
+      "learning_rate": 9.794230843533525e-07,
+      "loss": 0.8848,
+      "step": 12553
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2540090115006,
+      "learning_rate": 9.78807006834777e-07,
+      "loss": 0.9126,
+      "step": 12554
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2291766230126813,
+      "learning_rate": 9.781911021158181e-07,
+      "loss": 0.736,
+      "step": 12555
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1846376606465925,
+      "learning_rate": 9.775753702229452e-07,
+      "loss": 0.9937,
+      "step": 12556
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7575950045932682,
+      "learning_rate": 9.769598111826166e-07,
+      "loss": 0.7721,
+      "step": 12557
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7376308949385904,
+      "learning_rate": 9.763444250212855e-07,
+      "loss": 0.8038,
+      "step": 12558
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.324806643741342,
+      "learning_rate": 9.757292117653955e-07,
+      "loss": 0.8885,
+      "step": 12559
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.478717400806697,
+      "learning_rate": 9.751141714413836e-07,
+      "loss": 0.7743,
+      "step": 12560
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0017007837890308,
+      "learning_rate": 9.7449930407568e-07,
+      "loss": 0.574,
+      "step": 12561
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0775354220271802,
+      "learning_rate": 9.738846096947064e-07,
+      "loss": 0.5857,
+      "step": 12562
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7771104591221514,
+      "learning_rate": 9.732700883248763e-07,
+      "loss": 0.7302,
+      "step": 12563
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7351328119521738,
+      "learning_rate": 9.726557399925995e-07,
+      "loss": 0.7014,
+      "step": 12564
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8994326239819448,
+      "learning_rate": 9.72041564724277e-07,
+      "loss": 0.8476,
+      "step": 12565
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.469710921688951,
+      "learning_rate": 9.71427562546296e-07,
+      "loss": 0.8593,
+      "step": 12566
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7512605714088325,
+      "learning_rate": 9.708137334850465e-07,
+      "loss": 0.9266,
+      "step": 12567
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1978837967860563,
+      "learning_rate": 9.702000775669041e-07,
+      "loss": 0.7052,
+      "step": 12568
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3493683240228624,
+      "learning_rate": 9.695865948182392e-07,
+      "loss": 0.8268,
+      "step": 12569
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5021696212577136,
+      "learning_rate": 9.689732852654143e-07,
+      "loss": 0.8222,
+      "step": 12570
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0539023094310873,
+      "learning_rate": 9.683601489347843e-07,
+      "loss": 0.7815,
+      "step": 12571
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2075818372972766,
+      "learning_rate": 9.677471858526998e-07,
+      "loss": 0.6954,
+      "step": 12572
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2394401340616101,
+      "learning_rate": 9.671343960454965e-07,
+      "loss": 0.6311,
+      "step": 12573
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.280335367640486,
+      "learning_rate": 9.66521779539511e-07,
+      "loss": 0.7891,
+      "step": 12574
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.084314229843171,
+      "learning_rate": 9.659093363610678e-07,
+      "loss": 0.8842,
+      "step": 12575
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.467297327413809,
+      "learning_rate": 9.652970665364847e-07,
+      "loss": 0.7555,
+      "step": 12576
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1557100513198881,
+      "learning_rate": 9.646849700920729e-07,
+      "loss": 0.6764,
+      "step": 12577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.337052806896527,
+      "learning_rate": 9.640730470541343e-07,
+      "loss": 0.6786,
+      "step": 12578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3753522669240654,
+      "learning_rate": 9.63461297448966e-07,
+      "loss": 0.9287,
+      "step": 12579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.485582912190357,
+      "learning_rate": 9.628497213028553e-07,
+      "loss": 0.6025,
+      "step": 12580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.829957600789762,
+      "learning_rate": 9.622383186420818e-07,
+      "loss": 0.7634,
+      "step": 12581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.346582888462913,
+      "learning_rate": 9.616270894929219e-07,
+      "loss": 0.7147,
+      "step": 12582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5314515731894285,
+      "learning_rate": 9.610160338816404e-07,
+      "loss": 0.927,
+      "step": 12583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0627656964890217,
+      "learning_rate": 9.604051518344948e-07,
+      "loss": 0.6936,
+      "step": 12584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2862622958732475,
+      "learning_rate": 9.597944433777362e-07,
+      "loss": 0.7236,
+      "step": 12585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8852711381603073,
+      "learning_rate": 9.59183908537607e-07,
+      "loss": 0.56,
+      "step": 12586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5789761765351504,
+      "learning_rate": 9.585735473403479e-07,
+      "loss": 0.7674,
+      "step": 12587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.485860747373381,
+      "learning_rate": 9.579633598121824e-07,
+      "loss": 0.8894,
+      "step": 12588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8820766778321805,
+      "learning_rate": 9.57353345979332e-07,
+      "loss": 0.7934,
+      "step": 12589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.277749880829679,
+      "learning_rate": 9.567435058680146e-07,
+      "loss": 0.7347,
+      "step": 12590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.31923795623123,
+      "learning_rate": 9.56133839504431e-07,
+      "loss": 0.6734,
+      "step": 12591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2149021855201922,
+      "learning_rate": 9.55524346914784e-07,
+      "loss": 0.568,
+      "step": 12592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.278434571333933,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.9758,
+      "step": 12593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0838992613151577,
+      "learning_rate": 9.543058831620528e-07,
+      "loss": 0.7254,
+      "step": 12594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.091799237240556,
+      "learning_rate": 9.536969120513284e-07,
+      "loss": 0.7399,
+      "step": 12595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.113950955207978,
+      "learning_rate": 9.530881148192578e-07,
+      "loss": 0.9195,
+      "step": 12596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6971362043652998,
+      "learning_rate": 9.524794914920072e-07,
+      "loss": 0.7628,
+      "step": 12597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1846631826563887,
+      "learning_rate": 9.51871042095725e-07,
+      "loss": 0.6822,
+      "step": 12598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0236085856115864,
+      "learning_rate": 9.512627666565588e-07,
+      "loss": 0.6046,
+      "step": 12599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9134279580315001,
+      "learning_rate": 9.506546652006504e-07,
+      "loss": 0.637,
+      "step": 12600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5502547330831917,
+      "learning_rate": 9.500467377541289e-07,
+      "loss": 0.9519,
+      "step": 12601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.23889777853351,
+      "learning_rate": 9.49438984343119e-07,
+      "loss": 0.8171,
+      "step": 12602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.164743341307121,
+      "learning_rate": 9.48831404993737e-07,
+      "loss": 0.736,
+      "step": 12603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8472299362896845,
+      "learning_rate": 9.482239997320903e-07,
+      "loss": 0.7953,
+      "step": 12604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0922749837203845,
+      "learning_rate": 9.476167685842852e-07,
+      "loss": 0.9763,
+      "step": 12605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.990787345145949,
+      "learning_rate": 9.470097115764099e-07,
+      "loss": 0.6107,
+      "step": 12606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6270879905212552,
+      "learning_rate": 9.464028287345551e-07,
+      "loss": 0.7701,
+      "step": 12607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7925440208997316,
+      "learning_rate": 9.457961200847998e-07,
+      "loss": 0.7496,
+      "step": 12608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5831871213565836,
+      "learning_rate": 9.451895856532117e-07,
+      "loss": 0.8388,
+      "step": 12609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.314478009626272,
+      "learning_rate": 9.445832254658594e-07,
+      "loss": 0.7902,
+      "step": 12610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.118018624861795,
+      "learning_rate": 9.439770395487974e-07,
+      "loss": 0.7861,
+      "step": 12611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0585342773372557,
+      "learning_rate": 9.43371027928075e-07,
+      "loss": 0.6491,
+      "step": 12612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2892854057621914,
+      "learning_rate": 9.427651906297347e-07,
+      "loss": 0.9809,
+      "step": 12613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.210478487909907,
+      "learning_rate": 9.421595276798084e-07,
+      "loss": 0.6424,
+      "step": 12614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.934501238270288,
+      "learning_rate": 9.415540391043276e-07,
+      "loss": 0.9128,
+      "step": 12615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8611938064638034,
+      "learning_rate": 9.40948724929307e-07,
+      "loss": 0.7079,
+      "step": 12616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.123516946001674,
+      "learning_rate": 9.403435851807579e-07,
+      "loss": 0.6582,
+      "step": 12617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.597987668001577,
+      "learning_rate": 9.397386198846881e-07,
+      "loss": 0.6576,
+      "step": 12618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5973072313312917,
+      "learning_rate": 9.39133829067092e-07,
+      "loss": 0.8209,
+      "step": 12619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.219225091519967,
+      "learning_rate": 9.385292127539597e-07,
+      "loss": 0.8021,
+      "step": 12620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.303555325960793,
+      "learning_rate": 9.379247709712725e-07,
+      "loss": 0.8551,
+      "step": 12621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0116411603483701,
+      "learning_rate": 9.373205037450028e-07,
+      "loss": 0.6856,
+      "step": 12622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9982917445127963,
+      "learning_rate": 9.367164111011223e-07,
+      "loss": 0.7586,
+      "step": 12623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0688233855698286,
+      "learning_rate": 9.361124930655841e-07,
+      "loss": 0.8911,
+      "step": 12624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.018701839190244,
+      "learning_rate": 9.355087496643444e-07,
+      "loss": 0.6529,
+      "step": 12625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6227192457719153,
+      "learning_rate": 9.349051809233472e-07,
+      "loss": 0.6171,
+      "step": 12626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0778056530944373,
+      "learning_rate": 9.34301786868525e-07,
+      "loss": 0.7813,
+      "step": 12627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.568538096218923,
+      "learning_rate": 9.336985675258109e-07,
+      "loss": 0.9026,
+      "step": 12628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.329126660476712,
+      "learning_rate": 9.330955229211259e-07,
+      "loss": 0.8367,
+      "step": 12629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9999210757650006,
+      "learning_rate": 9.324926530803835e-07,
+      "loss": 0.9601,
+      "step": 12630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.46363325002627,
+      "learning_rate": 9.31889958029491e-07,
+      "loss": 0.8704,
+      "step": 12631
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8023007482014222,
+      "learning_rate": 9.312874377943454e-07,
+      "loss": 0.6881,
+      "step": 12632
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1473045728445834,
+      "learning_rate": 9.306850924008415e-07,
+      "loss": 0.5485,
+      "step": 12633
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1394010259555625,
+      "learning_rate": 9.300829218748625e-07,
+      "loss": 0.7289,
+      "step": 12634
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.178272218173307,
+      "learning_rate": 9.294809262422838e-07,
+      "loss": 0.6945,
+      "step": 12635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.247846237602465,
+      "learning_rate": 9.288791055289759e-07,
+      "loss": 0.8066,
+      "step": 12636
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1333090409341933,
+      "learning_rate": 9.282774597607991e-07,
+      "loss": 0.5892,
+      "step": 12637
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1052801708910895,
+      "learning_rate": 9.276759889636084e-07,
+      "loss": 0.6506,
+      "step": 12638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.303815397489276,
+      "learning_rate": 9.270746931632501e-07,
+      "loss": 0.6791,
+      "step": 12639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1407802271272476,
+      "learning_rate": 9.264735723855617e-07,
+      "loss": 0.7057,
+      "step": 12640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.685507890078458,
+      "learning_rate": 9.258726266563789e-07,
+      "loss": 0.734,
+      "step": 12641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.868229724743927,
+      "learning_rate": 9.2527185600152e-07,
+      "loss": 0.8801,
+      "step": 12642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7288409829960334,
+      "learning_rate": 9.246712604468061e-07,
+      "loss": 0.7567,
+      "step": 12643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.271635976032773,
+      "learning_rate": 9.240708400180437e-07,
+      "loss": 0.795,
+      "step": 12644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9345023021954342,
+      "learning_rate": 9.234705947410355e-07,
+      "loss": 0.8035,
+      "step": 12645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3744546223351204,
+      "learning_rate": 9.228705246415742e-07,
+      "loss": 0.6829,
+      "step": 12646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0698089343133317,
+      "learning_rate": 9.222706297454459e-07,
+      "loss": 0.7118,
+      "step": 12647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7727451874919216,
+      "learning_rate": 9.216709100784326e-07,
+      "loss": 0.8261,
+      "step": 12648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7846189284247718,
+      "learning_rate": 9.210713656663023e-07,
+      "loss": 0.6745,
+      "step": 12649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5182040249132704,
+      "learning_rate": 9.20471996534818e-07,
+      "loss": 0.6875,
+      "step": 12650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.005461830632184,
+      "learning_rate": 9.198728027097386e-07,
+      "loss": 0.6843,
+      "step": 12651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7067742368667052,
+      "learning_rate": 9.192737842168126e-07,
+      "loss": 0.7673,
+      "step": 12652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.306609228451892,
+      "learning_rate": 9.186749410817797e-07,
+      "loss": 0.9444,
+      "step": 12653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.483633800151044,
+      "learning_rate": 9.180762733303745e-07,
+      "loss": 0.6917,
+      "step": 12654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1553416539462584,
+      "learning_rate": 9.174777809883229e-07,
+      "loss": 0.8769,
+      "step": 12655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0565133076429767,
+      "learning_rate": 9.168794640813428e-07,
+      "loss": 0.6544,
+      "step": 12656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0409675794904403,
+      "learning_rate": 9.162813226351447e-07,
+      "loss": 0.7002,
+      "step": 12657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4440113665074223,
+      "learning_rate": 9.156833566754347e-07,
+      "loss": 0.7848,
+      "step": 12658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.158479099541898,
+      "learning_rate": 9.150855662279079e-07,
+      "loss": 0.8838,
+      "step": 12659
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4709326411591044,
+      "learning_rate": 9.144879513182498e-07,
+      "loss": 0.883,
+      "step": 12660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0059961455243904,
+      "learning_rate": 9.138905119721442e-07,
+      "loss": 0.8824,
+      "step": 12661
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.612879984220134,
+      "learning_rate": 9.13293248215264e-07,
+      "loss": 0.7654,
+      "step": 12662
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0536856445678189,
+      "learning_rate": 9.126961600732742e-07,
+      "loss": 0.5712,
+      "step": 12663
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1453325536884948,
+      "learning_rate": 9.120992475718333e-07,
+      "loss": 0.6805,
+      "step": 12664
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8277646968501322,
+      "learning_rate": 9.115025107365904e-07,
+      "loss": 0.713,
+      "step": 12665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.792994155262683,
+      "learning_rate": 9.109059495931932e-07,
+      "loss": 0.6522,
+      "step": 12666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.727463496039534,
+      "learning_rate": 9.103095641672732e-07,
+      "loss": 0.8511,
+      "step": 12667
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9840728735422197,
+      "learning_rate": 9.097133544844577e-07,
+      "loss": 0.7821,
+      "step": 12668
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1489649975437533,
+      "learning_rate": 9.091173205703708e-07,
+      "loss": 1.0133,
+      "step": 12669
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3169370587787568,
+      "learning_rate": 9.085214624506228e-07,
+      "loss": 0.9393,
+      "step": 12670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.60569940767874,
+      "learning_rate": 9.079257801508201e-07,
+      "loss": 0.8885,
+      "step": 12671
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.484285372342568,
+      "learning_rate": 9.0733027369656e-07,
+      "loss": 0.8307,
+      "step": 12672
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1718016468144357,
+      "learning_rate": 9.067349431134331e-07,
+      "loss": 0.711,
+      "step": 12673
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.967072280269619,
+      "learning_rate": 9.061397884270217e-07,
+      "loss": 0.6738,
+      "step": 12674
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8893890634151544,
+      "learning_rate": 9.055448096628999e-07,
+      "loss": 0.8829,
+      "step": 12675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1927291639058613,
+      "learning_rate": 9.049500068466377e-07,
+      "loss": 0.7326,
+      "step": 12676
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3076315861348657,
+      "learning_rate": 9.043553800037952e-07,
+      "loss": 0.7877,
+      "step": 12677
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0889290273324688,
+      "learning_rate": 9.037609291599214e-07,
+      "loss": 0.6273,
+      "step": 12678
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0374650528501952,
+      "learning_rate": 9.031666543405637e-07,
+      "loss": 0.7523,
+      "step": 12679
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.0662765477414995,
+      "learning_rate": 9.025725555712595e-07,
+      "loss": 0.7284,
+      "step": 12680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8534913370583948,
+      "learning_rate": 9.019786328775382e-07,
+      "loss": 0.7664,
+      "step": 12681
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.459091798542901,
+      "learning_rate": 9.013848862849217e-07,
+      "loss": 0.8514,
+      "step": 12682
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.336424855618919,
+      "learning_rate": 9.007913158189236e-07,
+      "loss": 1.0094,
+      "step": 12683
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3432639740384045,
+      "learning_rate": 9.001979215050544e-07,
+      "loss": 0.8568,
+      "step": 12684
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.328778066012769,
+      "learning_rate": 8.996047033688083e-07,
+      "loss": 0.9247,
+      "step": 12685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2556989129648854,
+      "learning_rate": 8.990116614356819e-07,
+      "loss": 0.6569,
+      "step": 12686
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1288613323481727,
+      "learning_rate": 8.984187957311579e-07,
+      "loss": 0.5503,
+      "step": 12687
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2332751773020894,
+      "learning_rate": 8.978261062807131e-07,
+      "loss": 0.7923,
+      "step": 12688
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5084567442396204,
+      "learning_rate": 8.972335931098159e-07,
+      "loss": 0.9741,
+      "step": 12689
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1773280230238288,
+      "learning_rate": 8.966412562439291e-07,
+      "loss": 0.6268,
+      "step": 12690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4049887148011266,
+      "learning_rate": 8.960490957085061e-07,
+      "loss": 0.9287,
+      "step": 12691
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0669594899959156,
+      "learning_rate": 8.954571115289934e-07,
+      "loss": 0.7887,
+      "step": 12692
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8314065891524693,
+      "learning_rate": 8.948653037308286e-07,
+      "loss": 0.7326,
+      "step": 12693
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.892994019859524,
+      "learning_rate": 8.942736723394458e-07,
+      "loss": 0.7786,
+      "step": 12694
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6069682510201306,
+      "learning_rate": 8.936822173802667e-07,
+      "loss": 0.7873,
+      "step": 12695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1092671163326964,
+      "learning_rate": 8.930909388787084e-07,
+      "loss": 0.5626,
+      "step": 12696
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9315706170269044,
+      "learning_rate": 8.92499836860179e-07,
+      "loss": 0.6481,
+      "step": 12697
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2684988310342113,
+      "learning_rate": 8.919089113500795e-07,
+      "loss": 0.8274,
+      "step": 12698
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8603632399218637,
+      "learning_rate": 8.913181623738032e-07,
+      "loss": 0.7754,
+      "step": 12699
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.548157881959274,
+      "learning_rate": 8.907275899567363e-07,
+      "loss": 0.8668,
+      "step": 12700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.109269633339493,
+      "learning_rate": 8.901371941242554e-07,
+      "loss": 0.599,
+      "step": 12701
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5798826122285754,
+      "learning_rate": 8.895469749017344e-07,
+      "loss": 0.8158,
+      "step": 12702
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9529311545469554,
+      "learning_rate": 8.889569323145325e-07,
+      "loss": 0.7112,
+      "step": 12703
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4489408643890047,
+      "learning_rate": 8.883670663880078e-07,
+      "loss": 0.9961,
+      "step": 12704
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.299009386126066,
+      "learning_rate": 8.877773771475074e-07,
+      "loss": 0.7826,
+      "step": 12705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2962450366498617,
+      "learning_rate": 8.871878646183718e-07,
+      "loss": 0.9176,
+      "step": 12706
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.641016710679177,
+      "learning_rate": 8.865985288259332e-07,
+      "loss": 0.5888,
+      "step": 12707
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.971990960356103,
+      "learning_rate": 8.860093697955152e-07,
+      "loss": 0.9489,
+      "step": 12708
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0652611794181763,
+      "learning_rate": 8.854203875524403e-07,
+      "loss": 0.841,
+      "step": 12709
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.947382393788665,
+      "learning_rate": 8.848315821220133e-07,
+      "loss": 0.947,
+      "step": 12710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.34098398689466,
+      "learning_rate": 8.842429535295366e-07,
+      "loss": 0.9185,
+      "step": 12711
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.319728847449923,
+      "learning_rate": 8.836545018003084e-07,
+      "loss": 0.9262,
+      "step": 12712
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6400088373530453,
+      "learning_rate": 8.830662269596135e-07,
+      "loss": 0.8655,
+      "step": 12713
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.861183194363288,
+      "learning_rate": 8.824781290327317e-07,
+      "loss": 0.9429,
+      "step": 12714
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.05693058208575,
+      "learning_rate": 8.818902080449348e-07,
+      "loss": 0.7146,
+      "step": 12715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.882930486515866,
+      "learning_rate": 8.813024640214873e-07,
+      "loss": 0.9269,
+      "step": 12716
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.379924351285326,
+      "learning_rate": 8.807148969876455e-07,
+      "loss": 0.7311,
+      "step": 12717
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3317958918358104,
+      "learning_rate": 8.801275069686593e-07,
+      "loss": 0.675,
+      "step": 12718
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5500867322984018,
+      "learning_rate": 8.795402939897679e-07,
+      "loss": 0.7448,
+      "step": 12719
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3338545773064236,
+      "learning_rate": 8.789532580762095e-07,
+      "loss": 0.957,
+      "step": 12720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.007043625684507,
+      "learning_rate": 8.783663992532048e-07,
+      "loss": 0.7407,
+      "step": 12721
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1180251102498888,
+      "learning_rate": 8.777797175459773e-07,
+      "loss": 0.7256,
+      "step": 12722
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8771550723901504,
+      "learning_rate": 8.771932129797356e-07,
+      "loss": 0.8046,
+      "step": 12723
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0838512598974532,
+      "learning_rate": 8.766068855796833e-07,
+      "loss": 0.663,
+      "step": 12724
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2388117355753447,
+      "learning_rate": 8.760207353710165e-07,
+      "loss": 0.7586,
+      "step": 12725
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7899728710438434,
+      "learning_rate": 8.754347623789222e-07,
+      "loss": 0.7848,
+      "step": 12726
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.861579123057241,
+      "learning_rate": 8.748489666285842e-07,
+      "loss": 0.746,
+      "step": 12727
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3441835601095646,
+      "learning_rate": 8.742633481451728e-07,
+      "loss": 0.8563,
+      "step": 12728
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2402364912711201,
+      "learning_rate": 8.736779069538521e-07,
+      "loss": 0.7188,
+      "step": 12729
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2853041956950766,
+      "learning_rate": 8.730926430797826e-07,
+      "loss": 0.8904,
+      "step": 12730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5847311025880626,
+      "learning_rate": 8.72507556548114e-07,
+      "loss": 0.7104,
+      "step": 12731
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.54324972490937,
+      "learning_rate": 8.719226473839876e-07,
+      "loss": 0.7341,
+      "step": 12732
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.159326760785018,
+      "learning_rate": 8.713379156125385e-07,
+      "loss": 0.9402,
+      "step": 12733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5868077033349484,
+      "learning_rate": 8.707533612588948e-07,
+      "loss": 0.7627,
+      "step": 12734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.599624869454871,
+      "learning_rate": 8.701689843481753e-07,
+      "loss": 0.9172,
+      "step": 12735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1281356489841294,
+      "learning_rate": 8.695847849054906e-07,
+      "loss": 0.8464,
+      "step": 12736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.059515040018947,
+      "learning_rate": 8.690007629559482e-07,
+      "loss": 0.7855,
+      "step": 12737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4376944065232196,
+      "learning_rate": 8.684169185246444e-07,
+      "loss": 0.8257,
+      "step": 12738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2783134026566572,
+      "learning_rate": 8.67833251636665e-07,
+      "loss": 0.6691,
+      "step": 12739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8882040917781506,
+      "learning_rate": 8.672497623170944e-07,
+      "loss": 0.6858,
+      "step": 12740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.692082661264149,
+      "learning_rate": 8.666664505910055e-07,
+      "loss": 0.8609,
+      "step": 12741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.133744055160752,
+      "learning_rate": 8.660833164834653e-07,
+      "loss": 0.8691,
+      "step": 12742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8895936695542708,
+      "learning_rate": 8.65500360019531e-07,
+      "loss": 0.9014,
+      "step": 12743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4368560285038776,
+      "learning_rate": 8.649175812242532e-07,
+      "loss": 0.8819,
+      "step": 12744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.501507451650237,
+      "learning_rate": 8.643349801226791e-07,
+      "loss": 0.8314,
+      "step": 12745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1977442498688653,
+      "learning_rate": 8.637525567398392e-07,
+      "loss": 0.817,
+      "step": 12746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2068399765760087,
+      "learning_rate": 8.631703111007645e-07,
+      "loss": 0.5896,
+      "step": 12747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.139333069043718,
+      "learning_rate": 8.625882432304749e-07,
+      "loss": 0.562,
+      "step": 12748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.053984501064385,
+      "learning_rate": 8.62006353153983e-07,
+      "loss": 0.5985,
+      "step": 12749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.049269049327842,
+      "learning_rate": 8.61424640896294e-07,
+      "loss": 1.0735,
+      "step": 12750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.527472126261148,
+      "learning_rate": 8.608431064824052e-07,
+      "loss": 0.8076,
+      "step": 12751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0150284160892933,
+      "learning_rate": 8.602617499373056e-07,
+      "loss": 0.7268,
+      "step": 12752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5618822642249692,
+      "learning_rate": 8.596805712859807e-07,
+      "loss": 0.9725,
+      "step": 12753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9672141371814906,
+      "learning_rate": 8.590995705533994e-07,
+      "loss": 0.7939,
+      "step": 12754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9480398245694348,
+      "learning_rate": 8.58518747764534e-07,
+      "loss": 0.8201,
+      "step": 12755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4544985886379487,
+      "learning_rate": 8.579381029443412e-07,
+      "loss": 0.7992,
+      "step": 12756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.158713181440663,
+      "learning_rate": 8.573576361177733e-07,
+      "loss": 0.7843,
+      "step": 12757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9607575549585867,
+      "learning_rate": 8.567773473097735e-07,
+      "loss": 0.6584,
+      "step": 12758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6627370177977339,
+      "learning_rate": 8.561972365452775e-07,
+      "loss": 0.856,
+      "step": 12759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8421741413891288,
+      "learning_rate": 8.556173038492171e-07,
+      "loss": 0.671,
+      "step": 12760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8076758470317094,
+      "learning_rate": 8.550375492465102e-07,
+      "loss": 0.6321,
+      "step": 12761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.934397600272884,
+      "learning_rate": 8.5445797276207e-07,
+      "loss": 0.673,
+      "step": 12762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0692474937673597,
+      "learning_rate": 8.538785744208062e-07,
+      "loss": 0.7635,
+      "step": 12763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.760384255648537,
+      "learning_rate": 8.532993542476108e-07,
+      "loss": 0.8624,
+      "step": 12764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7578897749733406,
+      "learning_rate": 8.527203122673789e-07,
+      "loss": 0.7313,
+      "step": 12765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3410369283651593,
+      "learning_rate": 8.521414485049917e-07,
+      "loss": 0.7325,
+      "step": 12766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9739960343053438,
+      "learning_rate": 8.51562762985324e-07,
+      "loss": 0.6764,
+      "step": 12767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.751255055291825,
+      "learning_rate": 8.509842557332437e-07,
+      "loss": 0.6324,
+      "step": 12768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0546114329668903,
+      "learning_rate": 8.504059267736097e-07,
+      "loss": 0.7705,
+      "step": 12769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3281514919886037,
+      "learning_rate": 8.49827776131274e-07,
+      "loss": 0.8205,
+      "step": 12770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.74360769297613,
+      "learning_rate": 8.492498038310843e-07,
+      "loss": 0.9457,
+      "step": 12771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2451548047913668,
+      "learning_rate": 8.486720098978718e-07,
+      "loss": 0.9538,
+      "step": 12772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.064792750140252,
+      "learning_rate": 8.480943943564701e-07,
+      "loss": 0.6021,
+      "step": 12773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2607683401684273,
+      "learning_rate": 8.475169572316988e-07,
+      "loss": 0.8675,
+      "step": 12774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7766590724348734,
+      "learning_rate": 8.469396985483724e-07,
+      "loss": 0.7908,
+      "step": 12775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.634827349709677,
+      "learning_rate": 8.463626183312962e-07,
+      "loss": 0.6805,
+      "step": 12776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.787782043490733,
+      "learning_rate": 8.457857166052674e-07,
+      "loss": 1.0806,
+      "step": 12777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.173302105179479,
+      "learning_rate": 8.452089933950813e-07,
+      "loss": 0.6942,
+      "step": 12778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8613956448169775,
+      "learning_rate": 8.446324487255164e-07,
+      "loss": 0.839,
+      "step": 12779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.492415063840819,
+      "learning_rate": 8.440560826213485e-07,
+      "loss": 0.6623,
+      "step": 12780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.063527565945278,
+      "learning_rate": 8.434798951073492e-07,
+      "loss": 0.7884,
+      "step": 12781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0901256478929398,
+      "learning_rate": 8.429038862082734e-07,
+      "loss": 0.7092,
+      "step": 12782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2370946004451833,
+      "learning_rate": 8.423280559488767e-07,
+      "loss": 0.8013,
+      "step": 12783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3416390805030676,
+      "learning_rate": 8.417524043539038e-07,
+      "loss": 0.7356,
+      "step": 12784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9075984042668093,
+      "learning_rate": 8.411769314480905e-07,
+      "loss": 0.906,
+      "step": 12785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0491649279795137,
+      "learning_rate": 8.406016372561665e-07,
+      "loss": 0.6821,
+      "step": 12786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9756847071520978,
+      "learning_rate": 8.400265218028525e-07,
+      "loss": 0.8444,
+      "step": 12787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3651963370016746,
+      "learning_rate": 8.394515851128654e-07,
+      "loss": 0.6639,
+      "step": 12788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8550579653830777,
+      "learning_rate": 8.388768272109105e-07,
+      "loss": 0.8221,
+      "step": 12789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1431006145798857,
+      "learning_rate": 8.383022481216829e-07,
+      "loss": 0.8933,
+      "step": 12790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2102827470510626,
+      "learning_rate": 8.37727847869878e-07,
+      "loss": 0.9373,
+      "step": 12791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7190818586476717,
+      "learning_rate": 8.371536264801772e-07,
+      "loss": 0.7849,
+      "step": 12792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5606921137442455,
+      "learning_rate": 8.365795839772561e-07,
+      "loss": 0.7612,
+      "step": 12793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.459643620344818,
+      "learning_rate": 8.36005720385783e-07,
+      "loss": 0.6524,
+      "step": 12794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.316400403303693,
+      "learning_rate": 8.354320357304163e-07,
+      "loss": 0.9158,
+      "step": 12795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.957749266834525,
+      "learning_rate": 8.34858530035813e-07,
+      "loss": 0.6012,
+      "step": 12796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.188081577734665,
+      "learning_rate": 8.342852033266119e-07,
+      "loss": 0.7098,
+      "step": 12797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.251492332127374,
+      "learning_rate": 8.337120556274548e-07,
+      "loss": 0.7354,
+      "step": 12798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.038963365303163,
+      "learning_rate": 8.331390869629702e-07,
+      "loss": 0.638,
+      "step": 12799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9457307983054912,
+      "learning_rate": 8.32566297357777e-07,
+      "loss": 0.7868,
+      "step": 12800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5358520901416224,
+      "learning_rate": 8.319936868364925e-07,
+      "loss": 0.8633,
+      "step": 12801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0929500390319624,
+      "learning_rate": 8.314212554237222e-07,
+      "loss": 0.6068,
+      "step": 12802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1483970814889217,
+      "learning_rate": 8.308490031440641e-07,
+      "loss": 0.7149,
+      "step": 12803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.187974241774146,
+      "learning_rate": 8.302769300221098e-07,
+      "loss": 0.7332,
+      "step": 12804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1054387764657891,
+      "learning_rate": 8.29705036082441e-07,
+      "loss": 0.7529,
+      "step": 12805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.073667694847114,
+      "learning_rate": 8.291333213496355e-07,
+      "loss": 0.6154,
+      "step": 12806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7003489231771798,
+      "learning_rate": 8.285617858482609e-07,
+      "loss": 0.883,
+      "step": 12807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0746441808100697,
+      "learning_rate": 8.279904296028757e-07,
+      "loss": 0.8697,
+      "step": 12808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8723032982513708,
+      "learning_rate": 8.274192526380337e-07,
+      "loss": 0.653,
+      "step": 12809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7733466784252183,
+      "learning_rate": 8.268482549782797e-07,
+      "loss": 0.9618,
+      "step": 12810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.054058329214567,
+      "learning_rate": 8.262774366481496e-07,
+      "loss": 0.9119,
+      "step": 12811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1004962843684654,
+      "learning_rate": 8.257067976721739e-07,
+      "loss": 0.6535,
+      "step": 12812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.657780788029301,
+      "learning_rate": 8.251363380748723e-07,
+      "loss": 0.8464,
+      "step": 12813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4150744535658397,
+      "learning_rate": 8.24566057880763e-07,
+      "loss": 0.7282,
+      "step": 12814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.161479965123715,
+      "learning_rate": 8.239959571143463e-07,
+      "loss": 0.8565,
+      "step": 12815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1244008359309627,
+      "learning_rate": 8.234260358001256e-07,
+      "loss": 0.7858,
+      "step": 12816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.112935526441189,
+      "learning_rate": 8.228562939625906e-07,
+      "loss": 0.6658,
+      "step": 12817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.388742890481402,
+      "learning_rate": 8.222867316262217e-07,
+      "loss": 0.616,
+      "step": 12818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.418655965168758,
+      "learning_rate": 8.217173488154972e-07,
+      "loss": 0.7546,
+      "step": 12819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3233896490618333,
+      "learning_rate": 8.211481455548837e-07,
+      "loss": 0.9804,
+      "step": 12820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6530945815961045,
+      "learning_rate": 8.205791218688414e-07,
+      "loss": 0.7911,
+      "step": 12821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1259911415871684,
+      "learning_rate": 8.20010277781822e-07,
+      "loss": 0.7807,
+      "step": 12822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.19155913572145,
+      "learning_rate": 8.194416133182693e-07,
+      "loss": 0.8459,
+      "step": 12823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.88545350084773,
+      "learning_rate": 8.188731285026219e-07,
+      "loss": 0.8808,
+      "step": 12824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.009628302822664,
+      "learning_rate": 8.183048233593083e-07,
+      "loss": 0.8531,
+      "step": 12825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.955297225293683,
+      "learning_rate": 8.177366979127499e-07,
+      "loss": 0.7229,
+      "step": 12826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.980462316489486,
+      "learning_rate": 8.171687521873595e-07,
+      "loss": 0.8541,
+      "step": 12827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2047441278057742,
+      "learning_rate": 8.166009862075435e-07,
+      "loss": 0.7903,
+      "step": 12828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1332704119185386,
+      "learning_rate": 8.160333999977004e-07,
+      "loss": 0.6161,
+      "step": 12829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.24274682582069,
+      "learning_rate": 8.154659935822201e-07,
+      "loss": 0.738,
+      "step": 12830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0921861468482463,
+      "learning_rate": 8.148987669854846e-07,
+      "loss": 0.9035,
+      "step": 12831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0564077917637469,
+      "learning_rate": 8.143317202318718e-07,
+      "loss": 0.6728,
+      "step": 12832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1921336614704297,
+      "learning_rate": 8.137648533457448e-07,
+      "loss": 0.734,
+      "step": 12833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.432502723460482,
+      "learning_rate": 8.131981663514665e-07,
+      "loss": 0.8275,
+      "step": 12834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5530934048659404,
+      "learning_rate": 8.126316592733874e-07,
+      "loss": 0.7717,
+      "step": 12835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0592986327505955,
+      "learning_rate": 8.120653321358519e-07,
+      "loss": 0.7518,
+      "step": 12836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8991022795239512,
+      "learning_rate": 8.114991849631953e-07,
+      "loss": 0.7213,
+      "step": 12837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5825392290652447,
+      "learning_rate": 8.109332177797469e-07,
+      "loss": 0.9072,
+      "step": 12838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7728368918872683,
+      "learning_rate": 8.103674306098291e-07,
+      "loss": 0.8878,
+      "step": 12839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.168083234565149,
+      "learning_rate": 8.098018234777533e-07,
+      "loss": 0.8,
+      "step": 12840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2870174070044857,
+      "learning_rate": 8.092363964078237e-07,
+      "loss": 0.7252,
+      "step": 12841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1286929790789926,
+      "learning_rate": 8.086711494243405e-07,
+      "loss": 0.7091,
+      "step": 12842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9720578017713293,
+      "learning_rate": 8.081060825515924e-07,
+      "loss": 0.7625,
+      "step": 12843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.265768924509737,
+      "learning_rate": 8.075411958138623e-07,
+      "loss": 0.7649,
+      "step": 12844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1935639927743513,
+      "learning_rate": 8.069764892354237e-07,
+      "loss": 0.9577,
+      "step": 12845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.370130097615622,
+      "learning_rate": 8.064119628405443e-07,
+      "loss": 0.8761,
+      "step": 12846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.261166048802068,
+      "learning_rate": 8.058476166534818e-07,
+      "loss": 0.6865,
+      "step": 12847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.787684838659057,
+      "learning_rate": 8.052834506984874e-07,
+      "loss": 0.9067,
+      "step": 12848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1748066440101534,
+      "learning_rate": 8.047194649998063e-07,
+      "loss": 0.8094,
+      "step": 12849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0980386032120495,
+      "learning_rate": 8.041556595816741e-07,
+      "loss": 0.6537,
+      "step": 12850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1931212529673925,
+      "learning_rate": 8.035920344683157e-07,
+      "loss": 0.8786,
+      "step": 12851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.936798875483398,
+      "learning_rate": 8.030285896839546e-07,
+      "loss": 0.8137,
+      "step": 12852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.333590446506807,
+      "learning_rate": 8.024653252528025e-07,
+      "loss": 0.783,
+      "step": 12853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7166970245612987,
+      "learning_rate": 8.019022411990634e-07,
+      "loss": 0.753,
+      "step": 12854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2215578889595093,
+      "learning_rate": 8.013393375469347e-07,
+      "loss": 0.9,
+      "step": 12855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.526746663255358,
+      "learning_rate": 8.007766143206047e-07,
+      "loss": 0.8883,
+      "step": 12856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.177782169037524,
+      "learning_rate": 8.00214071544258e-07,
+      "loss": 0.7018,
+      "step": 12857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6207917117302997,
+      "learning_rate": 7.996517092420636e-07,
+      "loss": 0.9469,
+      "step": 12858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.364257243389768,
+      "learning_rate": 7.99089527438191e-07,
+      "loss": 0.7252,
+      "step": 12859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.540298753774156,
+      "learning_rate": 7.985275261567971e-07,
+      "loss": 0.9491,
+      "step": 12860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1237238452414795,
+      "learning_rate": 7.97965705422033e-07,
+      "loss": 0.8069,
+      "step": 12861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3235957512537357,
+      "learning_rate": 7.974040652580412e-07,
+      "loss": 0.8301,
+      "step": 12862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6379281081761596,
+      "learning_rate": 7.968426056889561e-07,
+      "loss": 0.8016,
+      "step": 12863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3836301707077836,
+      "learning_rate": 7.962813267389052e-07,
+      "loss": 0.701,
+      "step": 12864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0169978890181217,
+      "learning_rate": 7.957202284320076e-07,
+      "loss": 0.7043,
+      "step": 12865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2332277597246653,
+      "learning_rate": 7.951593107923744e-07,
+      "loss": 0.6721,
+      "step": 12866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0527977545134135,
+      "learning_rate": 7.945985738441114e-07,
+      "loss": 0.9419,
+      "step": 12867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0090031323063298,
+      "learning_rate": 7.940380176113149e-07,
+      "loss": 0.7174,
+      "step": 12868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1430343259708615,
+      "learning_rate": 7.93477642118069e-07,
+      "loss": 0.9639,
+      "step": 12869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.720975903578251,
+      "learning_rate": 7.929174473884593e-07,
+      "loss": 0.6078,
+      "step": 12870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.824603906451362,
+      "learning_rate": 7.923574334465562e-07,
+      "loss": 0.6401,
+      "step": 12871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4228171461639514,
+      "learning_rate": 7.917976003164246e-07,
+      "loss": 0.7928,
+      "step": 12872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0421130766080595,
+      "learning_rate": 7.912379480221228e-07,
+      "loss": 0.8034,
+      "step": 12873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1866229565092548,
+      "learning_rate": 7.906784765876985e-07,
+      "loss": 0.7253,
+      "step": 12874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.114017965079287,
+      "learning_rate": 7.90119186037197e-07,
+      "loss": 0.5233,
+      "step": 12875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1242016269094717,
+      "learning_rate": 7.895600763946482e-07,
+      "loss": 0.8821,
+      "step": 12876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2704254066133633,
+      "learning_rate": 7.890011476840809e-07,
+      "loss": 0.9439,
+      "step": 12877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.265360787369283,
+      "learning_rate": 7.884423999295122e-07,
+      "loss": 0.8382,
+      "step": 12878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1426375007158427,
+      "learning_rate": 7.878838331549538e-07,
+      "loss": 0.8576,
+      "step": 12879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.779735102881908,
+      "learning_rate": 7.873254473844077e-07,
+      "loss": 0.893,
+      "step": 12880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.172891132455814,
+      "learning_rate": 7.867672426418693e-07,
+      "loss": 0.7062,
+      "step": 12881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1610749811658805,
+      "learning_rate": 7.86209218951326e-07,
+      "loss": 0.69,
+      "step": 12882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9749139095634773,
+      "learning_rate": 7.856513763367569e-07,
+      "loss": 0.6212,
+      "step": 12883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.113742644950813,
+      "learning_rate": 7.850937148221332e-07,
+      "loss": 0.7864,
+      "step": 12884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9794914680627078,
+      "learning_rate": 7.845362344314206e-07,
+      "loss": 0.7433,
+      "step": 12885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4170886423404414,
+      "learning_rate": 7.839789351885746e-07,
+      "loss": 0.8306,
+      "step": 12886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5226816775194782,
+      "learning_rate": 7.834218171175428e-07,
+      "loss": 0.8262,
+      "step": 12887
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3274088445038923,
+      "learning_rate": 7.828648802422667e-07,
+      "loss": 0.8239,
+      "step": 12888
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0013465572263702,
+      "learning_rate": 7.823081245866776e-07,
+      "loss": 0.8153,
+      "step": 12889
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.185762437891575,
+      "learning_rate": 7.817515501747041e-07,
+      "loss": 0.9038,
+      "step": 12890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.420143060692576,
+      "learning_rate": 7.811951570302594e-07,
+      "loss": 0.7468,
+      "step": 12891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.719240561291769,
+      "learning_rate": 7.806389451772539e-07,
+      "loss": 0.864,
+      "step": 12892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.108790746256074,
+      "learning_rate": 7.800829146395922e-07,
+      "loss": 0.6217,
+      "step": 12893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1122091991212333,
+      "learning_rate": 7.795270654411635e-07,
+      "loss": 0.7979,
+      "step": 12894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6220204082810485,
+      "learning_rate": 7.789713976058578e-07,
+      "loss": 0.8815,
+      "step": 12895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3093552082821844,
+      "learning_rate": 7.784159111575512e-07,
+      "loss": 0.9623,
+      "step": 12896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.179539583517809,
+      "learning_rate": 7.778606061201154e-07,
+      "loss": 0.8258,
+      "step": 12897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.030432537514625,
+      "learning_rate": 7.773054825174126e-07,
+      "loss": 1.0427,
+      "step": 12898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2377703866588645,
+      "learning_rate": 7.767505403732961e-07,
+      "loss": 0.6264,
+      "step": 12899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2317432368309067,
+      "learning_rate": 7.761957797116171e-07,
+      "loss": 0.8538,
+      "step": 12900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1892168786898414,
+      "learning_rate": 7.756412005562114e-07,
+      "loss": 0.8502,
+      "step": 12901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8388404749105376,
+      "learning_rate": 7.750868029309099e-07,
+      "loss": 0.7896,
+      "step": 12902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.557097997496471,
+      "learning_rate": 7.745325868595394e-07,
+      "loss": 0.6648,
+      "step": 12903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.796923714712998,
+      "learning_rate": 7.739785523659144e-07,
+      "loss": 0.8748,
+      "step": 12904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8534962255389162,
+      "learning_rate": 7.734246994738426e-07,
+      "loss": 0.6943,
+      "step": 12905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1518914825622906,
+      "learning_rate": 7.728710282071244e-07,
+      "loss": 0.8124,
+      "step": 12906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0590756350994237,
+      "learning_rate": 7.723175385895509e-07,
+      "loss": 0.9257,
+      "step": 12907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7966301022430176,
+      "learning_rate": 7.717642306449113e-07,
+      "loss": 0.8251,
+      "step": 12908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.576634942449756,
+      "learning_rate": 7.712111043969772e-07,
+      "loss": 0.8374,
+      "step": 12909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.789953660884544,
+      "learning_rate": 7.706581598695207e-07,
+      "loss": 0.9403,
+      "step": 12910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8193546941579637,
+      "learning_rate": 7.701053970863037e-07,
+      "loss": 0.8242,
+      "step": 12911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.40470212715261,
+      "learning_rate": 7.695528160710764e-07,
+      "loss": 0.9063,
+      "step": 12912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1053478163567034,
+      "learning_rate": 7.690004168475868e-07,
+      "loss": 0.6218,
+      "step": 12913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7061886582102264,
+      "learning_rate": 7.684481994395726e-07,
+      "loss": 0.8237,
+      "step": 12914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.046646568190313,
+      "learning_rate": 7.678961638707633e-07,
+      "loss": 0.8938,
+      "step": 12915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.097589482278179,
+      "learning_rate": 7.673443101648808e-07,
+      "loss": 0.8469,
+      "step": 12916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0926004249498678,
+      "learning_rate": 7.667926383456392e-07,
+      "loss": 0.7553,
+      "step": 12917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9904864847473758,
+      "learning_rate": 7.662411484367482e-07,
+      "loss": 0.6905,
+      "step": 12918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.056799007439221,
+      "learning_rate": 7.656898404619029e-07,
+      "loss": 0.6365,
+      "step": 12919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2179108161716123,
+      "learning_rate": 7.651387144447942e-07,
+      "loss": 0.7669,
+      "step": 12920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.075483767981557,
+      "learning_rate": 7.645877704091082e-07,
+      "loss": 0.6257,
+      "step": 12921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0216484931829322,
+      "learning_rate": 7.640370083785175e-07,
+      "loss": 0.6925,
+      "step": 12922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.170285499751965,
+      "learning_rate": 7.634864283766913e-07,
+      "loss": 0.7641,
+      "step": 12923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2430109180984545,
+      "learning_rate": 7.629360304272882e-07,
+      "loss": 0.8618,
+      "step": 12924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0647769493656827,
+      "learning_rate": 7.623858145539592e-07,
+      "loss": 0.8163,
+      "step": 12925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1852793463435174,
+      "learning_rate": 7.618357807803517e-07,
+      "loss": 0.6746,
+      "step": 12926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5479259534364904,
+      "learning_rate": 7.612859291300973e-07,
+      "loss": 0.8786,
+      "step": 12927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4984640167645975,
+      "learning_rate": 7.607362596268281e-07,
+      "loss": 0.6956,
+      "step": 12928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9722446784919483,
+      "learning_rate": 7.601867722941642e-07,
+      "loss": 0.814,
+      "step": 12929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1751320579896485,
+      "learning_rate": 7.596374671557144e-07,
+      "loss": 0.6691,
+      "step": 12930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2605116482635732,
+      "learning_rate": 7.590883442350883e-07,
+      "loss": 0.6763,
+      "step": 12931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4537330014297734,
+      "learning_rate": 7.585394035558807e-07,
+      "loss": 0.7122,
+      "step": 12932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0139671755060107,
+      "learning_rate": 7.579906451416813e-07,
+      "loss": 0.6989,
+      "step": 12933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.789406891925793,
+      "learning_rate": 7.57442069016071e-07,
+      "loss": 0.7814,
+      "step": 12934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.397640496827254,
+      "learning_rate": 7.568936752026229e-07,
+      "loss": 0.7763,
+      "step": 12935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.346909647787656,
+      "learning_rate": 7.563454637249056e-07,
+      "loss": 0.7504,
+      "step": 12936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0839669917166748,
+      "learning_rate": 7.557974346064728e-07,
+      "loss": 0.5855,
+      "step": 12937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0195275299655915,
+      "learning_rate": 7.552495878708776e-07,
+      "loss": 0.8003,
+      "step": 12938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.107288706092745,
+      "learning_rate": 7.547019235416609e-07,
+      "loss": 0.7532,
+      "step": 12939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5982677838410395,
+      "learning_rate": 7.541544416423574e-07,
+      "loss": 0.6967,
+      "step": 12940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1226075253827306,
+      "learning_rate": 7.536071421964941e-07,
+      "loss": 0.8629,
+      "step": 12941
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.436576313713447,
+      "learning_rate": 7.530600252275888e-07,
+      "loss": 0.8507,
+      "step": 12942
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9141906419233286,
+      "learning_rate": 7.52513090759151e-07,
+      "loss": 0.7702,
+      "step": 12943
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8882713846255488,
+      "learning_rate": 7.519663388146886e-07,
+      "loss": 0.9846,
+      "step": 12944
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.479716344085468,
+      "learning_rate": 7.514197694176905e-07,
+      "loss": 0.6385,
+      "step": 12945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1467515607288803,
+      "learning_rate": 7.50873382591648e-07,
+      "loss": 0.6826,
+      "step": 12946
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.078274859676262,
+      "learning_rate": 7.503271783600402e-07,
+      "loss": 0.6831,
+      "step": 12947
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.670332206872755,
+      "learning_rate": 7.497811567463381e-07,
+      "loss": 0.6536,
+      "step": 12948
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.2663162640936285,
+      "learning_rate": 7.492353177740047e-07,
+      "loss": 0.8611,
+      "step": 12949
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8160276281264458,
+      "learning_rate": 7.486896614664962e-07,
+      "loss": 0.9592,
+      "step": 12950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5827072832224514,
+      "learning_rate": 7.481441878472629e-07,
+      "loss": 0.7955,
+      "step": 12951
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2084188289340245,
+      "learning_rate": 7.475988969397424e-07,
+      "loss": 1.0009,
+      "step": 12952
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4658975867598785,
+      "learning_rate": 7.470537887673667e-07,
+      "loss": 0.9084,
+      "step": 12953
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.849009108612345,
+      "learning_rate": 7.465088633535639e-07,
+      "loss": 0.7171,
+      "step": 12954
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.037940665005681,
+      "learning_rate": 7.459641207217461e-07,
+      "loss": 0.7771,
+      "step": 12955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4997078050742925,
+      "learning_rate": 7.454195608953252e-07,
+      "loss": 0.723,
+      "step": 12956
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6822186165407045,
+      "learning_rate": 7.448751838977014e-07,
+      "loss": 0.635,
+      "step": 12957
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9035016927768813,
+      "learning_rate": 7.443309897522671e-07,
+      "loss": 0.7481,
+      "step": 12958
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0542472153365834,
+      "learning_rate": 7.437869784824086e-07,
+      "loss": 0.8622,
+      "step": 12959
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4028896317767026,
+      "learning_rate": 7.432431501115012e-07,
+      "loss": 0.7123,
+      "step": 12960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1445410655244823,
+      "learning_rate": 7.426995046629176e-07,
+      "loss": 0.7783,
+      "step": 12961
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1695622940472754,
+      "learning_rate": 7.421560421600182e-07,
+      "loss": 0.7576,
+      "step": 12962
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0710560592599427,
+      "learning_rate": 7.416127626261549e-07,
+      "loss": 0.6171,
+      "step": 12963
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8801056646907843,
+      "learning_rate": 7.410696660846761e-07,
+      "loss": 0.9074,
+      "step": 12964
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0895598222892904,
+      "learning_rate": 7.405267525589183e-07,
+      "loss": 0.7396,
+      "step": 12965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.222573438341829,
+      "learning_rate": 7.399840220722127e-07,
+      "loss": 0.6264,
+      "step": 12966
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0472988919133877,
+      "learning_rate": 7.394414746478817e-07,
+      "loss": 0.6917,
+      "step": 12967
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4535407327166348,
+      "learning_rate": 7.388991103092374e-07,
+      "loss": 0.864,
+      "step": 12968
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.317761157500878,
+      "learning_rate": 7.383569290795911e-07,
+      "loss": 0.7653,
+      "step": 12969
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.604858143847363,
+      "learning_rate": 7.378149309822374e-07,
+      "loss": 0.8335,
+      "step": 12970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.164687171326805,
+      "learning_rate": 7.372731160404672e-07,
+      "loss": 0.8075,
+      "step": 12971
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1519363840422665,
+      "learning_rate": 7.367314842775674e-07,
+      "loss": 0.7313,
+      "step": 12972
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1795875690593276,
+      "learning_rate": 7.361900357168084e-07,
+      "loss": 0.6611,
+      "step": 12973
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0518168582637855,
+      "learning_rate": 7.356487703814602e-07,
+      "loss": 0.7858,
+      "step": 12974
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7971567466927412,
+      "learning_rate": 7.351076882947817e-07,
+      "loss": 0.7097,
+      "step": 12975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3207554750351402,
+      "learning_rate": 7.345667894800246e-07,
+      "loss": 0.7968,
+      "step": 12976
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.884359966727753,
+      "learning_rate": 7.340260739604316e-07,
+      "loss": 0.7764,
+      "step": 12977
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8815840095670557,
+      "learning_rate": 7.334855417592385e-07,
+      "loss": 0.8074,
+      "step": 12978
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5554804817206476,
+      "learning_rate": 7.329451928996745e-07,
+      "loss": 0.7372,
+      "step": 12979
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1031399951783083,
+      "learning_rate": 7.324050274049599e-07,
+      "loss": 0.9261,
+      "step": 12980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.619387971927757,
+      "learning_rate": 7.318650452983039e-07,
+      "loss": 0.5437,
+      "step": 12981
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0661629281111218,
+      "learning_rate": 7.313252466029131e-07,
+      "loss": 0.5671,
+      "step": 12982
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4392336872189744,
+      "learning_rate": 7.307856313419837e-07,
+      "loss": 0.6352,
+      "step": 12983
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8438721858248506,
+      "learning_rate": 7.302461995387033e-07,
+      "loss": 0.8342,
+      "step": 12984
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9663299811113548,
+      "learning_rate": 7.297069512162535e-07,
+      "loss": 0.8111,
+      "step": 12985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3819030413872535,
+      "learning_rate": 7.291678863978052e-07,
+      "loss": 0.6426,
+      "step": 12986
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2913697031127955,
+      "learning_rate": 7.286290051065264e-07,
+      "loss": 0.6822,
+      "step": 12987
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.079006935449832,
+      "learning_rate": 7.280903073655704e-07,
+      "loss": 0.9164,
+      "step": 12988
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.998322728593761,
+      "learning_rate": 7.275517931980886e-07,
+      "loss": 0.7418,
+      "step": 12989
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.307563296036029,
+      "learning_rate": 7.270134626272229e-07,
+      "loss": 0.7315,
+      "step": 12990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.042776017696044,
+      "learning_rate": 7.264753156761028e-07,
+      "loss": 0.5304,
+      "step": 12991
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2534319436713255,
+      "learning_rate": 7.25937352367857e-07,
+      "loss": 0.8156,
+      "step": 12992
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5310638788308872,
+      "learning_rate": 7.253995727256024e-07,
+      "loss": 0.8568,
+      "step": 12993
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1311804346665504,
+      "learning_rate": 7.24861976772448e-07,
+      "loss": 0.7366,
+      "step": 12994
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.34926051139654,
+      "learning_rate": 7.243245645314955e-07,
+      "loss": 0.6296,
+      "step": 12995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.469645276479637,
+      "learning_rate": 7.23787336025838e-07,
+      "loss": 0.7759,
+      "step": 12996
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4622734618570687,
+      "learning_rate": 7.232502912785633e-07,
+      "loss": 0.9041,
+      "step": 12997
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.236790239496884,
+      "learning_rate": 7.227134303127487e-07,
+      "loss": 0.7681,
+      "step": 12998
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.286621011327973,
+      "learning_rate": 7.22176753151464e-07,
+      "loss": 0.7696,
+      "step": 12999
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.285269506048492,
+      "learning_rate": 7.216402598177713e-07,
+      "loss": 0.6805,
+      "step": 13000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2808415447391566,
+      "learning_rate": 7.21103950334725e-07,
+      "loss": 0.7875,
+      "step": 13001
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0294398309874135,
+      "learning_rate": 7.205678247253717e-07,
+      "loss": 0.7844,
+      "step": 13002
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7951406421199867,
+      "learning_rate": 7.200318830127495e-07,
+      "loss": 0.7892,
+      "step": 13003
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.137663620435326,
+      "learning_rate": 7.194961252198885e-07,
+      "loss": 0.5858,
+      "step": 13004
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.887841641186075,
+      "learning_rate": 7.189605513698139e-07,
+      "loss": 0.8471,
+      "step": 13005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.225187461191698,
+      "learning_rate": 7.184251614855369e-07,
+      "loss": 0.9407,
+      "step": 13006
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.55470627284321,
+      "learning_rate": 7.178899555900675e-07,
+      "loss": 0.7977,
+      "step": 13007
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.162262104031614,
+      "learning_rate": 7.173549337064035e-07,
+      "loss": 0.7342,
+      "step": 13008
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9955571441727216,
+      "learning_rate": 7.168200958575361e-07,
+      "loss": 0.8554,
+      "step": 13009
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4175497092417917,
+      "learning_rate": 7.162854420664478e-07,
+      "loss": 0.9361,
+      "step": 13010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0668652792139635,
+      "learning_rate": 7.157509723561134e-07,
+      "loss": 0.6516,
+      "step": 13011
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.312430858534827,
+      "learning_rate": 7.152166867495042e-07,
+      "loss": 0.8235,
+      "step": 13012
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1410785326991606,
+      "learning_rate": 7.146825852695749e-07,
+      "loss": 0.5714,
+      "step": 13013
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.331749018991275,
+      "learning_rate": 7.141486679392778e-07,
+      "loss": 0.6341,
+      "step": 13014
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.110198773241593,
+      "learning_rate": 7.136149347815591e-07,
+      "loss": 0.8205,
+      "step": 13015
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5422208702993836,
+      "learning_rate": 7.130813858193525e-07,
+      "loss": 0.8228,
+      "step": 13016
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8665956742211436,
+      "learning_rate": 7.125480210755858e-07,
+      "loss": 0.6479,
+      "step": 13017
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.62487236740899,
+      "learning_rate": 7.120148405731803e-07,
+      "loss": 0.9372,
+      "step": 13018
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7330174138746477,
+      "learning_rate": 7.114818443350463e-07,
+      "loss": 0.8821,
+      "step": 13019
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.034206863880083,
+      "learning_rate": 7.109490323840884e-07,
+      "loss": 0.7527,
+      "step": 13020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.088106247994855,
+      "learning_rate": 7.104164047432022e-07,
+      "loss": 0.6232,
+      "step": 13021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0918100773882604,
+      "learning_rate": 7.098839614352775e-07,
+      "loss": 0.8153,
+      "step": 13022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0770166464554523,
+      "learning_rate": 7.093517024831941e-07,
+      "loss": 0.7384,
+      "step": 13023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1225454816697917,
+      "learning_rate": 7.088196279098225e-07,
+      "loss": 0.712,
+      "step": 13024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.606658097474879,
+      "learning_rate": 7.082877377380293e-07,
+      "loss": 0.581,
+      "step": 13025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1591399089007313,
+      "learning_rate": 7.077560319906696e-07,
+      "loss": 0.8261,
+      "step": 13026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.117563742729296,
+      "learning_rate": 7.072245106905928e-07,
+      "loss": 0.6099,
+      "step": 13027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9398495469147116,
+      "learning_rate": 7.066931738606397e-07,
+      "loss": 0.6187,
+      "step": 13028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2581684156404602,
+      "learning_rate": 7.061620215236415e-07,
+      "loss": 0.7257,
+      "step": 13029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.733258501158793,
+      "learning_rate": 7.056310537024257e-07,
+      "loss": 0.8055,
+      "step": 13030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.129441411303036,
+      "learning_rate": 7.051002704198074e-07,
+      "loss": 0.7722,
+      "step": 13031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0661710795938384,
+      "learning_rate": 7.045696716985939e-07,
+      "loss": 0.8459,
+      "step": 13032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.492825916118557,
+      "learning_rate": 7.040392575615896e-07,
+      "loss": 0.7819,
+      "step": 13033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5176713617800632,
+      "learning_rate": 7.035090280315854e-07,
+      "loss": 0.8644,
+      "step": 13034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5000543892861025,
+      "learning_rate": 7.029789831313677e-07,
+      "loss": 0.7161,
+      "step": 13035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6507711802275544,
+      "learning_rate": 7.024491228837127e-07,
+      "loss": 0.8346,
+      "step": 13036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4005285910238143,
+      "learning_rate": 7.0191944731139e-07,
+      "loss": 0.8648,
+      "step": 13037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1637609227901677,
+      "learning_rate": 7.013899564371613e-07,
+      "loss": 0.7381,
+      "step": 13038
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.192172214387327,
+      "learning_rate": 7.008606502837784e-07,
+      "loss": 0.6807,
+      "step": 13039
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7739812735823606,
+      "learning_rate": 7.003315288739893e-07,
+      "loss": 0.7973,
+      "step": 13040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2913124263251023,
+      "learning_rate": 6.998025922305313e-07,
+      "loss": 0.8691,
+      "step": 13041
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8839291175106354,
+      "learning_rate": 6.992738403761307e-07,
+      "loss": 0.8077,
+      "step": 13042
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7877635937523606,
+      "learning_rate": 6.987452733335126e-07,
+      "loss": 0.7274,
+      "step": 13043
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.284924224625645,
+      "learning_rate": 6.982168911253895e-07,
+      "loss": 0.999,
+      "step": 13044
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.074176662754249,
+      "learning_rate": 6.97688693774467e-07,
+      "loss": 0.6716,
+      "step": 13045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1585822342870409,
+      "learning_rate": 6.971606813034432e-07,
+      "loss": 0.6775,
+      "step": 13046
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0144351913861303,
+      "learning_rate": 6.96632853735007e-07,
+      "loss": 0.7452,
+      "step": 13047
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.016091180269757,
+      "learning_rate": 6.961052110918432e-07,
+      "loss": 0.8451,
+      "step": 13048
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6871980859913955,
+      "learning_rate": 6.955777533966212e-07,
+      "loss": 0.7018,
+      "step": 13049
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.990391290312797,
+      "learning_rate": 6.950504806720115e-07,
+      "loss": 0.6713,
+      "step": 13050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.856994298851848,
+      "learning_rate": 6.945233929406697e-07,
+      "loss": 0.7164,
+      "step": 13051
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0085597021754054,
+      "learning_rate": 6.939964902252466e-07,
+      "loss": 0.6108,
+      "step": 13052
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.650994730173584,
+      "learning_rate": 6.934697725483846e-07,
+      "loss": 0.783,
+      "step": 13053
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2280170688869205,
+      "learning_rate": 6.929432399327174e-07,
+      "loss": 0.6899,
+      "step": 13054
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3683496686478516,
+      "learning_rate": 6.924168924008712e-07,
+      "loss": 1.0606,
+      "step": 13055
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8392636321848306,
+      "learning_rate": 6.918907299754651e-07,
+      "loss": 0.8092,
+      "step": 13056
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2009100975695894,
+      "learning_rate": 6.913647526791078e-07,
+      "loss": 0.5793,
+      "step": 13057
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.159432836100693,
+      "learning_rate": 6.90838960534404e-07,
+      "loss": 0.8746,
+      "step": 13058
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.013324106013998,
+      "learning_rate": 6.903133535639467e-07,
+      "loss": 0.7868,
+      "step": 13059
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7960244782375672,
+      "learning_rate": 6.897879317903233e-07,
+      "loss": 0.9556,
+      "step": 13060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1389546482940176,
+      "learning_rate": 6.89262695236112e-07,
+      "loss": 0.7361,
+      "step": 13061
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7717701676352555,
+      "learning_rate": 6.887376439238813e-07,
+      "loss": 0.7709,
+      "step": 13062
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.460219477591642,
+      "learning_rate": 6.882127778761987e-07,
+      "loss": 0.8678,
+      "step": 13063
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.575428491410733,
+      "learning_rate": 6.876880971156147e-07,
+      "loss": 0.8185,
+      "step": 13064
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9006022443682744,
+      "learning_rate": 6.871636016646754e-07,
+      "loss": 0.6898,
+      "step": 13065
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3686332436475936,
+      "learning_rate": 6.866392915459241e-07,
+      "loss": 0.8242,
+      "step": 13066
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3708118950443637,
+      "learning_rate": 6.861151667818866e-07,
+      "loss": 0.7958,
+      "step": 13067
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7618486805071063,
+      "learning_rate": 6.855912273950893e-07,
+      "loss": 0.5849,
+      "step": 13068
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4347018907947775,
+      "learning_rate": 6.850674734080454e-07,
+      "loss": 0.7981,
+      "step": 13069
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6301004460948767,
+      "learning_rate": 6.845439048432617e-07,
+      "loss": 0.6972,
+      "step": 13070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.255659188398972,
+      "learning_rate": 6.840205217232376e-07,
+      "loss": 0.7781,
+      "step": 13071
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.689576370986861,
+      "learning_rate": 6.834973240704629e-07,
+      "loss": 0.8771,
+      "step": 13072
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2504253036461774,
+      "learning_rate": 6.829743119074239e-07,
+      "loss": 0.8575,
+      "step": 13073
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1843895650204725,
+      "learning_rate": 6.824514852565922e-07,
+      "loss": 0.6705,
+      "step": 13074
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6510182102072113,
+      "learning_rate": 6.819288441404343e-07,
+      "loss": 0.8293,
+      "step": 13075
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1903747999038043,
+      "learning_rate": 6.814063885814127e-07,
+      "loss": 0.7338,
+      "step": 13076
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3272189959689644,
+      "learning_rate": 6.80884118601976e-07,
+      "loss": 0.7953,
+      "step": 13077
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9870603253753805,
+      "learning_rate": 6.803620342245676e-07,
+      "loss": 0.7184,
+      "step": 13078
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.47648093791232,
+      "learning_rate": 6.798401354716233e-07,
+      "loss": 0.7451,
+      "step": 13079
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.410943564080077,
+      "learning_rate": 6.793184223655691e-07,
+      "loss": 0.7889,
+      "step": 13080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2500895459149497,
+      "learning_rate": 6.78796894928827e-07,
+      "loss": 0.6045,
+      "step": 13081
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.228266365827384,
+      "learning_rate": 6.782755531838047e-07,
+      "loss": 0.8615,
+      "step": 13082
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2412828517086436,
+      "learning_rate": 6.77754397152906e-07,
+      "loss": 0.7603,
+      "step": 13083
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9994349946897119,
+      "learning_rate": 6.772334268585296e-07,
+      "loss": 0.6815,
+      "step": 13084
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1374084623063492,
+      "learning_rate": 6.767126423230574e-07,
+      "loss": 0.7067,
+      "step": 13085
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2093815178791094,
+      "learning_rate": 6.76192043568873e-07,
+      "loss": 0.8325,
+      "step": 13086
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.937531088961867,
+      "learning_rate": 6.756716306183459e-07,
+      "loss": 0.8403,
+      "step": 13087
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5762500068521126,
+      "learning_rate": 6.751514034938395e-07,
+      "loss": 0.7965,
+      "step": 13088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.354947533331709,
+      "learning_rate": 6.746313622177097e-07,
+      "loss": 0.6357,
+      "step": 13089
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0641428553789205,
+      "learning_rate": 6.741115068123017e-07,
+      "loss": 0.5395,
+      "step": 13090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9629263379923394,
+      "learning_rate": 6.73591837299959e-07,
+      "loss": 0.6417,
+      "step": 13091
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.036446848530865,
+      "learning_rate": 6.730723537030092e-07,
+      "loss": 0.7355,
+      "step": 13092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.57226327384957,
+      "learning_rate": 6.725530560437759e-07,
+      "loss": 0.9809,
+      "step": 13093
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7267681418545386,
+      "learning_rate": 6.720339443445772e-07,
+      "loss": 0.7845,
+      "step": 13094
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2089942223757313,
+      "learning_rate": 6.71515018627718e-07,
+      "loss": 0.9831,
+      "step": 13095
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.012416542004465,
+      "learning_rate": 6.709962789154989e-07,
+      "loss": 0.8367,
+      "step": 13096
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.453616803150407,
+      "learning_rate": 6.704777252302108e-07,
+      "loss": 0.946,
+      "step": 13097
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7818328187904573,
+      "learning_rate": 6.699593575941365e-07,
+      "loss": 0.7173,
+      "step": 13098
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.380816475070019,
+      "learning_rate": 6.694411760295538e-07,
+      "loss": 0.7612,
+      "step": 13099
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0758900772527547,
+      "learning_rate": 6.689231805587265e-07,
+      "loss": 0.6551,
+      "step": 13100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2987026078461756,
+      "learning_rate": 6.684053712039168e-07,
+      "loss": 0.7988,
+      "step": 13101
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3280063067094647,
+      "learning_rate": 6.678877479873769e-07,
+      "loss": 0.8565,
+      "step": 13102
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1233971735393036,
+      "learning_rate": 6.673703109313456e-07,
+      "loss": 0.943,
+      "step": 13103
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8141866660146237,
+      "learning_rate": 6.66853060058063e-07,
+      "loss": 0.8281,
+      "step": 13104
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8665270627503323,
+      "learning_rate": 6.663359953897547e-07,
+      "loss": 0.799,
+      "step": 13105
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0872434539638824,
+      "learning_rate": 6.658191169486405e-07,
+      "loss": 0.9289,
+      "step": 13106
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2609551799474903,
+      "learning_rate": 6.653024247569312e-07,
+      "loss": 0.7574,
+      "step": 13107
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.680466956001853,
+      "learning_rate": 6.647859188368294e-07,
+      "loss": 0.8571,
+      "step": 13108
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.672328232811073,
+      "learning_rate": 6.642695992105347e-07,
+      "loss": 0.756,
+      "step": 13109
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2511698925114962,
+      "learning_rate": 6.637534659002287e-07,
+      "loss": 0.6866,
+      "step": 13110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1383360619143996,
+      "learning_rate": 6.632375189280948e-07,
+      "loss": 0.9323,
+      "step": 13111
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1746056815581827,
+      "learning_rate": 6.627217583163031e-07,
+      "loss": 0.6145,
+      "step": 13112
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.022499492126583,
+      "learning_rate": 6.622061840870169e-07,
+      "loss": 0.8405,
+      "step": 13113
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.872322084967792,
+      "learning_rate": 6.61690796262392e-07,
+      "loss": 0.6748,
+      "step": 13114
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.015789049937498,
+      "learning_rate": 6.61175594864576e-07,
+      "loss": 0.7732,
+      "step": 13115
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.652865748290971,
+      "learning_rate": 6.606605799157062e-07,
+      "loss": 0.7095,
+      "step": 13116
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.494571826751658,
+      "learning_rate": 6.601457514379173e-07,
+      "loss": 0.7602,
+      "step": 13117
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.971516325792218,
+      "learning_rate": 6.596311094533292e-07,
+      "loss": 0.7627,
+      "step": 13118
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5657512335104045,
+      "learning_rate": 6.591166539840599e-07,
+      "loss": 0.9745,
+      "step": 13119
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3888292449672743,
+      "learning_rate": 6.586023850522156e-07,
+      "loss": 1.0932,
+      "step": 13120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4462957545358033,
+      "learning_rate": 6.580883026798957e-07,
+      "loss": 0.8249,
+      "step": 13121
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.415913776520867,
+      "learning_rate": 6.575744068891909e-07,
+      "loss": 0.7563,
+      "step": 13122
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.586307252181426,
+      "learning_rate": 6.570606977021842e-07,
+      "loss": 0.6631,
+      "step": 13123
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1314161236361007,
+      "learning_rate": 6.565471751409541e-07,
+      "loss": 0.8263,
+      "step": 13124
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0815428625718853,
+      "learning_rate": 6.56033839227564e-07,
+      "loss": 0.7854,
+      "step": 13125
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.332076725838633,
+      "learning_rate": 6.555206899840727e-07,
+      "loss": 0.9452,
+      "step": 13126
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0572402189334273,
+      "learning_rate": 6.550077274325345e-07,
+      "loss": 0.6802,
+      "step": 13127
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7990147617148504,
+      "learning_rate": 6.544949515949911e-07,
+      "loss": 0.7302,
+      "step": 13128
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.854554039539272,
+      "learning_rate": 6.539823624934777e-07,
+      "loss": 0.7879,
+      "step": 13129
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.450539423686322,
+      "learning_rate": 6.53469960150021e-07,
+      "loss": 0.8931,
+      "step": 13130
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.657443624147589,
+      "learning_rate": 6.529577445866409e-07,
+      "loss": 0.8179,
+      "step": 13131
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.210298322463161,
+      "learning_rate": 6.524457158253472e-07,
+      "loss": 0.5729,
+      "step": 13132
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.084106436469271,
+      "learning_rate": 6.519338738881442e-07,
+      "loss": 0.7637,
+      "step": 13133
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.015074907657567,
+      "learning_rate": 6.514222187970248e-07,
+      "loss": 0.9415,
+      "step": 13134
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2816929255407508,
+      "learning_rate": 6.509107505739798e-07,
+      "loss": 0.8406,
+      "step": 13135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4230799696099363,
+      "learning_rate": 6.503994692409838e-07,
+      "loss": 0.7181,
+      "step": 13136
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1363451904415087,
+      "learning_rate": 6.498883748200102e-07,
+      "loss": 0.8742,
+      "step": 13137
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6543223606312707,
+      "learning_rate": 6.493774673330217e-07,
+      "loss": 0.6084,
+      "step": 13138
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.691569294723058,
+      "learning_rate": 6.488667468019727e-07,
+      "loss": 0.7803,
+      "step": 13139
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3991057019135136,
+      "learning_rate": 6.483562132488103e-07,
+      "loss": 0.791,
+      "step": 13140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9990966609095913,
+      "learning_rate": 6.478458666954718e-07,
+      "loss": 0.7598,
+      "step": 13141
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2306449628235225,
+      "learning_rate": 6.473357071638914e-07,
+      "loss": 0.8176,
+      "step": 13142
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.6702809772393215,
+      "learning_rate": 6.468257346759888e-07,
+      "loss": 0.842,
+      "step": 13143
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9256309228967239,
+      "learning_rate": 6.46315949253678e-07,
+      "loss": 0.7315,
+      "step": 13144
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.112787562987841,
+      "learning_rate": 6.458063509188678e-07,
+      "loss": 0.6511,
+      "step": 13145
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3912606392502007,
+      "learning_rate": 6.452969396934567e-07,
+      "loss": 0.7117,
+      "step": 13146
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.017046976988462,
+      "learning_rate": 6.447877155993343e-07,
+      "loss": 0.766,
+      "step": 13147
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3411161016901123,
+      "learning_rate": 6.442786786583832e-07,
+      "loss": 0.7993,
+      "step": 13148
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0650129502650794,
+      "learning_rate": 6.437698288924777e-07,
+      "loss": 0.7943,
+      "step": 13149
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.118397216200135,
+      "learning_rate": 6.432611663234851e-07,
+      "loss": 0.7704,
+      "step": 13150
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.154025723929018,
+      "learning_rate": 6.427526909732617e-07,
+      "loss": 0.6461,
+      "step": 13151
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.757150200770983,
+      "learning_rate": 6.422444028636604e-07,
+      "loss": 0.7551,
+      "step": 13152
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3272699797149183,
+      "learning_rate": 6.417363020165235e-07,
+      "loss": 0.622,
+      "step": 13153
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.376580927187344,
+      "learning_rate": 6.412283884536818e-07,
+      "loss": 0.7956,
+      "step": 13154
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9829320175320333,
+      "learning_rate": 6.407206621969648e-07,
+      "loss": 0.6864,
+      "step": 13155
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.286947786552048,
+      "learning_rate": 6.402131232681897e-07,
+      "loss": 0.8038,
+      "step": 13156
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.33001702033553,
+      "learning_rate": 6.397057716891669e-07,
+      "loss": 0.7011,
+      "step": 13157
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9563371043398011,
+      "learning_rate": 6.391986074816975e-07,
+      "loss": 0.6378,
+      "step": 13158
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3366922143068845,
+      "learning_rate": 6.38691630667575e-07,
+      "loss": 0.8753,
+      "step": 13159
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9380214961198474,
+      "learning_rate": 6.381848412685882e-07,
+      "loss": 0.8173,
+      "step": 13160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.254514306078125,
+      "learning_rate": 6.376782393065111e-07,
+      "loss": 0.7157,
+      "step": 13161
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3327178730504365,
+      "learning_rate": 6.371718248031167e-07,
+      "loss": 0.7653,
+      "step": 13162
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.814909523474691,
+      "learning_rate": 6.36665597780165e-07,
+      "loss": 0.8354,
+      "step": 13163
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.102319742317624,
+      "learning_rate": 6.36159558259411e-07,
+      "loss": 1.0993,
+      "step": 13164
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0183541152629716,
+      "learning_rate": 6.35653706262599e-07,
+      "loss": 0.7241,
+      "step": 13165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.798062449177311,
+      "learning_rate": 6.351480418114675e-07,
+      "loss": 0.8823,
+      "step": 13166
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.292684304751193,
+      "learning_rate": 6.346425649277454e-07,
+      "loss": 0.7547,
+      "step": 13167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9701153858664398,
+      "learning_rate": 6.341372756331543e-07,
+      "loss": 0.7578,
+      "step": 13168
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3377487466884763,
+      "learning_rate": 6.336321739494072e-07,
+      "loss": 0.8781,
+      "step": 13169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0247356909831473,
+      "learning_rate": 6.331272598982108e-07,
+      "loss": 0.7106,
+      "step": 13170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6410421579617687,
+      "learning_rate": 6.326225335012615e-07,
+      "loss": 0.7034,
+      "step": 13171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.284815374683778,
+      "learning_rate": 6.321179947802492e-07,
+      "loss": 0.6716,
+      "step": 13172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0631757213183852,
+      "learning_rate": 6.316136437568538e-07,
+      "loss": 0.8168,
+      "step": 13173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.372739230846321,
+      "learning_rate": 6.31109480452749e-07,
+      "loss": 0.8257,
+      "step": 13174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.040404125906226,
+      "learning_rate": 6.306055048896004e-07,
+      "loss": 0.7858,
+      "step": 13175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.595099684820404,
+      "learning_rate": 6.301017170890639e-07,
+      "loss": 0.6719,
+      "step": 13176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7179300069976569,
+      "learning_rate": 6.295981170727883e-07,
+      "loss": 0.7456,
+      "step": 13177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3483709991151547,
+      "learning_rate": 6.290947048624169e-07,
+      "loss": 0.6865,
+      "step": 13178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9571701953087355,
+      "learning_rate": 6.285914804795784e-07,
+      "loss": 0.8064,
+      "step": 13179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2023931468185232,
+      "learning_rate": 6.280884439459006e-07,
+      "loss": 0.6995,
+      "step": 13180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6045317200205584,
+      "learning_rate": 6.275855952829995e-07,
+      "loss": 0.6782,
+      "step": 13181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4093141122499238,
+      "learning_rate": 6.270829345124829e-07,
+      "loss": 0.7955,
+      "step": 13182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.713871822581323,
+      "learning_rate": 6.265804616559518e-07,
+      "loss": 0.747,
+      "step": 13183
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.165858076532707,
+      "learning_rate": 6.260781767349983e-07,
+      "loss": 0.8947,
+      "step": 13184
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2640319965986326,
+      "learning_rate": 6.255760797712068e-07,
+      "loss": 0.861,
+      "step": 13185
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.36686162837186,
+      "learning_rate": 6.250741707861535e-07,
+      "loss": 0.9126,
+      "step": 13186
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.105641777283503,
+      "learning_rate": 6.245724498014055e-07,
+      "loss": 0.8305,
+      "step": 13187
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7417477874342915,
+      "learning_rate": 6.240709168385251e-07,
+      "loss": 0.6517,
+      "step": 13188
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.158187512639674,
+      "learning_rate": 6.235695719190632e-07,
+      "loss": 0.7866,
+      "step": 13189
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1055820240103513,
+      "learning_rate": 6.230684150645633e-07,
+      "loss": 0.6727,
+      "step": 13190
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.087064029830792,
+      "learning_rate": 6.22567446296562e-07,
+      "loss": 0.7818,
+      "step": 13191
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.371422997175664,
+      "learning_rate": 6.220666656365865e-07,
+      "loss": 0.8449,
+      "step": 13192
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.20578809292672,
+      "learning_rate": 6.215660731061563e-07,
+      "loss": 0.7061,
+      "step": 13193
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8841347267554363,
+      "learning_rate": 6.210656687267835e-07,
+      "loss": 0.7861,
+      "step": 13194
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3705774111550513,
+      "learning_rate": 6.2056545251997e-07,
+      "loss": 0.8092,
+      "step": 13195
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2835382669044804,
+      "learning_rate": 6.200654245072146e-07,
+      "loss": 0.7449,
+      "step": 13196
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3882455636626199,
+      "learning_rate": 6.195655847100008e-07,
+      "loss": 0.5691,
+      "step": 13197
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9280227843183256,
+      "learning_rate": 6.1906593314981e-07,
+      "loss": 0.6186,
+      "step": 13198
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1597222111835044,
+      "learning_rate": 6.185664698481137e-07,
+      "loss": 0.9473,
+      "step": 13199
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0801002418957144,
+      "learning_rate": 6.180671948263739e-07,
+      "loss": 0.7142,
+      "step": 13200
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.21862564525794,
+      "learning_rate": 6.175681081060453e-07,
+      "loss": 0.7191,
+      "step": 13201
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.247160036387803,
+      "learning_rate": 6.170692097085751e-07,
+      "loss": 0.7718,
+      "step": 13202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9193566710558772,
+      "learning_rate": 6.165704996554039e-07,
+      "loss": 0.7958,
+      "step": 13203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0722531524983077,
+      "learning_rate": 6.160719779679597e-07,
+      "loss": 0.6222,
+      "step": 13204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.41610503859043,
+      "learning_rate": 6.155736446676652e-07,
+      "loss": 0.7166,
+      "step": 13205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0911817123539616,
+      "learning_rate": 6.150754997759367e-07,
+      "loss": 0.6621,
+      "step": 13206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.218582078496116,
+      "learning_rate": 6.145775433141798e-07,
+      "loss": 0.8307,
+      "step": 13207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.532882561491495,
+      "learning_rate": 6.140797753037925e-07,
+      "loss": 0.7992,
+      "step": 13208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5684085443978493,
+      "learning_rate": 6.135821957661658e-07,
+      "loss": 0.6551,
+      "step": 13209
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0769545572323298,
+      "learning_rate": 6.130848047226806e-07,
+      "loss": 0.6997,
+      "step": 13210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1243117100773254,
+      "learning_rate": 6.125876021947119e-07,
+      "loss": 0.7388,
+      "step": 13211
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.95833615937326,
+      "learning_rate": 6.120905882036244e-07,
+      "loss": 0.8224,
+      "step": 13212
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1912278573497366,
+      "learning_rate": 6.115937627707775e-07,
+      "loss": 0.7971,
+      "step": 13213
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.138566104973175,
+      "learning_rate": 6.110971259175208e-07,
+      "loss": 0.7478,
+      "step": 13214
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.40320516991673,
+      "learning_rate": 6.106006776651934e-07,
+      "loss": 0.8335,
+      "step": 13215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4602598818357797,
+      "learning_rate": 6.101044180351318e-07,
+      "loss": 0.7955,
+      "step": 13216
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5890486244566002,
+      "learning_rate": 6.096083470486602e-07,
+      "loss": 0.7745,
+      "step": 13217
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6691587985274274,
+      "learning_rate": 6.091124647270958e-07,
+      "loss": 0.8835,
+      "step": 13218
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7538855544105814,
+      "learning_rate": 6.086167710917479e-07,
+      "loss": 0.7853,
+      "step": 13219
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7245546095401962,
+      "learning_rate": 6.081212661639163e-07,
+      "loss": 0.7664,
+      "step": 13220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1851160858050642,
+      "learning_rate": 6.076259499648968e-07,
+      "loss": 0.8369,
+      "step": 13221
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.179422427853114,
+      "learning_rate": 6.071308225159712e-07,
+      "loss": 0.7283,
+      "step": 13222
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.421318007507849,
+      "learning_rate": 6.066358838384184e-07,
+      "loss": 0.8357,
+      "step": 13223
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2632492462133156,
+      "learning_rate": 6.061411339535062e-07,
+      "loss": 0.7442,
+      "step": 13224
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1060776777180426,
+      "learning_rate": 6.056465728824951e-07,
+      "loss": 0.9648,
+      "step": 13225
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9666606749062976,
+      "learning_rate": 6.051522006466376e-07,
+      "loss": 0.8828,
+      "step": 13226
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.668019748091422,
+      "learning_rate": 6.046580172671773e-07,
+      "loss": 0.8337,
+      "step": 13227
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1436465380126646,
+      "learning_rate": 6.041640227653517e-07,
+      "loss": 0.703,
+      "step": 13228
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.261746229747446,
+      "learning_rate": 6.036702171623876e-07,
+      "loss": 0.7143,
+      "step": 13229
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7718448926185735,
+      "learning_rate": 6.031766004795047e-07,
+      "loss": 0.8051,
+      "step": 13230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8454001458766274,
+      "learning_rate": 6.026831727379168e-07,
+      "loss": 0.8007,
+      "step": 13231
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.232044976662782,
+      "learning_rate": 6.021899339588266e-07,
+      "loss": 0.9418,
+      "step": 13232
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.079068248733983,
+      "learning_rate": 6.016968841634279e-07,
+      "loss": 0.9205,
+      "step": 13233
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.52068840379966,
+      "learning_rate": 6.012040233729105e-07,
+      "loss": 0.5423,
+      "step": 13234
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0086160882568649,
+      "learning_rate": 6.007113516084528e-07,
+      "loss": 0.6449,
+      "step": 13235
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2062570391223484,
+      "learning_rate": 6.002188688912264e-07,
+      "loss": 0.796,
+      "step": 13236
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.303615123353396,
+      "learning_rate": 5.997265752423936e-07,
+      "loss": 0.7263,
+      "step": 13237
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9267483787650097,
+      "learning_rate": 5.992344706831088e-07,
+      "loss": 0.7235,
+      "step": 13238
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0588254359793434,
+      "learning_rate": 5.987425552345222e-07,
+      "loss": 0.7353,
+      "step": 13239
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6100307359337496,
+      "learning_rate": 5.98250828917768e-07,
+      "loss": 0.9259,
+      "step": 13240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0680418363569797,
+      "learning_rate": 5.977592917539799e-07,
+      "loss": 0.6897,
+      "step": 13241
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.169995075236578,
+      "learning_rate": 5.972679437642793e-07,
+      "loss": 0.7404,
+      "step": 13242
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.97642179269,
+      "learning_rate": 5.967767849697803e-07,
+      "loss": 0.7562,
+      "step": 13243
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5653639693921324,
+      "learning_rate": 5.962858153915896e-07,
+      "loss": 0.8204,
+      "step": 13244
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9902945313362836,
+      "learning_rate": 5.957950350508052e-07,
+      "loss": 0.77,
+      "step": 13245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2997735833404533,
+      "learning_rate": 5.953044439685157e-07,
+      "loss": 0.5417,
+      "step": 13246
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.726638867179351,
+      "learning_rate": 5.948140421658061e-07,
+      "loss": 0.8833,
+      "step": 13247
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2352016220335043,
+      "learning_rate": 5.943238296637455e-07,
+      "loss": 0.8968,
+      "step": 13248
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.696533151013981,
+      "learning_rate": 5.938338064834037e-07,
+      "loss": 0.6421,
+      "step": 13249
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8383707248127403,
+      "learning_rate": 5.933439726458357e-07,
+      "loss": 0.9836,
+      "step": 13250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.598990008167074,
+      "learning_rate": 5.928543281720917e-07,
+      "loss": 0.7729,
+      "step": 13251
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.593828701454406,
+      "learning_rate": 5.923648730832121e-07,
+      "loss": 0.8422,
+      "step": 13252
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8517614613663942,
+      "learning_rate": 5.918756074002291e-07,
+      "loss": 0.8178,
+      "step": 13253
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2031825007898367,
+      "learning_rate": 5.913865311441714e-07,
+      "loss": 0.7784,
+      "step": 13254
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8683945961192954,
+      "learning_rate": 5.908976443360515e-07,
+      "loss": 0.7187,
+      "step": 13255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8215171996133002,
+      "learning_rate": 5.904089469968782e-07,
+      "loss": 0.8877,
+      "step": 13256
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.686639821205569,
+      "learning_rate": 5.899204391476554e-07,
+      "loss": 0.8591,
+      "step": 13257
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4055342614695663,
+      "learning_rate": 5.894321208093712e-07,
+      "loss": 0.9479,
+      "step": 13258
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3997919570636044,
+      "learning_rate": 5.889439920030127e-07,
+      "loss": 0.9531,
+      "step": 13259
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0342931535768836,
+      "learning_rate": 5.884560527495542e-07,
+      "loss": 0.9582,
+      "step": 13260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.153409891003474,
+      "learning_rate": 5.879683030699645e-07,
+      "loss": 0.7708,
+      "step": 13261
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9724545849751352,
+      "learning_rate": 5.874807429852031e-07,
+      "loss": 0.7278,
+      "step": 13262
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.524506487675849,
+      "learning_rate": 5.869933725162203e-07,
+      "loss": 0.7864,
+      "step": 13263
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1042903421537176,
+      "learning_rate": 5.865061916839615e-07,
+      "loss": 0.643,
+      "step": 13264
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0757544366937326,
+      "learning_rate": 5.860192005093624e-07,
+      "loss": 0.6637,
+      "step": 13265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.614032641507949,
+      "learning_rate": 5.855323990133465e-07,
+      "loss": 0.7982,
+      "step": 13266
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.023592914775755,
+      "learning_rate": 5.850457872168358e-07,
+      "loss": 0.837,
+      "step": 13267
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3194554942539014,
+      "learning_rate": 5.845593651407405e-07,
+      "loss": 0.7041,
+      "step": 13268
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9598060473761192,
+      "learning_rate": 5.840731328059629e-07,
+      "loss": 0.8399,
+      "step": 13269
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.035699906198908,
+      "learning_rate": 5.835870902333974e-07,
+      "loss": 0.6669,
+      "step": 13270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.448504638898995,
+      "learning_rate": 5.831012374439293e-07,
+      "loss": 0.8445,
+      "step": 13271
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8873292169533227,
+      "learning_rate": 5.826155744584405e-07,
+      "loss": 0.8285,
+      "step": 13272
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.046720277338231,
+      "learning_rate": 5.821301012977965e-07,
+      "loss": 0.7259,
+      "step": 13273
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.507834605856472,
+      "learning_rate": 5.816448179828616e-07,
+      "loss": 0.6187,
+      "step": 13274
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.306374613166635,
+      "learning_rate": 5.811597245344908e-07,
+      "loss": 0.8849,
+      "step": 13275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4191741615706936,
+      "learning_rate": 5.806748209735252e-07,
+      "loss": 0.6715,
+      "step": 13276
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9170347201097753,
+      "learning_rate": 5.801901073208066e-07,
+      "loss": 0.6353,
+      "step": 13277
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9590459436095302,
+      "learning_rate": 5.797055835971621e-07,
+      "loss": 0.7604,
+      "step": 13278
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.319333793550608,
+      "learning_rate": 5.792212498234134e-07,
+      "loss": 0.9189,
+      "step": 13279
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9641901946784075,
+      "learning_rate": 5.787371060203728e-07,
+      "loss": 0.7497,
+      "step": 13280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1668296098155608,
+      "learning_rate": 5.782531522088452e-07,
+      "loss": 0.6522,
+      "step": 13281
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6608830620794315,
+      "learning_rate": 5.777693884096275e-07,
+      "loss": 0.7354,
+      "step": 13282
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.998982561077007,
+      "learning_rate": 5.772858146435095e-07,
+      "loss": 0.6699,
+      "step": 13283
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2875589410070276,
+      "learning_rate": 5.768024309312681e-07,
+      "loss": 0.7162,
+      "step": 13284
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9311275926021159,
+      "learning_rate": 5.763192372936782e-07,
+      "loss": 0.8356,
+      "step": 13285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1111516075752186,
+      "learning_rate": 5.758362337515028e-07,
+      "loss": 0.7332,
+      "step": 13286
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0474072883297496,
+      "learning_rate": 5.753534203254973e-07,
+      "loss": 0.6336,
+      "step": 13287
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8676201535964427,
+      "learning_rate": 5.748707970364092e-07,
+      "loss": 0.9036,
+      "step": 13288
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6470189454435022,
+      "learning_rate": 5.74388363904978e-07,
+      "loss": 0.9169,
+      "step": 13289
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3058757221968404,
+      "learning_rate": 5.73906120951937e-07,
+      "loss": 0.8306,
+      "step": 13290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.191816371654996,
+      "learning_rate": 5.734240681980052e-07,
+      "loss": 0.7346,
+      "step": 13291
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1236758519364747,
+      "learning_rate": 5.729422056639006e-07,
+      "loss": 0.7756,
+      "step": 13292
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.433677131740416,
+      "learning_rate": 5.724605333703303e-07,
+      "loss": 1.0315,
+      "step": 13293
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.388160174083008,
+      "learning_rate": 5.719790513379891e-07,
+      "loss": 0.6428,
+      "step": 13294
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.900342476436824,
+      "learning_rate": 5.714977595875715e-07,
+      "loss": 0.7503,
+      "step": 13295
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.221363007208128,
+      "learning_rate": 5.710166581397575e-07,
+      "loss": 0.824,
+      "step": 13296
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.243122671764095,
+      "learning_rate": 5.705357470152218e-07,
+      "loss": 0.7919,
+      "step": 13297
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0812134478137339,
+      "learning_rate": 5.700550262346294e-07,
+      "loss": 0.61,
+      "step": 13298
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9615552273604215,
+      "learning_rate": 5.695744958186383e-07,
+      "loss": 0.7177,
+      "step": 13299
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0726431728764285,
+      "learning_rate": 5.690941557878988e-07,
+      "loss": 0.8247,
+      "step": 13300
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.307960693736726,
+      "learning_rate": 5.686140061630519e-07,
+      "loss": 0.8123,
+      "step": 13301
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3696300321298227,
+      "learning_rate": 5.6813404696473e-07,
+      "loss": 0.7685,
+      "step": 13302
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.11941670631463,
+      "learning_rate": 5.676542782135585e-07,
+      "loss": 0.6088,
+      "step": 13303
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2628753715045917,
+      "learning_rate": 5.671746999301542e-07,
+      "loss": 0.7688,
+      "step": 13304
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.157424853099777,
+      "learning_rate": 5.666953121351254e-07,
+      "loss": 0.8815,
+      "step": 13305
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2323421162085526,
+      "learning_rate": 5.662161148490725e-07,
+      "loss": 0.7228,
+      "step": 13306
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8410255286152606,
+      "learning_rate": 5.657371080925866e-07,
+      "loss": 0.6921,
+      "step": 13307
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.83695153548724,
+      "learning_rate": 5.652582918862548e-07,
+      "loss": 0.8414,
+      "step": 13308
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5831547781694186,
+      "learning_rate": 5.647796662506493e-07,
+      "loss": 0.9609,
+      "step": 13309
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9295463696366957,
+      "learning_rate": 5.6430123120634e-07,
+      "loss": 0.8894,
+      "step": 13310
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2322828238433052,
+      "learning_rate": 5.638229867738854e-07,
+      "loss": 0.7027,
+      "step": 13311
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.878544013770195,
+      "learning_rate": 5.633449329738372e-07,
+      "loss": 0.696,
+      "step": 13312
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.219098488497359,
+      "learning_rate": 5.628670698267386e-07,
+      "loss": 0.8116,
+      "step": 13313
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5217468754930783,
+      "learning_rate": 5.623893973531225e-07,
+      "loss": 0.8929,
+      "step": 13314
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0194117133563045,
+      "learning_rate": 5.619119155735192e-07,
+      "loss": 0.7455,
+      "step": 13315
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.187419016244252,
+      "learning_rate": 5.61434624508444e-07,
+      "loss": 0.8285,
+      "step": 13316
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4900150688064167,
+      "learning_rate": 5.609575241784071e-07,
+      "loss": 0.7141,
+      "step": 13317
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.992581853863396,
+      "learning_rate": 5.604806146039126e-07,
+      "loss": 0.7788,
+      "step": 13318
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.744395645360005,
+      "learning_rate": 5.600038958054538e-07,
+      "loss": 0.6743,
+      "step": 13319
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1732039841155646,
+      "learning_rate": 5.595273678035156e-07,
+      "loss": 0.7325,
+      "step": 13320
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.093117779327854,
+      "learning_rate": 5.590510306185765e-07,
+      "loss": 0.7736,
+      "step": 13321
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7703916548415017,
+      "learning_rate": 5.585748842711047e-07,
+      "loss": 0.8089,
+      "step": 13322
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9118502191300526,
+      "learning_rate": 5.580989287815619e-07,
+      "loss": 0.7714,
+      "step": 13323
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3968584536529978,
+      "learning_rate": 5.576231641703994e-07,
+      "loss": 0.7709,
+      "step": 13324
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2125837747881705,
+      "learning_rate": 5.571475904580648e-07,
+      "loss": 0.9109,
+      "step": 13325
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7652297998828488,
+      "learning_rate": 5.566722076649933e-07,
+      "loss": 0.7595,
+      "step": 13326
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5365409851488425,
+      "learning_rate": 5.561970158116114e-07,
+      "loss": 0.861,
+      "step": 13327
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9322527636302325,
+      "learning_rate": 5.557220149183412e-07,
+      "loss": 0.6585,
+      "step": 13328
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1297345413869793,
+      "learning_rate": 5.552472050055946e-07,
+      "loss": 0.8096,
+      "step": 13329
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6733727875257256,
+      "learning_rate": 5.547725860937741e-07,
+      "loss": 0.7085,
+      "step": 13330
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.287432862035517,
+      "learning_rate": 5.542981582032759e-07,
+      "loss": 0.8421,
+      "step": 13331
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1897889500177672,
+      "learning_rate": 5.538239213544855e-07,
+      "loss": 0.7949,
+      "step": 13332
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.744068793924442,
+      "learning_rate": 5.533498755677857e-07,
+      "loss": 0.7854,
+      "step": 13333
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1493117021806127,
+      "learning_rate": 5.528760208635436e-07,
+      "loss": 0.6844,
+      "step": 13334
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1345374048560477,
+      "learning_rate": 5.524023572621229e-07,
+      "loss": 0.9317,
+      "step": 13335
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.076263494631379,
+      "learning_rate": 5.519288847838783e-07,
+      "loss": 0.6435,
+      "step": 13336
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5945627768602058,
+      "learning_rate": 5.514556034491569e-07,
+      "loss": 0.8366,
+      "step": 13337
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8896568903306705,
+      "learning_rate": 5.509825132782948e-07,
+      "loss": 0.8596,
+      "step": 13338
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.195575748028536,
+      "learning_rate": 5.505096142916233e-07,
+      "loss": 0.6423,
+      "step": 13339
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0932466537881342,
+      "learning_rate": 5.50036906509463e-07,
+      "loss": 0.6661,
+      "step": 13340
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.701503088676176,
+      "learning_rate": 5.495643899521269e-07,
+      "loss": 0.9638,
+      "step": 13341
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6456958923206737,
+      "learning_rate": 5.4909206463992e-07,
+      "loss": 0.7812,
+      "step": 13342
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0554946165575416,
+      "learning_rate": 5.486199305931411e-07,
+      "loss": 0.623,
+      "step": 13343
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2582590540520378,
+      "learning_rate": 5.481479878320784e-07,
+      "loss": 0.7593,
+      "step": 13344
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0153554788334935,
+      "learning_rate": 5.476762363770088e-07,
+      "loss": 0.618,
+      "step": 13345
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9604032020350672,
+      "learning_rate": 5.472046762482092e-07,
+      "loss": 0.7282,
+      "step": 13346
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7542212283005176,
+      "learning_rate": 5.467333074659409e-07,
+      "loss": 0.8169,
+      "step": 13347
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8383521713338533,
+      "learning_rate": 5.462621300504606e-07,
+      "loss": 0.788,
+      "step": 13348
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1314578416316223,
+      "learning_rate": 5.457911440220154e-07,
+      "loss": 0.7818,
+      "step": 13349
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.461295818907835,
+      "learning_rate": 5.453203494008435e-07,
+      "loss": 0.813,
+      "step": 13350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.85067906091478,
+      "learning_rate": 5.4484974620718e-07,
+      "loss": 0.8565,
+      "step": 13351
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9583428303241304,
+      "learning_rate": 5.443793344612425e-07,
+      "loss": 0.7738,
+      "step": 13352
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8759499099763746,
+      "learning_rate": 5.439091141832498e-07,
+      "loss": 0.7462,
+      "step": 13353
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 23.19467450733939,
+      "learning_rate": 5.434390853934063e-07,
+      "loss": 0.8094,
+      "step": 13354
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.725943937869342,
+      "learning_rate": 5.429692481119103e-07,
+      "loss": 0.8159,
+      "step": 13355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9765212710855152,
+      "learning_rate": 5.424996023589524e-07,
+      "loss": 0.6513,
+      "step": 13356
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.4110868296872185,
+      "learning_rate": 5.420301481547141e-07,
+      "loss": 0.7464,
+      "step": 13357
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3996976858336145,
+      "learning_rate": 5.415608855193688e-07,
+      "loss": 0.963,
+      "step": 13358
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.808407615011808,
+      "learning_rate": 5.410918144730815e-07,
+      "loss": 0.6471,
+      "step": 13359
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.210804773460904,
+      "learning_rate": 5.406229350360087e-07,
+      "loss": 0.8871,
+      "step": 13360
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.487085465712158,
+      "learning_rate": 5.401542472283006e-07,
+      "loss": 0.7815,
+      "step": 13361
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1694637608136822,
+      "learning_rate": 5.39685751070097e-07,
+      "loss": 0.6284,
+      "step": 13362
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3693488422760054,
+      "learning_rate": 5.392174465815308e-07,
+      "loss": 0.7744,
+      "step": 13363
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8197373725260833,
+      "learning_rate": 5.387493337827254e-07,
+      "loss": 0.8141,
+      "step": 13364
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.877953271990523,
+      "learning_rate": 5.382814126937969e-07,
+      "loss": 0.7794,
+      "step": 13365
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.182441270534838,
+      "learning_rate": 5.378136833348529e-07,
+      "loss": 0.8992,
+      "step": 13366
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.054259444822116,
+      "learning_rate": 5.373461457259926e-07,
+      "loss": 0.9201,
+      "step": 13367
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.008454253271142,
+      "learning_rate": 5.368787998873065e-07,
+      "loss": 0.6868,
+      "step": 13368
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.630466032034131,
+      "learning_rate": 5.364116458388802e-07,
+      "loss": 0.7964,
+      "step": 13369
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2186898874906205,
+      "learning_rate": 5.359446836007842e-07,
+      "loss": 0.817,
+      "step": 13370
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2153382451940464,
+      "learning_rate": 5.354779131930882e-07,
+      "loss": 0.717,
+      "step": 13371
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1430993912614995,
+      "learning_rate": 5.350113346358493e-07,
+      "loss": 0.739,
+      "step": 13372
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8794762680556998,
+      "learning_rate": 5.345449479491177e-07,
+      "loss": 0.6724,
+      "step": 13373
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8931754496691395,
+      "learning_rate": 5.340787531529346e-07,
+      "loss": 0.6823,
+      "step": 13374
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.013087444187271,
+      "learning_rate": 5.336127502673328e-07,
+      "loss": 0.7235,
+      "step": 13375
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8485444099698567,
+      "learning_rate": 5.331469393123401e-07,
+      "loss": 0.6932,
+      "step": 13376
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.090043195002415,
+      "learning_rate": 5.326813203079706e-07,
+      "loss": 0.8461,
+      "step": 13377
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0745139494406217,
+      "learning_rate": 5.322158932742328e-07,
+      "loss": 0.726,
+      "step": 13378
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1025417048461725,
+      "learning_rate": 5.3175065823113e-07,
+      "loss": 0.8811,
+      "step": 13379
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4945058698549554,
+      "learning_rate": 5.312856151986523e-07,
+      "loss": 0.5974,
+      "step": 13380
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1486771935809448,
+      "learning_rate": 5.308207641967839e-07,
+      "loss": 0.8004,
+      "step": 13381
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5493128009954065,
+      "learning_rate": 5.303561052455009e-07,
+      "loss": 0.8099,
+      "step": 13382
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3887145580161913,
+      "learning_rate": 5.298916383647695e-07,
+      "loss": 0.7329,
+      "step": 13383
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0098970277729546,
+      "learning_rate": 5.294273635745517e-07,
+      "loss": 0.6667,
+      "step": 13384
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.786537085722121,
+      "learning_rate": 5.289632808947959e-07,
+      "loss": 0.6483,
+      "step": 13385
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.309694845778335,
+      "learning_rate": 5.284993903454444e-07,
+      "loss": 0.8304,
+      "step": 13386
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0448141643680855,
+      "learning_rate": 5.280356919464341e-07,
+      "loss": 0.9654,
+      "step": 13387
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3986992771083306,
+      "learning_rate": 5.275721857176885e-07,
+      "loss": 0.8446,
+      "step": 13388
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2300950707916045,
+      "learning_rate": 5.271088716791273e-07,
+      "loss": 0.7316,
+      "step": 13389
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9954832836244898,
+      "learning_rate": 5.266457498506594e-07,
+      "loss": 0.6479,
+      "step": 13390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.131221554460724,
+      "learning_rate": 5.261828202521868e-07,
+      "loss": 0.7388,
+      "step": 13391
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9886985773736412,
+      "learning_rate": 5.257200829036018e-07,
+      "loss": 0.6998,
+      "step": 13392
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9866190636077163,
+      "learning_rate": 5.252575378247887e-07,
+      "loss": 0.6166,
+      "step": 13393
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.177511358619168,
+      "learning_rate": 5.24795185035627e-07,
+      "loss": 0.7326,
+      "step": 13394
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.590004342888544,
+      "learning_rate": 5.243330245559819e-07,
+      "loss": 0.8547,
+      "step": 13395
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1249156849105333,
+      "learning_rate": 5.238710564057137e-07,
+      "loss": 0.8206,
+      "step": 13396
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.189940047711482,
+      "learning_rate": 5.234092806046759e-07,
+      "loss": 0.8352,
+      "step": 13397
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.742997486185129,
+      "learning_rate": 5.229476971727115e-07,
+      "loss": 0.8024,
+      "step": 13398
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8389151167869626,
+      "learning_rate": 5.224863061296553e-07,
+      "loss": 0.9549,
+      "step": 13399
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.909441638491454,
+      "learning_rate": 5.220251074953348e-07,
+      "loss": 0.9269,
+      "step": 13400
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4396754328877277,
+      "learning_rate": 5.215641012895667e-07,
+      "loss": 0.8491,
+      "step": 13401
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.702193462020962,
+      "learning_rate": 5.211032875321659e-07,
+      "loss": 0.8895,
+      "step": 13402
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2190628904572445,
+      "learning_rate": 5.206426662429298e-07,
+      "loss": 0.7597,
+      "step": 13403
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.16997920954371,
+      "learning_rate": 5.201822374416549e-07,
+      "loss": 0.5792,
+      "step": 13404
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1568049684919073,
+      "learning_rate": 5.197220011481274e-07,
+      "loss": 0.747,
+      "step": 13405
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0421165501509546,
+      "learning_rate": 5.192619573821223e-07,
+      "loss": 0.6983,
+      "step": 13406
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9370826674793162,
+      "learning_rate": 5.188021061634102e-07,
+      "loss": 0.7074,
+      "step": 13407
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.557243317304714,
+      "learning_rate": 5.183424475117521e-07,
+      "loss": 0.7819,
+      "step": 13408
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9602698706269068,
+      "learning_rate": 5.178829814469006e-07,
+      "loss": 0.8308,
+      "step": 13409
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2581019239116344,
+      "learning_rate": 5.174237079885991e-07,
+      "loss": 0.7032,
+      "step": 13410
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9324367931135973,
+      "learning_rate": 5.169646271565831e-07,
+      "loss": 0.8272,
+      "step": 13411
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.531283150771301,
+      "learning_rate": 5.165057389705835e-07,
+      "loss": 0.7857,
+      "step": 13412
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.004903984656461,
+      "learning_rate": 5.160470434503156e-07,
+      "loss": 0.8788,
+      "step": 13413
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0502682341258316,
+      "learning_rate": 5.155885406154937e-07,
+      "loss": 0.7798,
+      "step": 13414
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.537237078900171,
+      "learning_rate": 5.151302304858192e-07,
+      "loss": 0.7653,
+      "step": 13415
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3912001924707855,
+      "learning_rate": 5.146721130809873e-07,
+      "loss": 0.787,
+      "step": 13416
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0566770360999365,
+      "learning_rate": 5.142141884206842e-07,
+      "loss": 0.6236,
+      "step": 13417
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3570198915901246,
+      "learning_rate": 5.137564565245873e-07,
+      "loss": 0.8075,
+      "step": 13418
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7480868995269807,
+      "learning_rate": 5.132989174123659e-07,
+      "loss": 0.7096,
+      "step": 13419
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.117077278679434,
+      "learning_rate": 5.128415711036849e-07,
+      "loss": 0.8448,
+      "step": 13420
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.257664537794276,
+      "learning_rate": 5.12384417618193e-07,
+      "loss": 0.8813,
+      "step": 13421
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7906844896049983,
+      "learning_rate": 5.119274569755378e-07,
+      "loss": 0.7177,
+      "step": 13422
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8774326327061255,
+      "learning_rate": 5.114706891953552e-07,
+      "loss": 0.8109,
+      "step": 13423
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7507510130262047,
+      "learning_rate": 5.110141142972735e-07,
+      "loss": 0.8274,
+      "step": 13424
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4098140662600556,
+      "learning_rate": 5.105577323009131e-07,
+      "loss": 0.8097,
+      "step": 13425
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6426047614971186,
+      "learning_rate": 5.101015432258843e-07,
+      "loss": 0.834,
+      "step": 13426
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8477655341164296,
+      "learning_rate": 5.096455470917943e-07,
+      "loss": 0.8082,
+      "step": 13427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3886294007984117,
+      "learning_rate": 5.091897439182342e-07,
+      "loss": 0.8534,
+      "step": 13428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4070477037672418,
+      "learning_rate": 5.087341337247914e-07,
+      "loss": 0.7355,
+      "step": 13429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.074969652481809,
+      "learning_rate": 5.082787165310471e-07,
+      "loss": 0.8338,
+      "step": 13430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3751181811907847,
+      "learning_rate": 5.078234923565684e-07,
+      "loss": 0.6432,
+      "step": 13431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3719019301481468,
+      "learning_rate": 5.073684612209201e-07,
+      "loss": 0.6591,
+      "step": 13432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.327236178153354,
+      "learning_rate": 5.069136231436539e-07,
+      "loss": 0.7253,
+      "step": 13433
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9931388997484962,
+      "learning_rate": 5.064589781443163e-07,
+      "loss": 0.8305,
+      "step": 13434
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8174824565586807,
+      "learning_rate": 5.060045262424446e-07,
+      "loss": 0.9193,
+      "step": 13435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.160202684984713,
+      "learning_rate": 5.055502674575663e-07,
+      "loss": 0.578,
+      "step": 13436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0857061959595833,
+      "learning_rate": 5.05096201809202e-07,
+      "loss": 0.651,
+      "step": 13437
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5187553819068347,
+      "learning_rate": 5.046423293168667e-07,
+      "loss": 0.7637,
+      "step": 13438
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.980904212384179,
+      "learning_rate": 5.041886500000603e-07,
+      "loss": 0.6577,
+      "step": 13439
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.766669136163777,
+      "learning_rate": 5.037351638782812e-07,
+      "loss": 0.7359,
+      "step": 13440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4586797034745382,
+      "learning_rate": 5.032818709710152e-07,
+      "loss": 0.9684,
+      "step": 13441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4381879971794596,
+      "learning_rate": 5.028287712977425e-07,
+      "loss": 0.754,
+      "step": 13442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.503088256682058,
+      "learning_rate": 5.023758648779336e-07,
+      "loss": 0.8436,
+      "step": 13443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0104216225439697,
+      "learning_rate": 5.019231517310491e-07,
+      "loss": 0.7026,
+      "step": 13444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0989006220662902,
+      "learning_rate": 5.014706318765466e-07,
+      "loss": 0.5905,
+      "step": 13445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9109818558194624,
+      "learning_rate": 5.010183053338685e-07,
+      "loss": 0.8009,
+      "step": 13446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8094781345719166,
+      "learning_rate": 5.00566172122453e-07,
+      "loss": 0.89,
+      "step": 13447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.389158956871156,
+      "learning_rate": 5.001142322617314e-07,
+      "loss": 0.8619,
+      "step": 13448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2724857894928956,
+      "learning_rate": 4.996624857711219e-07,
+      "loss": 0.745,
+      "step": 13449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3738326741791433,
+      "learning_rate": 4.992109326700384e-07,
+      "loss": 0.7548,
+      "step": 13450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9894763975701163,
+      "learning_rate": 4.987595729778849e-07,
+      "loss": 0.7638,
+      "step": 13451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3341582904328484,
+      "learning_rate": 4.983084067140575e-07,
+      "loss": 0.765,
+      "step": 13452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.159451029668141,
+      "learning_rate": 4.978574338979442e-07,
+      "loss": 0.8384,
+      "step": 13453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3630850382704132,
+      "learning_rate": 4.97406654548922e-07,
+      "loss": 0.7219,
+      "step": 13454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8924241367958021,
+      "learning_rate": 4.96956068686365e-07,
+      "loss": 0.8273,
+      "step": 13455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4649583348387805,
+      "learning_rate": 4.965056763296356e-07,
+      "loss": 0.7076,
+      "step": 13456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.881349153516498,
+      "learning_rate": 4.960554774980853e-07,
+      "loss": 0.7611,
+      "step": 13457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2202564128549223,
+      "learning_rate": 4.956054722110626e-07,
+      "loss": 0.772,
+      "step": 13458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.271983630453135,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.7232,
+      "step": 13459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1834887056119703,
+      "learning_rate": 4.94706042347941e-07,
+      "loss": 0.7558,
+      "step": 13460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.643549300028163,
+      "learning_rate": 4.942566178104924e-07,
+      "loss": 0.8493,
+      "step": 13461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.162905279405926,
+      "learning_rate": 4.938073868948707e-07,
+      "loss": 0.726,
+      "step": 13462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7947263924769021,
+      "learning_rate": 4.933583496203831e-07,
+      "loss": 0.673,
+      "step": 13463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1244557660413725,
+      "learning_rate": 4.929095060063227e-07,
+      "loss": 0.6468,
+      "step": 13464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.271166196622326,
+      "learning_rate": 4.92460856071979e-07,
+      "loss": 0.7177,
+      "step": 13465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4230973240951825,
+      "learning_rate": 4.920123998366327e-07,
+      "loss": 0.8287,
+      "step": 13466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.790804092355941,
+      "learning_rate": 4.915641373195512e-07,
+      "loss": 0.804,
+      "step": 13467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4790041327284573,
+      "learning_rate": 4.911160685400008e-07,
+      "loss": 0.6475,
+      "step": 13468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.057883728942431,
+      "learning_rate": 4.906681935172342e-07,
+      "loss": 0.9276,
+      "step": 13469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2445139810362793,
+      "learning_rate": 4.902205122704983e-07,
+      "loss": 0.8466,
+      "step": 13470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0441503515087494,
+      "learning_rate": 4.897730248190313e-07,
+      "loss": 0.8299,
+      "step": 13471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4718191101314613,
+      "learning_rate": 4.893257311820609e-07,
+      "loss": 0.7206,
+      "step": 13472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6359508696624205,
+      "learning_rate": 4.888786313788108e-07,
+      "loss": 0.5693,
+      "step": 13473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.475835893066225,
+      "learning_rate": 4.88431725428492e-07,
+      "loss": 0.8469,
+      "step": 13474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4815266921209322,
+      "learning_rate": 4.879850133503106e-07,
+      "loss": 0.6223,
+      "step": 13475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9110390465900013,
+      "learning_rate": 4.875384951634616e-07,
+      "loss": 0.7498,
+      "step": 13476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9759291788866935,
+      "learning_rate": 4.870921708871334e-07,
+      "loss": 0.6783,
+      "step": 13477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.485443571504746,
+      "learning_rate": 4.866460405405054e-07,
+      "loss": 0.8433,
+      "step": 13478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.02995968113646,
+      "learning_rate": 4.862001041427488e-07,
+      "loss": 0.6392,
+      "step": 13479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6760549132942626,
+      "learning_rate": 4.857543617130256e-07,
+      "loss": 0.8537,
+      "step": 13480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.036304483194039,
+      "learning_rate": 4.85308813270493e-07,
+      "loss": 0.7463,
+      "step": 13481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5885692593302125,
+      "learning_rate": 4.848634588342932e-07,
+      "loss": 0.7969,
+      "step": 13482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2129100773246444,
+      "learning_rate": 4.844182984235679e-07,
+      "loss": 0.7239,
+      "step": 13483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8960282123882664,
+      "learning_rate": 4.839733320574457e-07,
+      "loss": 0.7807,
+      "step": 13484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.337771681942927,
+      "learning_rate": 4.835285597550454e-07,
+      "loss": 0.8195,
+      "step": 13485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5531323246342867,
+      "learning_rate": 4.830839815354827e-07,
+      "loss": 0.7814,
+      "step": 13486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2055305854311995,
+      "learning_rate": 4.826395974178605e-07,
+      "loss": 0.6738,
+      "step": 13487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6161663856261272,
+      "learning_rate": 4.821954074212759e-07,
+      "loss": 0.6354,
+      "step": 13488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9746444541277117,
+      "learning_rate": 4.817514115648164e-07,
+      "loss": 0.7448,
+      "step": 13489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1729869844301852,
+      "learning_rate": 4.813076098675601e-07,
+      "loss": 0.6175,
+      "step": 13490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9723625046255526,
+      "learning_rate": 4.808640023485805e-07,
+      "loss": 0.6822,
+      "step": 13491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8122079560205913,
+      "learning_rate": 4.804205890269398e-07,
+      "loss": 0.7966,
+      "step": 13492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.495489246942559,
+      "learning_rate": 4.799773699216914e-07,
+      "loss": 0.6968,
+      "step": 13493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4340243863065694,
+      "learning_rate": 4.795343450518825e-07,
+      "loss": 0.8212,
+      "step": 13494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8573137556165977,
+      "learning_rate": 4.790915144365504e-07,
+      "loss": 0.7455,
+      "step": 13495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7809381080906335,
+      "learning_rate": 4.786488780947246e-07,
+      "loss": 0.7372,
+      "step": 13496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5045306614756857,
+      "learning_rate": 4.782064360454258e-07,
+      "loss": 0.8003,
+      "step": 13497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.067024358406838,
+      "learning_rate": 4.777641883076661e-07,
+      "loss": 0.7737,
+      "step": 13498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9013513235683654,
+      "learning_rate": 4.773221349004531e-07,
+      "loss": 0.6633,
+      "step": 13499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0296529855420447,
+      "learning_rate": 4.768802758427782e-07,
+      "loss": 0.7134,
+      "step": 13500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.437664097831097,
+      "learning_rate": 4.7643861115363234e-07,
+      "loss": 0.7918,
+      "step": 13501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5299341277613054,
+      "learning_rate": 4.759971408519937e-07,
+      "loss": 1.0019,
+      "step": 13502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9930459264203269,
+      "learning_rate": 4.755558649568337e-07,
+      "loss": 0.6537,
+      "step": 13503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8702686904261316,
+      "learning_rate": 4.7511478348711447e-07,
+      "loss": 0.8361,
+      "step": 13504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7631051302866148,
+      "learning_rate": 4.7467389646178976e-07,
+      "loss": 0.8287,
+      "step": 13505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5034738743735474,
+      "learning_rate": 4.742332038998082e-07,
+      "loss": 0.9618,
+      "step": 13506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0321122889098233,
+      "learning_rate": 4.7379270582010416e-07,
+      "loss": 0.7558,
+      "step": 13507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8866758077628887,
+      "learning_rate": 4.7335240224160695e-07,
+      "loss": 0.789,
+      "step": 13508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7346241989256936,
+      "learning_rate": 4.729122931832392e-07,
+      "loss": 0.8037,
+      "step": 13509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1324747582588113,
+      "learning_rate": 4.7247237866391236e-07,
+      "loss": 0.6603,
+      "step": 13510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.58710670761524,
+      "learning_rate": 4.7203265870253144e-07,
+      "loss": 0.7046,
+      "step": 13511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6496306715496583,
+      "learning_rate": 4.715931333179913e-07,
+      "loss": 0.6358,
+      "step": 13512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9328890079168644,
+      "learning_rate": 4.7115380252917897e-07,
+      "loss": 0.6477,
+      "step": 13513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5249702398122014,
+      "learning_rate": 4.707146663549744e-07,
+      "loss": 1.0557,
+      "step": 13514
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2435565107637676,
+      "learning_rate": 4.7027572481424753e-07,
+      "loss": 0.6391,
+      "step": 13515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.160486629388527,
+      "learning_rate": 4.6983697792586103e-07,
+      "loss": 0.8456,
+      "step": 13516
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.515725335966096,
+      "learning_rate": 4.6939842570867034e-07,
+      "loss": 0.9135,
+      "step": 13517
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1577303623192314,
+      "learning_rate": 4.6896006818151706e-07,
+      "loss": 0.6489,
+      "step": 13518
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8622391455450795,
+      "learning_rate": 4.685219053632423e-07,
+      "loss": 0.737,
+      "step": 13519
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.438352387763515,
+      "learning_rate": 4.6808393727267307e-07,
+      "loss": 0.7799,
+      "step": 13520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.974717604314722,
+      "learning_rate": 4.676461639286306e-07,
+      "loss": 0.7756,
+      "step": 13521
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0561015238261438,
+      "learning_rate": 4.6720858534992584e-07,
+      "loss": 0.5866,
+      "step": 13522
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8971010355291562,
+      "learning_rate": 4.667712015553627e-07,
+      "loss": 0.7352,
+      "step": 13523
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0552645365516249,
+      "learning_rate": 4.663340125637389e-07,
+      "loss": 0.7244,
+      "step": 13524
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.710908811266251,
+      "learning_rate": 4.658970183938383e-07,
+      "loss": 0.7312,
+      "step": 13525
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0502562665562323,
+      "learning_rate": 4.654602190644408e-07,
+      "loss": 0.6435,
+      "step": 13526
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0764080913913667,
+      "learning_rate": 4.650236145943171e-07,
+      "loss": 0.6693,
+      "step": 13527
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.290198687279886,
+      "learning_rate": 4.6458720500222873e-07,
+      "loss": 0.7215,
+      "step": 13528
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2436069367800864,
+      "learning_rate": 4.6415099030692914e-07,
+      "loss": 0.6475,
+      "step": 13529
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4650474319595364,
+      "learning_rate": 4.6371497052716265e-07,
+      "loss": 0.7996,
+      "step": 13530
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.241731298237602,
+      "learning_rate": 4.6327914568166763e-07,
+      "loss": 0.6137,
+      "step": 13531
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1259181750618996,
+      "learning_rate": 4.628435157891709e-07,
+      "loss": 0.5676,
+      "step": 13532
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2408508412788604,
+      "learning_rate": 4.624080808683923e-07,
+      "loss": 0.626,
+      "step": 13533
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.789750962287008,
+      "learning_rate": 4.619728409380453e-07,
+      "loss": 0.6521,
+      "step": 13534
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9391600895977091,
+      "learning_rate": 4.615377960168321e-07,
+      "loss": 0.5514,
+      "step": 13535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.4122755754510194,
+      "learning_rate": 4.611029461234473e-07,
+      "loss": 0.899,
+      "step": 13536
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3678543910340928,
+      "learning_rate": 4.6066829127657807e-07,
+      "loss": 0.8046,
+      "step": 13537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.133481109974179,
+      "learning_rate": 4.6023383149490066e-07,
+      "loss": 0.6953,
+      "step": 13538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.088414983154378,
+      "learning_rate": 4.597995667970878e-07,
+      "loss": 0.7306,
+      "step": 13539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9938939398731714,
+      "learning_rate": 4.5936549720179855e-07,
+      "loss": 0.7237,
+      "step": 13540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3638616821304996,
+      "learning_rate": 4.589316227276852e-07,
+      "loss": 0.6573,
+      "step": 13541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.122163524786143,
+      "learning_rate": 4.584979433933956e-07,
+      "loss": 0.7017,
+      "step": 13542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1574992928584211,
+      "learning_rate": 4.5806445921756146e-07,
+      "loss": 0.6869,
+      "step": 13543
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.394203464248417,
+      "learning_rate": 4.5763117021881467e-07,
+      "loss": 0.6816,
+      "step": 13544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.313605912660832,
+      "learning_rate": 4.571980764157724e-07,
+      "loss": 0.8875,
+      "step": 13545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3068013231287585,
+      "learning_rate": 4.567651778270465e-07,
+      "loss": 0.8285,
+      "step": 13546
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.198948384365828,
+      "learning_rate": 4.563324744712389e-07,
+      "loss": 0.8639,
+      "step": 13547
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.794308248416161,
+      "learning_rate": 4.5589996636694387e-07,
+      "loss": 0.8401,
+      "step": 13548
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.0859465042419,
+      "learning_rate": 4.5546765353274846e-07,
+      "loss": 0.788,
+      "step": 13549
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6269122058390524,
+      "learning_rate": 4.5503553598722826e-07,
+      "loss": 0.9294,
+      "step": 13550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.690958119574634,
+      "learning_rate": 4.546036137489529e-07,
+      "loss": 0.8444,
+      "step": 13551
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0326535266156496,
+      "learning_rate": 4.5417188683648417e-07,
+      "loss": 0.7144,
+      "step": 13552
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.20797128249038,
+      "learning_rate": 4.5374035526837334e-07,
+      "loss": 0.6859,
+      "step": 13553
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0477077835006776,
+      "learning_rate": 4.5330901906316506e-07,
+      "loss": 0.7561,
+      "step": 13554
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1880507320001246,
+      "learning_rate": 4.528778782393939e-07,
+      "loss": 0.8988,
+      "step": 13555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.365903625692899,
+      "learning_rate": 4.5244693281558615e-07,
+      "loss": 0.8212,
+      "step": 13556
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1945427460697733,
+      "learning_rate": 4.5201618281026374e-07,
+      "loss": 0.6926,
+      "step": 13557
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.049645892849436,
+      "learning_rate": 4.5158562824193395e-07,
+      "loss": 0.8986,
+      "step": 13558
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8620973787401438,
+      "learning_rate": 4.511552691290988e-07,
+      "loss": 0.55,
+      "step": 13559
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.360301547705526,
+      "learning_rate": 4.507251054902539e-07,
+      "loss": 0.6768,
+      "step": 13560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.000701152312974,
+      "learning_rate": 4.5029513734388187e-07,
+      "loss": 0.7543,
+      "step": 13561
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.317199518434685,
+      "learning_rate": 4.498653647084611e-07,
+      "loss": 0.7207,
+      "step": 13562
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.078670557578651,
+      "learning_rate": 4.494357876024591e-07,
+      "loss": 0.7618,
+      "step": 13563
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0689188749225433,
+      "learning_rate": 4.490064060443361e-07,
+      "loss": 0.8998,
+      "step": 13564
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1346882331212815,
+      "learning_rate": 4.4857722005254345e-07,
+      "loss": 0.7343,
+      "step": 13565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.294907440982035,
+      "learning_rate": 4.4814822964552363e-07,
+      "loss": 0.8031,
+      "step": 13566
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8024031986541376,
+      "learning_rate": 4.47719434841713e-07,
+      "loss": 0.8882,
+      "step": 13567
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2295910967706245,
+      "learning_rate": 4.4729083565953626e-07,
+      "loss": 0.8215,
+      "step": 13568
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0231055906186683,
+      "learning_rate": 4.468624321174109e-07,
+      "loss": 0.8262,
+      "step": 13569
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.4683811072826325,
+      "learning_rate": 4.4643422423374825e-07,
+      "loss": 0.89,
+      "step": 13570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7287899474808084,
+      "learning_rate": 4.460062120269476e-07,
+      "loss": 0.8609,
+      "step": 13571
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6136021754463346,
+      "learning_rate": 4.4557839551540303e-07,
+      "loss": 0.9217,
+      "step": 13572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3476314831017753,
+      "learning_rate": 4.4515077471749767e-07,
+      "loss": 0.775,
+      "step": 13573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.076514265918856,
+      "learning_rate": 4.4472334965160736e-07,
+      "loss": 0.8166,
+      "step": 13574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2338175518088965,
+      "learning_rate": 4.442961203361013e-07,
+      "loss": 0.8431,
+      "step": 13575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2423531164933626,
+      "learning_rate": 4.438690867893347e-07,
+      "loss": 0.6875,
+      "step": 13576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1383479658960156,
+      "learning_rate": 4.4344224902966194e-07,
+      "loss": 0.7916,
+      "step": 13577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3263616462293766,
+      "learning_rate": 4.430156070754249e-07,
+      "loss": 0.7391,
+      "step": 13578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.971698585041339,
+      "learning_rate": 4.4258916094495394e-07,
+      "loss": 0.8564,
+      "step": 13579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9085281026322503,
+      "learning_rate": 4.421629106565778e-07,
+      "loss": 0.7493,
+      "step": 13580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.913444519548335,
+      "learning_rate": 4.417368562286117e-07,
+      "loss": 0.8745,
+      "step": 13581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.03713160314638,
+      "learning_rate": 4.4131099767936504e-07,
+      "loss": 0.6978,
+      "step": 13582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.562826037914657,
+      "learning_rate": 4.4088533502713803e-07,
+      "loss": 0.8642,
+      "step": 13583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0653098648703763,
+      "learning_rate": 4.4045986829022e-07,
+      "loss": 0.9091,
+      "step": 13584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2632515439856924,
+      "learning_rate": 4.400345974868986e-07,
+      "loss": 0.6425,
+      "step": 13585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.227123617894183,
+      "learning_rate": 4.396095226354441e-07,
+      "loss": 0.6402,
+      "step": 13586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.631108932295734,
+      "learning_rate": 4.391846437541258e-07,
+      "loss": 0.9117,
+      "step": 13587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5560627544703314,
+      "learning_rate": 4.3875996086120085e-07,
+      "loss": 0.9766,
+      "step": 13588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.114285745106508,
+      "learning_rate": 4.38335473974919e-07,
+      "loss": 0.729,
+      "step": 13589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8623571602555575,
+      "learning_rate": 4.379111831135213e-07,
+      "loss": 0.778,
+      "step": 13590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0218781230127862,
+      "learning_rate": 4.374870882952398e-07,
+      "loss": 0.5573,
+      "step": 13591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1003307508823754,
+      "learning_rate": 4.370631895382993e-07,
+      "loss": 0.8545,
+      "step": 13592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.685715740669534,
+      "learning_rate": 4.3663948686091694e-07,
+      "loss": 0.7793,
+      "step": 13593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.831554522192043,
+      "learning_rate": 4.362159802812971e-07,
+      "loss": 0.9309,
+      "step": 13594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.996369275468818,
+      "learning_rate": 4.357926698176418e-07,
+      "loss": 0.7524,
+      "step": 13595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.105809183091435,
+      "learning_rate": 4.3536955548814153e-07,
+      "loss": 0.7734,
+      "step": 13596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2007148805659864,
+      "learning_rate": 4.3494663731097564e-07,
+      "loss": 0.8452,
+      "step": 13597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.473794724791232,
+      "learning_rate": 4.3452391530432016e-07,
+      "loss": 0.7602,
+      "step": 13598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0580951480290413,
+      "learning_rate": 4.341013894863405e-07,
+      "loss": 0.8348,
+      "step": 13599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.113443902670121,
+      "learning_rate": 4.3367905987519275e-07,
+      "loss": 0.7238,
+      "step": 13600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0688417021903622,
+      "learning_rate": 4.332569264890252e-07,
+      "loss": 0.6141,
+      "step": 13601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1406655188165575,
+      "learning_rate": 4.328349893459777e-07,
+      "loss": 0.7432,
+      "step": 13602
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1459022322326238,
+      "learning_rate": 4.3241324846418355e-07,
+      "loss": 0.6583,
+      "step": 13603
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.339188971286666,
+      "learning_rate": 4.3199170386176325e-07,
+      "loss": 0.7703,
+      "step": 13604
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.730820231548082,
+      "learning_rate": 4.31570355556834e-07,
+      "loss": 0.8238,
+      "step": 13605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.172603295129102,
+      "learning_rate": 4.3114920356750076e-07,
+      "loss": 0.8278,
+      "step": 13606
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3478259290771253,
+      "learning_rate": 4.3072824791186187e-07,
+      "loss": 0.939,
+      "step": 13607
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0337214536119548,
+      "learning_rate": 4.3030748860800606e-07,
+      "loss": 0.6026,
+      "step": 13608
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.687159195169746,
+      "learning_rate": 4.2988692567401515e-07,
+      "loss": 0.8772,
+      "step": 13609
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.203343808122298,
+      "learning_rate": 4.294665591279601e-07,
+      "loss": 0.6603,
+      "step": 13610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.145562728637036,
+      "learning_rate": 4.290463889879082e-07,
+      "loss": 0.7502,
+      "step": 13611
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.995370765770525,
+      "learning_rate": 4.286264152719111e-07,
+      "loss": 1.0197,
+      "step": 13612
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.053067536814452,
+      "learning_rate": 4.2820663799801876e-07,
+      "loss": 0.5875,
+      "step": 13613
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2732328257625074,
+      "learning_rate": 4.2778705718426907e-07,
+      "loss": 0.8913,
+      "step": 13614
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.743746289396073,
+      "learning_rate": 4.273676728486925e-07,
+      "loss": 0.6963,
+      "step": 13615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9962228057292506,
+      "learning_rate": 4.2694848500931086e-07,
+      "loss": 0.7508,
+      "step": 13616
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.876757873437536,
+      "learning_rate": 4.2652949368413634e-07,
+      "loss": 0.7723,
+      "step": 13617
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.146825084822663,
+      "learning_rate": 4.2611069889117727e-07,
+      "loss": 0.9546,
+      "step": 13618
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1216806303519404,
+      "learning_rate": 4.2569210064842716e-07,
+      "loss": 0.6691,
+      "step": 13619
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6352926220714585,
+      "learning_rate": 4.252736989738737e-07,
+      "loss": 0.6518,
+      "step": 13620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.224586669589755,
+      "learning_rate": 4.2485549388549927e-07,
+      "loss": 0.9272,
+      "step": 13621
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2206562256334386,
+      "learning_rate": 4.244374854012734e-07,
+      "loss": 0.6989,
+      "step": 13622
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9465094219078103,
+      "learning_rate": 4.2401967353915885e-07,
+      "loss": 0.8765,
+      "step": 13623
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.104663192286205,
+      "learning_rate": 4.236020583171108e-07,
+      "loss": 0.6312,
+      "step": 13624
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4876125646676854,
+      "learning_rate": 4.2318463975307423e-07,
+      "loss": 0.8051,
+      "step": 13625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7744444477577006,
+      "learning_rate": 4.227674178649865e-07,
+      "loss": 0.9892,
+      "step": 13626
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.346362901625551,
+      "learning_rate": 4.2235039267077615e-07,
+      "loss": 0.7439,
+      "step": 13627
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.033436096146876,
+      "learning_rate": 4.219335641883654e-07,
+      "loss": 0.8616,
+      "step": 13628
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.461687553046537,
+      "learning_rate": 4.215169324356666e-07,
+      "loss": 0.7822,
+      "step": 13629
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4141403431227,
+      "learning_rate": 4.2110049743057933e-07,
+      "loss": 0.9772,
+      "step": 13630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3638583167353384,
+      "learning_rate": 4.206842591910032e-07,
+      "loss": 0.7527,
+      "step": 13631
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2155498713835127,
+      "learning_rate": 4.202682177348227e-07,
+      "loss": 0.6829,
+      "step": 13632
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1547402785140282,
+      "learning_rate": 4.198523730799164e-07,
+      "loss": 0.7027,
+      "step": 13633
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4226093226624297,
+      "learning_rate": 4.194367252441545e-07,
+      "loss": 0.8267,
+      "step": 13634
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.730843669429906,
+      "learning_rate": 4.1902127424539695e-07,
+      "loss": 0.8659,
+      "step": 13635
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.260117530293354,
+      "learning_rate": 4.186060201014991e-07,
+      "loss": 0.6288,
+      "step": 13636
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.464212529980822,
+      "learning_rate": 4.181909628303027e-07,
+      "loss": 0.9064,
+      "step": 13637
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.037347286171343,
+      "learning_rate": 4.177761024496452e-07,
+      "loss": 0.6817,
+      "step": 13638
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3754779028681794,
+      "learning_rate": 4.1736143897735394e-07,
+      "loss": 0.7759,
+      "step": 13639
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.715265076300565,
+      "learning_rate": 4.1694697243124804e-07,
+      "loss": 0.9294,
+      "step": 13640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7755485306597447,
+      "learning_rate": 4.1653270282913715e-07,
+      "loss": 0.7202,
+      "step": 13641
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0741040570402134,
+      "learning_rate": 4.1611863018882424e-07,
+      "loss": 0.6187,
+      "step": 13642
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8507904278571494,
+      "learning_rate": 4.157047545281029e-07,
+      "loss": 0.9589,
+      "step": 13643
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2662066003525614,
+      "learning_rate": 4.152910758647577e-07,
+      "loss": 0.8068,
+      "step": 13644
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.450574231971,
+      "learning_rate": 4.14877594216565e-07,
+      "loss": 0.865,
+      "step": 13645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8909233955795022,
+      "learning_rate": 4.14464309601294e-07,
+      "loss": 0.6505,
+      "step": 13646
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.117536771897203,
+      "learning_rate": 4.14051222036706e-07,
+      "loss": 1.0071,
+      "step": 13647
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.449472765145988,
+      "learning_rate": 4.1363833154054786e-07,
+      "loss": 0.6889,
+      "step": 13648
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3716884207198867,
+      "learning_rate": 4.1322563813056606e-07,
+      "loss": 0.8802,
+      "step": 13649
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.848734672392301,
+      "learning_rate": 4.1281314182449405e-07,
+      "loss": 0.8139,
+      "step": 13650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.069979325954283,
+      "learning_rate": 4.1240084264005774e-07,
+      "loss": 0.9606,
+      "step": 13651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.15378949428568,
+      "learning_rate": 4.1198874059497405e-07,
+      "loss": 0.6709,
+      "step": 13652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0684909386141213,
+      "learning_rate": 4.1157683570695105e-07,
+      "loss": 0.6469,
+      "step": 13653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8316455542948515,
+      "learning_rate": 4.111651279936929e-07,
+      "loss": 0.7433,
+      "step": 13654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5634055426668345,
+      "learning_rate": 4.1075361747288657e-07,
+      "loss": 0.7905,
+      "step": 13655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.985996439753649,
+      "learning_rate": 4.103423041622195e-07,
+      "loss": 0.8129,
+      "step": 13656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.497924648494981,
+      "learning_rate": 4.099311880793655e-07,
+      "loss": 0.792,
+      "step": 13657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3926154791712557,
+      "learning_rate": 4.0952026924199027e-07,
+      "loss": 0.6689,
+      "step": 13658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2552744462165397,
+      "learning_rate": 4.091095476677531e-07,
+      "loss": 0.8132,
+      "step": 13659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.198752564938521,
+      "learning_rate": 4.0869902337430323e-07,
+      "loss": 0.9071,
+      "step": 13660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1990612735501767,
+      "learning_rate": 4.082886963792815e-07,
+      "loss": 0.7898,
+      "step": 13661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3228452055727686,
+      "learning_rate": 4.0787856670032044e-07,
+      "loss": 0.8397,
+      "step": 13662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.270499913401409,
+      "learning_rate": 4.0746863435504434e-07,
+      "loss": 0.7852,
+      "step": 13663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1642643550302072,
+      "learning_rate": 4.070588993610697e-07,
+      "loss": 0.632,
+      "step": 13664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1717955413342547,
+      "learning_rate": 4.066493617360029e-07,
+      "loss": 0.6362,
+      "step": 13665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.035227003468503,
+      "learning_rate": 4.062400214974432e-07,
+      "loss": 0.7447,
+      "step": 13666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.929390978550983,
+      "learning_rate": 4.0583087866298056e-07,
+      "loss": 0.9362,
+      "step": 13667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1243952617491573,
+      "learning_rate": 4.054219332501974e-07,
+      "loss": 0.7041,
+      "step": 13668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3096935875092286,
+      "learning_rate": 4.050131852766659e-07,
+      "loss": 0.7596,
+      "step": 13669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.1545933776246535,
+      "learning_rate": 4.046046347599514e-07,
+      "loss": 0.9009,
+      "step": 13670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8133111147791183,
+      "learning_rate": 4.0419628171760927e-07,
+      "loss": 0.737,
+      "step": 13671
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.960264119253207,
+      "learning_rate": 4.0378812616719056e-07,
+      "loss": 0.7779,
+      "step": 13672
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6719283318247484,
+      "learning_rate": 4.0338016812623007e-07,
+      "loss": 0.9014,
+      "step": 13673
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8677118294257689,
+      "learning_rate": 4.029724076122621e-07,
+      "loss": 0.7638,
+      "step": 13674
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.853731472010144,
+      "learning_rate": 4.025648446428082e-07,
+      "loss": 0.7192,
+      "step": 13675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1787739739881973,
+      "learning_rate": 4.021574792353816e-07,
+      "loss": 0.6256,
+      "step": 13676
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.472495331641524,
+      "learning_rate": 4.017503114074883e-07,
+      "loss": 0.855,
+      "step": 13677
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2447114704311815,
+      "learning_rate": 4.0134334117662375e-07,
+      "loss": 0.5837,
+      "step": 13678
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.41878140873371,
+      "learning_rate": 4.009365685602795e-07,
+      "loss": 0.6711,
+      "step": 13679
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2819838867144666,
+      "learning_rate": 4.005299935759327e-07,
+      "loss": 0.84,
+      "step": 13680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.063695732891486,
+      "learning_rate": 4.001236162410549e-07,
+      "loss": 0.7254,
+      "step": 13681
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.119456926758034,
+      "learning_rate": 3.997174365731105e-07,
+      "loss": 0.7498,
+      "step": 13682
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3852531476998013,
+      "learning_rate": 3.993114545895532e-07,
+      "loss": 0.8293,
+      "step": 13683
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.122897065573846,
+      "learning_rate": 3.989056703078292e-07,
+      "loss": 0.7895,
+      "step": 13684
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0202671573567836,
+      "learning_rate": 3.985000837453756e-07,
+      "loss": 0.6544,
+      "step": 13685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3105508065020754,
+      "learning_rate": 3.9809469491962126e-07,
+      "loss": 0.8326,
+      "step": 13686
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4237538664047293,
+      "learning_rate": 3.976895038479872e-07,
+      "loss": 0.9916,
+      "step": 13687
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.580621221815072,
+      "learning_rate": 3.9728451054788396e-07,
+      "loss": 0.7521,
+      "step": 13688
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3320732668642816,
+      "learning_rate": 3.968797150367171e-07,
+      "loss": 0.8281,
+      "step": 13689
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2940162084233835,
+      "learning_rate": 3.9647511733188147e-07,
+      "loss": 0.6891,
+      "step": 13690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9411209464788994,
+      "learning_rate": 3.960707174507611e-07,
+      "loss": 0.8489,
+      "step": 13691
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9926455720126417,
+      "learning_rate": 3.9566651541073586e-07,
+      "loss": 0.8212,
+      "step": 13692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2176552693362472,
+      "learning_rate": 3.9526251122917526e-07,
+      "loss": 0.6396,
+      "step": 13693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8877044579084329,
+      "learning_rate": 3.948587049234398e-07,
+      "loss": 0.5825,
+      "step": 13694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4993445905336893,
+      "learning_rate": 3.944550965108818e-07,
+      "loss": 0.8852,
+      "step": 13695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.422509764456094,
+      "learning_rate": 3.940516860088445e-07,
+      "loss": 1.0998,
+      "step": 13696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0342382075515486,
+      "learning_rate": 3.936484734346663e-07,
+      "loss": 0.8554,
+      "step": 13697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2386944733748875,
+      "learning_rate": 3.932454588056711e-07,
+      "loss": 0.7732,
+      "step": 13698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1405546514366622,
+      "learning_rate": 3.928426421391773e-07,
+      "loss": 0.5993,
+      "step": 13699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4608627436287804,
+      "learning_rate": 3.924400234524972e-07,
+      "loss": 0.8941,
+      "step": 13700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8204591331308362,
+      "learning_rate": 3.9203760276293024e-07,
+      "loss": 0.861,
+      "step": 13701
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7579672140315803,
+      "learning_rate": 3.9163538008777035e-07,
+      "loss": 0.8879,
+      "step": 13702
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4257868674289105,
+      "learning_rate": 3.9123335544430153e-07,
+      "loss": 0.7592,
+      "step": 13703
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0858275500677002,
+      "learning_rate": 3.9083152884979935e-07,
+      "loss": 0.7279,
+      "step": 13704
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5834130767777377,
+      "learning_rate": 3.9042990032153227e-07,
+      "loss": 0.7619,
+      "step": 13705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.813265498966219,
+      "learning_rate": 3.9002846987675704e-07,
+      "loss": 0.8086,
+      "step": 13706
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.460882643746461,
+      "learning_rate": 3.8962723753272645e-07,
+      "loss": 0.6992,
+      "step": 13707
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5652420909293219,
+      "learning_rate": 3.892262033066818e-07,
+      "loss": 0.6049,
+      "step": 13708
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0136816998168623,
+      "learning_rate": 3.8882536721585486e-07,
+      "loss": 0.8443,
+      "step": 13709
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8595037919907587,
+      "learning_rate": 3.884247292774718e-07,
+      "loss": 0.773,
+      "step": 13710
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0361621834571766,
+      "learning_rate": 3.8802428950874884e-07,
+      "loss": 0.7238,
+      "step": 13711
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9392410718378053,
+      "learning_rate": 3.876240479268939e-07,
+      "loss": 0.7434,
+      "step": 13712
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0799300501941063,
+      "learning_rate": 3.872240045491055e-07,
+      "loss": 0.8563,
+      "step": 13713
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9262065003953486,
+      "learning_rate": 3.868241593925742e-07,
+      "loss": 0.7621,
+      "step": 13714
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.567068875249258,
+      "learning_rate": 3.8642451247448477e-07,
+      "loss": 0.689,
+      "step": 13715
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4479356276798967,
+      "learning_rate": 3.8602506381200666e-07,
+      "loss": 0.7872,
+      "step": 13716
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8692926403824184,
+      "learning_rate": 3.8562581342230897e-07,
+      "loss": 0.7629,
+      "step": 13717
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1949107705139794,
+      "learning_rate": 3.8522676132254635e-07,
+      "loss": 0.8095,
+      "step": 13718
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.249608147742702,
+      "learning_rate": 3.848279075298678e-07,
+      "loss": 0.7854,
+      "step": 13719
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9574515121293525,
+      "learning_rate": 3.8442925206141237e-07,
+      "loss": 0.9301,
+      "step": 13720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7900708712400901,
+      "learning_rate": 3.840307949343114e-07,
+      "loss": 0.9039,
+      "step": 13721
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0591189437707489,
+      "learning_rate": 3.8363253616568784e-07,
+      "loss": 0.6786,
+      "step": 13722
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.35193926747685,
+      "learning_rate": 3.8323447577265525e-07,
+      "loss": 0.8174,
+      "step": 13723
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.098514152036401,
+      "learning_rate": 3.828366137723183e-07,
+      "loss": 0.986,
+      "step": 13724
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7301752313606094,
+      "learning_rate": 3.82438950181776e-07,
+      "loss": 0.8212,
+      "step": 13725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.019137595800307,
+      "learning_rate": 3.820414850181153e-07,
+      "loss": 0.9171,
+      "step": 13726
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8359917272469004,
+      "learning_rate": 3.8164421829841756e-07,
+      "loss": 0.8112,
+      "step": 13727
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8667138181043932,
+      "learning_rate": 3.81247150039753e-07,
+      "loss": 0.6849,
+      "step": 13728
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9735558380352607,
+      "learning_rate": 3.80850280259184e-07,
+      "loss": 0.7015,
+      "step": 13729
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.138104821551085,
+      "learning_rate": 3.8045360897376814e-07,
+      "loss": 0.9009,
+      "step": 13730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3343709238678545,
+      "learning_rate": 3.800571362005473e-07,
+      "loss": 0.8002,
+      "step": 13731
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0417374087843838,
+      "learning_rate": 3.7966086195656e-07,
+      "loss": 0.6668,
+      "step": 13732
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0330375255087496,
+      "learning_rate": 3.7926478625883725e-07,
+      "loss": 0.8587,
+      "step": 13733
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8365836137113924,
+      "learning_rate": 3.7886890912439633e-07,
+      "loss": 0.8011,
+      "step": 13734
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3374716783300995,
+      "learning_rate": 3.784732305702504e-07,
+      "loss": 0.828,
+      "step": 13735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7001018775482555,
+      "learning_rate": 3.7807775061340257e-07,
+      "loss": 0.7915,
+      "step": 13736
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2994377562119284,
+      "learning_rate": 3.776824692708475e-07,
+      "loss": 0.7795,
+      "step": 13737
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.214676763354476,
+      "learning_rate": 3.772873865595711e-07,
+      "loss": 0.807,
+      "step": 13738
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0702102828988864,
+      "learning_rate": 3.768925024965503e-07,
+      "loss": 0.7117,
+      "step": 13739
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.335569002708931,
+      "learning_rate": 3.76497817098756e-07,
+      "loss": 0.8075,
+      "step": 13740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5582396268445313,
+      "learning_rate": 3.761033303831474e-07,
+      "loss": 0.9354,
+      "step": 13741
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.065447053806981,
+      "learning_rate": 3.7570904236667536e-07,
+      "loss": 0.9074,
+      "step": 13742
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.244321938886477,
+      "learning_rate": 3.7531495306628584e-07,
+      "loss": 0.7674,
+      "step": 13743
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0552533871428755,
+      "learning_rate": 3.749210624989125e-07,
+      "loss": 0.7421,
+      "step": 13744
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.481799209217219,
+      "learning_rate": 3.7452737068148173e-07,
+      "loss": 0.7495,
+      "step": 13745
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.485501900921298,
+      "learning_rate": 3.741338776309111e-07,
+      "loss": 0.683,
+      "step": 13746
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.205659635719587,
+      "learning_rate": 3.7374058336410945e-07,
+      "loss": 0.5604,
+      "step": 13747
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8658402836315673,
+      "learning_rate": 3.733474878979798e-07,
+      "loss": 0.8451,
+      "step": 13748
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.883144928230791,
+      "learning_rate": 3.729545912494115e-07,
+      "loss": 0.8238,
+      "step": 13749
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3147690146280293,
+      "learning_rate": 3.7256189343528926e-07,
+      "loss": 0.7095,
+      "step": 13750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1952549669226977,
+      "learning_rate": 3.721693944724897e-07,
+      "loss": 0.9328,
+      "step": 13751
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5527773169978087,
+      "learning_rate": 3.717770943778759e-07,
+      "loss": 0.7894,
+      "step": 13752
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.48169743832175,
+      "learning_rate": 3.7138499316830946e-07,
+      "loss": 0.7858,
+      "step": 13753
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4272143355769837,
+      "learning_rate": 3.7099309086063794e-07,
+      "loss": 0.9705,
+      "step": 13754
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.021838444449947,
+      "learning_rate": 3.706013874717024e-07,
+      "loss": 0.8941,
+      "step": 13755
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.379962196157043,
+      "learning_rate": 3.702098830183359e-07,
+      "loss": 0.7765,
+      "step": 13756
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8129412115966148,
+      "learning_rate": 3.698185775173607e-07,
+      "loss": 0.8606,
+      "step": 13757
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0645232537441305,
+      "learning_rate": 3.694274709855944e-07,
+      "loss": 0.6845,
+      "step": 13758
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2081557877410605,
+      "learning_rate": 3.6903656343984293e-07,
+      "loss": 0.8588,
+      "step": 13759
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.295889617281703,
+      "learning_rate": 3.6864585489690297e-07,
+      "loss": 0.681,
+      "step": 13760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2713218997320057,
+      "learning_rate": 3.6825534537356545e-07,
+      "loss": 0.7476,
+      "step": 13761
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1112147850999121,
+      "learning_rate": 3.678650348866114e-07,
+      "loss": 0.6422,
+      "step": 13762
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4769805133913856,
+      "learning_rate": 3.6747492345281297e-07,
+      "loss": 0.9395,
+      "step": 13763
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.950215852050583,
+      "learning_rate": 3.670850110889346e-07,
+      "loss": 0.9003,
+      "step": 13764
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.275866838880002,
+      "learning_rate": 3.6669529781173053e-07,
+      "loss": 0.9477,
+      "step": 13765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.017707724256171,
+      "learning_rate": 3.6630578363794965e-07,
+      "loss": 0.8093,
+      "step": 13766
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.885904794578305,
+      "learning_rate": 3.659164685843275e-07,
+      "loss": 0.7949,
+      "step": 13767
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4004865572255687,
+      "learning_rate": 3.655273526675962e-07,
+      "loss": 0.9369,
+      "step": 13768
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3557932640668287,
+      "learning_rate": 3.651384359044774e-07,
+      "loss": 0.7701,
+      "step": 13769
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.766935515087786,
+      "learning_rate": 3.647497183116799e-07,
+      "loss": 0.8686,
+      "step": 13770
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6433213363260923,
+      "learning_rate": 3.6436119990591214e-07,
+      "loss": 0.6618,
+      "step": 13771
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.241711221610153,
+      "learning_rate": 3.639728807038667e-07,
+      "loss": 0.827,
+      "step": 13772
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.37163823331968,
+      "learning_rate": 3.6358476072223205e-07,
+      "loss": 0.6026,
+      "step": 13773
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8883023231764405,
+      "learning_rate": 3.631968399776864e-07,
+      "loss": 0.5821,
+      "step": 13774
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.802368132543842,
+      "learning_rate": 3.628091184868976e-07,
+      "loss": 0.8906,
+      "step": 13775
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.204771775563643,
+      "learning_rate": 3.6242159626653004e-07,
+      "loss": 0.8949,
+      "step": 13776
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3851180390036326,
+      "learning_rate": 3.6203427333323495e-07,
+      "loss": 0.8756,
+      "step": 13777
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.530335662843583,
+      "learning_rate": 3.616471497036561e-07,
+      "loss": 0.7983,
+      "step": 13778
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.409363030199415,
+      "learning_rate": 3.6126022539442975e-07,
+      "loss": 0.7969,
+      "step": 13779
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.187334731587904,
+      "learning_rate": 3.6087350042218194e-07,
+      "loss": 0.9687,
+      "step": 13780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.59238671578501,
+      "learning_rate": 3.6048697480353225e-07,
+      "loss": 1.0155,
+      "step": 13781
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8344414272576401,
+      "learning_rate": 3.601006485550895e-07,
+      "loss": 0.7983,
+      "step": 13782
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.150105005147501,
+      "learning_rate": 3.597145216934556e-07,
+      "loss": 1.09,
+      "step": 13783
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.498488704648107,
+      "learning_rate": 3.593285942352237e-07,
+      "loss": 0.8765,
+      "step": 13784
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7995332513386524,
+      "learning_rate": 3.5894286619697627e-07,
+      "loss": 0.6806,
+      "step": 13785
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0911707412268448,
+      "learning_rate": 3.585573375952911e-07,
+      "loss": 0.9662,
+      "step": 13786
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2686584653766917,
+      "learning_rate": 3.581720084467344e-07,
+      "loss": 0.6309,
+      "step": 13787
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4719122127287436,
+      "learning_rate": 3.57786878767864e-07,
+      "loss": 0.5981,
+      "step": 13788
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.485025787454318,
+      "learning_rate": 3.5740194857523e-07,
+      "loss": 0.7228,
+      "step": 13789
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1740402343817324,
+      "learning_rate": 3.570172178853731e-07,
+      "loss": 0.8529,
+      "step": 13790
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.821122762825559,
+      "learning_rate": 3.56632686714829e-07,
+      "loss": 0.7654,
+      "step": 13791
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.397076836089336,
+      "learning_rate": 3.5624835508011824e-07,
+      "loss": 0.868,
+      "step": 13792
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.07136242781258,
+      "learning_rate": 3.5586422299775714e-07,
+      "loss": 0.653,
+      "step": 13793
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1796393571140977,
+      "learning_rate": 3.554802904842547e-07,
+      "loss": 0.6404,
+      "step": 13794
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2967360419001785,
+      "learning_rate": 3.5509655755610773e-07,
+      "loss": 0.7802,
+      "step": 13795
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6509087572068877,
+      "learning_rate": 3.547130242298064e-07,
+      "loss": 0.7566,
+      "step": 13796
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2546328242732936,
+      "learning_rate": 3.5432969052183186e-07,
+      "loss": 0.7793,
+      "step": 13797
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.250443075507316,
+      "learning_rate": 3.539465564486577e-07,
+      "loss": 0.6974,
+      "step": 13798
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2891408111196279,
+      "learning_rate": 3.5356362202674687e-07,
+      "loss": 0.5936,
+      "step": 13799
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0773520665058323,
+      "learning_rate": 3.5318088727255506e-07,
+      "loss": 0.862,
+      "step": 13800
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7654271733795976,
+      "learning_rate": 3.527983522025291e-07,
+      "loss": 0.8556,
+      "step": 13801
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8631948983101239,
+      "learning_rate": 3.5241601683310924e-07,
+      "loss": 0.6208,
+      "step": 13802
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8368874794811991,
+      "learning_rate": 3.520338811807222e-07,
+      "loss": 0.9069,
+      "step": 13803
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.289756303576698,
+      "learning_rate": 3.516519452617922e-07,
+      "loss": 0.8779,
+      "step": 13804
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1620562049390952,
+      "learning_rate": 3.5127020909273e-07,
+      "loss": 0.7658,
+      "step": 13805
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3707086844575436,
+      "learning_rate": 3.508886726899402e-07,
+      "loss": 0.8034,
+      "step": 13806
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.932724337885513,
+      "learning_rate": 3.5050733606981855e-07,
+      "loss": 0.63,
+      "step": 13807
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.680303230564355,
+      "learning_rate": 3.501261992487509e-07,
+      "loss": 0.8593,
+      "step": 13808
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8748017848024279,
+      "learning_rate": 3.4974526224311744e-07,
+      "loss": 0.8713,
+      "step": 13809
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.201537958923116,
+      "learning_rate": 3.493645250692862e-07,
+      "loss": 0.7183,
+      "step": 13810
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5097302399758354,
+      "learning_rate": 3.4898398774361854e-07,
+      "loss": 0.8446,
+      "step": 13811
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.312216650301658,
+      "learning_rate": 3.486036502824675e-07,
+      "loss": 0.6651,
+      "step": 13812
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1902480634904615,
+      "learning_rate": 3.4822351270217656e-07,
+      "loss": 0.6454,
+      "step": 13813
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1919100183767473,
+      "learning_rate": 3.478435750190817e-07,
+      "loss": 0.942,
+      "step": 13814
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2513077706760893,
+      "learning_rate": 3.474638372495098e-07,
+      "loss": 0.8202,
+      "step": 13815
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.915090927700682,
+      "learning_rate": 3.4708429940977785e-07,
+      "loss": 0.9401,
+      "step": 13816
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1145147237589204,
+      "learning_rate": 3.467049615161966e-07,
+      "loss": 0.7047,
+      "step": 13817
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.986634432594425,
+      "learning_rate": 3.463258235850653e-07,
+      "loss": 0.6433,
+      "step": 13818
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7916759223563523,
+      "learning_rate": 3.459468856326792e-07,
+      "loss": 0.7347,
+      "step": 13819
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.21449016806626,
+      "learning_rate": 3.455681476753209e-07,
+      "loss": 0.9195,
+      "step": 13820
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4498727518486594,
+      "learning_rate": 3.451896097292634e-07,
+      "loss": 0.8319,
+      "step": 13821
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.190852706459011,
+      "learning_rate": 3.448112718107766e-07,
+      "loss": 0.622,
+      "step": 13822
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.126821252501146,
+      "learning_rate": 3.444331339361168e-07,
+      "loss": 0.6328,
+      "step": 13823
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.887181129485368,
+      "learning_rate": 3.4405519612153326e-07,
+      "loss": 0.8283,
+      "step": 13824
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1029768005295466,
+      "learning_rate": 3.4367745838326807e-07,
+      "loss": 0.8673,
+      "step": 13825
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1264956696460962,
+      "learning_rate": 3.432999207375515e-07,
+      "loss": 0.7244,
+      "step": 13826
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7423783203949565,
+      "learning_rate": 3.4292258320061e-07,
+      "loss": 0.5107,
+      "step": 13827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.338734780169489,
+      "learning_rate": 3.425454457886551e-07,
+      "loss": 0.6964,
+      "step": 13828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1892592254440877,
+      "learning_rate": 3.4216850851789663e-07,
+      "loss": 0.7442,
+      "step": 13829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.195431431984995,
+      "learning_rate": 3.4179177140453045e-07,
+      "loss": 1.0798,
+      "step": 13830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.787961049326795,
+      "learning_rate": 3.414152344647459e-07,
+      "loss": 0.7639,
+      "step": 13831
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.44859508378395,
+      "learning_rate": 3.410388977147244e-07,
+      "loss": 0.7378,
+      "step": 13832
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.019424232002275,
+      "learning_rate": 3.406627611706376e-07,
+      "loss": 0.9847,
+      "step": 13833
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.426708135291803,
+      "learning_rate": 3.402868248486485e-07,
+      "loss": 0.7875,
+      "step": 13834
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8950572066051024,
+      "learning_rate": 3.3991108876491263e-07,
+      "loss": 0.7687,
+      "step": 13835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2260330779468842,
+      "learning_rate": 3.3953555293557483e-07,
+      "loss": 0.6514,
+      "step": 13836
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.011154801133645,
+      "learning_rate": 3.391602173767744e-07,
+      "loss": 0.8604,
+      "step": 13837
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.572197178375245,
+      "learning_rate": 3.387850821046401e-07,
+      "loss": 0.8152,
+      "step": 13838
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2748633685041737,
+      "learning_rate": 3.3841014713529184e-07,
+      "loss": 0.6834,
+      "step": 13839
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0357095014353386,
+      "learning_rate": 3.380354124848412e-07,
+      "loss": 0.6094,
+      "step": 13840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.147999550418396,
+      "learning_rate": 3.376608781693913e-07,
+      "loss": 0.7579,
+      "step": 13841
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.821343214056771,
+      "learning_rate": 3.372865442050377e-07,
+      "loss": 0.706,
+      "step": 13842
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.393715153392344,
+      "learning_rate": 3.3691241060786586e-07,
+      "loss": 0.8701,
+      "step": 13843
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4463909706773737,
+      "learning_rate": 3.3653847739395174e-07,
+      "loss": 0.846,
+      "step": 13844
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1083875112779875,
+      "learning_rate": 3.36164744579367e-07,
+      "loss": 0.9108,
+      "step": 13845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6723614092472334,
+      "learning_rate": 3.357912121801682e-07,
+      "loss": 0.7975,
+      "step": 13846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9284547768335307,
+      "learning_rate": 3.354178802124103e-07,
+      "loss": 0.7209,
+      "step": 13847
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7496745406966934,
+      "learning_rate": 3.350447486921338e-07,
+      "loss": 0.6034,
+      "step": 13848
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.619676757101361,
+      "learning_rate": 3.346718176353747e-07,
+      "loss": 0.9209,
+      "step": 13849
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0750710812762034,
+      "learning_rate": 3.342990870581575e-07,
+      "loss": 0.7142,
+      "step": 13850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.178360711691437,
+      "learning_rate": 3.339265569764993e-07,
+      "loss": 0.9011,
+      "step": 13851
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4096411770606783,
+      "learning_rate": 3.3355422740640905e-07,
+      "loss": 0.7912,
+      "step": 13852
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1483162396629822,
+      "learning_rate": 3.331820983638867e-07,
+      "loss": 0.8593,
+      "step": 13853
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2198564176473383,
+      "learning_rate": 3.3281016986492165e-07,
+      "loss": 0.6197,
+      "step": 13854
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4250037675463187,
+      "learning_rate": 3.3243844192549947e-07,
+      "loss": 0.7616,
+      "step": 13855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1151503111818477,
+      "learning_rate": 3.3206691456159233e-07,
+      "loss": 0.941,
+      "step": 13856
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9303291049387696,
+      "learning_rate": 3.316955877891659e-07,
+      "loss": 0.942,
+      "step": 13857
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.123710497491418,
+      "learning_rate": 3.313244616241773e-07,
+      "loss": 0.7822,
+      "step": 13858
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.277927236390682,
+      "learning_rate": 3.3095353608257385e-07,
+      "loss": 0.7573,
+      "step": 13859
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.268979490949993,
+      "learning_rate": 3.3058281118029553e-07,
+      "loss": 0.667,
+      "step": 13860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3060929638247645,
+      "learning_rate": 3.302122869332736e-07,
+      "loss": 0.7625,
+      "step": 13861
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.996798258470727,
+      "learning_rate": 3.2984196335742844e-07,
+      "loss": 0.6411,
+      "step": 13862
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.14155523592197,
+      "learning_rate": 3.294718404686775e-07,
+      "loss": 0.6614,
+      "step": 13863
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1975204607982366,
+      "learning_rate": 3.2910191828292083e-07,
+      "loss": 0.6487,
+      "step": 13864
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4701737296117736,
+      "learning_rate": 3.2873219681605894e-07,
+      "loss": 0.8626,
+      "step": 13865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4013002955718137,
+      "learning_rate": 3.283626760839775e-07,
+      "loss": 0.7688,
+      "step": 13866
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7322019470079235,
+      "learning_rate": 3.279933561025567e-07,
+      "loss": 0.7644,
+      "step": 13867
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0601453175314317,
+      "learning_rate": 3.276242368876664e-07,
+      "loss": 0.6716,
+      "step": 13868
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.142005447276918,
+      "learning_rate": 3.2725531845516744e-07,
+      "loss": 0.7192,
+      "step": 13869
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.001656251040427,
+      "learning_rate": 3.2688660082091594e-07,
+      "loss": 0.642,
+      "step": 13870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.216680002256527,
+      "learning_rate": 3.2651808400075426e-07,
+      "loss": 0.877,
+      "step": 13871
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3292068332005698,
+      "learning_rate": 3.2614976801051745e-07,
+      "loss": 0.647,
+      "step": 13872
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9331610225933753,
+      "learning_rate": 3.257816528660357e-07,
+      "loss": 0.6086,
+      "step": 13873
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4367192463699845,
+      "learning_rate": 3.254137385831263e-07,
+      "loss": 0.7894,
+      "step": 13874
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.035718014135998,
+      "learning_rate": 3.250460251775994e-07,
+      "loss": 0.6969,
+      "step": 13875
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7975244714661005,
+      "learning_rate": 3.246785126652563e-07,
+      "loss": 0.8432,
+      "step": 13876
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.171927074415016,
+      "learning_rate": 3.243112010618893e-07,
+      "loss": 0.9169,
+      "step": 13877
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1415064205857788,
+      "learning_rate": 3.239440903832852e-07,
+      "loss": 0.8229,
+      "step": 13878
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.410950484954864,
+      "learning_rate": 3.2357718064521594e-07,
+      "loss": 0.6391,
+      "step": 13879
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3956730474138097,
+      "learning_rate": 3.2321047186345047e-07,
+      "loss": 0.8722,
+      "step": 13880
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9656912070002264,
+      "learning_rate": 3.2284396405374787e-07,
+      "loss": 0.8322,
+      "step": 13881
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.448920834340466,
+      "learning_rate": 3.2247765723185556e-07,
+      "loss": 0.7967,
+      "step": 13882
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2614273869231747,
+      "learning_rate": 3.221115514135159e-07,
+      "loss": 0.8602,
+      "step": 13883
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0945963155916214,
+      "learning_rate": 3.217456466144614e-07,
+      "loss": 0.6803,
+      "step": 13884
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2479246686681345,
+      "learning_rate": 3.213799428504155e-07,
+      "loss": 0.7354,
+      "step": 13885
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0891691551265303,
+      "learning_rate": 3.210144401370935e-07,
+      "loss": 0.7928,
+      "step": 13886
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.193293821358984,
+      "learning_rate": 3.206491384902005e-07,
+      "loss": 0.8199,
+      "step": 13887
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8497939817998839,
+      "learning_rate": 3.202840379254374e-07,
+      "loss": 0.7955,
+      "step": 13888
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.618520236693685,
+      "learning_rate": 3.199191384584893e-07,
+      "loss": 0.6237,
+      "step": 13889
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2648288930113465,
+      "learning_rate": 3.1955444010504044e-07,
+      "loss": 0.7292,
+      "step": 13890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.608600273068545,
+      "learning_rate": 3.19189942880761e-07,
+      "loss": 0.8822,
+      "step": 13891
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2761400438061186,
+      "learning_rate": 3.18825646801314e-07,
+      "loss": 0.8075,
+      "step": 13892
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1828333982431603,
+      "learning_rate": 3.184615518823547e-07,
+      "loss": 0.8248,
+      "step": 13893
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9535472449066698,
+      "learning_rate": 3.180976581395295e-07,
+      "loss": 0.7594,
+      "step": 13894
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.30091318880168,
+      "learning_rate": 3.177339655884737e-07,
+      "loss": 0.9033,
+      "step": 13895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0856409397544728,
+      "learning_rate": 3.173704742448186e-07,
+      "loss": 0.6565,
+      "step": 13896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.147647227699981,
+      "learning_rate": 3.170071841241823e-07,
+      "loss": 0.8259,
+      "step": 13897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3551866217651987,
+      "learning_rate": 3.1664409524217677e-07,
+      "loss": 0.9075,
+      "step": 13898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9055761705247916,
+      "learning_rate": 3.1628120761440616e-07,
+      "loss": 0.5546,
+      "step": 13899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.339198002076789,
+      "learning_rate": 3.1591852125646196e-07,
+      "loss": 0.7827,
+      "step": 13900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.498035365142846,
+      "learning_rate": 3.1555603618393106e-07,
+      "loss": 0.7748,
+      "step": 13901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0117919928214356,
+      "learning_rate": 3.151937524123905e-07,
+      "loss": 0.6604,
+      "step": 13902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8724712396628558,
+      "learning_rate": 3.148316699574078e-07,
+      "loss": 0.8176,
+      "step": 13903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6929018192692915,
+      "learning_rate": 3.144697888345427e-07,
+      "loss": 0.6504,
+      "step": 13904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.72714384650188,
+      "learning_rate": 3.1410810905934564e-07,
+      "loss": 0.7303,
+      "step": 13905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1718295632826856,
+      "learning_rate": 3.137466306473602e-07,
+      "loss": 0.8307,
+      "step": 13906
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.489389746729111,
+      "learning_rate": 3.1338535361411737e-07,
+      "loss": 0.8477,
+      "step": 13907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8398161927997259,
+      "learning_rate": 3.1302427797514465e-07,
+      "loss": 0.8071,
+      "step": 13908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9417839291938486,
+      "learning_rate": 3.1266340374595693e-07,
+      "loss": 0.6688,
+      "step": 13909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4571410879725217,
+      "learning_rate": 3.1230273094206233e-07,
+      "loss": 0.6992,
+      "step": 13910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0266787283027476,
+      "learning_rate": 3.1194225957895895e-07,
+      "loss": 0.6093,
+      "step": 13911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7800986771944363,
+      "learning_rate": 3.115819896721378e-07,
+      "loss": 0.8339,
+      "step": 13912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.601288263506989,
+      "learning_rate": 3.1122192123707985e-07,
+      "loss": 0.9323,
+      "step": 13913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0405107587724385,
+      "learning_rate": 3.108620542892593e-07,
+      "loss": 0.6523,
+      "step": 13914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9860149273651857,
+      "learning_rate": 3.105023888441383e-07,
+      "loss": 0.8817,
+      "step": 13915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.458400482325686,
+      "learning_rate": 3.1014292491717444e-07,
+      "loss": 0.8375,
+      "step": 13916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8791127286178673,
+      "learning_rate": 3.0978366252381376e-07,
+      "loss": 0.8202,
+      "step": 13917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2761091357323817,
+      "learning_rate": 3.0942460167949495e-07,
+      "loss": 1.0003,
+      "step": 13918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0497191577253149,
+      "learning_rate": 3.0906574239964795e-07,
+      "loss": 0.5066,
+      "step": 13919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3113222692594118,
+      "learning_rate": 3.08707084699692e-07,
+      "loss": 0.8661,
+      "step": 13920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.278842481005865,
+      "learning_rate": 3.083486285950421e-07,
+      "loss": 0.9416,
+      "step": 13921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.023746348429317,
+      "learning_rate": 3.0799037410109976e-07,
+      "loss": 0.6257,
+      "step": 13922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.780628887971418,
+      "learning_rate": 3.076323212332605e-07,
+      "loss": 0.601,
+      "step": 13923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.418995523898367,
+      "learning_rate": 3.072744700069119e-07,
+      "loss": 0.7255,
+      "step": 13924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.011003408215872,
+      "learning_rate": 3.06916820437429e-07,
+      "loss": 0.8743,
+      "step": 13925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2296329743168584,
+      "learning_rate": 3.065593725401833e-07,
+      "loss": 0.8008,
+      "step": 13926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.251201258083949,
+      "learning_rate": 3.0620212633053424e-07,
+      "loss": 0.7375,
+      "step": 13927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1918836389627496,
+      "learning_rate": 3.0584508182383346e-07,
+      "loss": 0.7857,
+      "step": 13928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.022533414429424,
+      "learning_rate": 3.054882390354241e-07,
+      "loss": 0.5953,
+      "step": 13929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1182288091410428,
+      "learning_rate": 3.0513159798063906e-07,
+      "loss": 0.608,
+      "step": 13930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2776173469579124,
+      "learning_rate": 3.0477515867480655e-07,
+      "loss": 0.7742,
+      "step": 13931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4336683757983595,
+      "learning_rate": 3.0441892113324265e-07,
+      "loss": 0.9784,
+      "step": 13932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.905337563433695,
+      "learning_rate": 3.0406288537125403e-07,
+      "loss": 0.9672,
+      "step": 13933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2861923245564917,
+      "learning_rate": 3.0370705140414293e-07,
+      "loss": 0.706,
+      "step": 13934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8990897716973072,
+      "learning_rate": 3.0335141924719813e-07,
+      "loss": 0.7724,
+      "step": 13935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.020625270509504,
+      "learning_rate": 3.029959889157036e-07,
+      "loss": 0.9446,
+      "step": 13936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8411058412640722,
+      "learning_rate": 3.026407604249315e-07,
+      "loss": 0.6935,
+      "step": 13937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0144428270946806,
+      "learning_rate": 3.0228573379014745e-07,
+      "loss": 0.8372,
+      "step": 13938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.157710004473366,
+      "learning_rate": 3.019309090266087e-07,
+      "loss": 0.7865,
+      "step": 13939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0734430004303364,
+      "learning_rate": 3.015762861495608e-07,
+      "loss": 0.9176,
+      "step": 13940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0101321364119924,
+      "learning_rate": 3.0122186517424435e-07,
+      "loss": 0.6552,
+      "step": 13941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3301664646675437,
+      "learning_rate": 3.0086764611589047e-07,
+      "loss": 0.8562,
+      "step": 13942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2810687168961499,
+      "learning_rate": 3.0051362898971704e-07,
+      "loss": 0.6048,
+      "step": 13943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1419664173756845,
+      "learning_rate": 3.0015981381094073e-07,
+      "loss": 0.6042,
+      "step": 13944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9912235493039825,
+      "learning_rate": 2.9980620059476384e-07,
+      "loss": 0.7248,
+      "step": 13945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.929508797836272,
+      "learning_rate": 2.9945278935638255e-07,
+      "loss": 0.806,
+      "step": 13946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.091132629998741,
+      "learning_rate": 2.990995801109836e-07,
+      "loss": 0.7656,
+      "step": 13947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7619885117047633,
+      "learning_rate": 2.987465728737443e-07,
+      "loss": 0.616,
+      "step": 13948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7354271661134093,
+      "learning_rate": 2.9839376765983583e-07,
+      "loss": 0.8102,
+      "step": 13949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.433094171444925,
+      "learning_rate": 2.980411644844189e-07,
+      "loss": 0.7861,
+      "step": 13950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0340509666656017,
+      "learning_rate": 2.976887633626435e-07,
+      "loss": 0.6804,
+      "step": 13951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.943905790819303,
+      "learning_rate": 2.973365643096554e-07,
+      "loss": 0.6438,
+      "step": 13952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7725732236458707,
+      "learning_rate": 2.9698456734058856e-07,
+      "loss": 0.6582,
+      "step": 13953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6620257616640988,
+      "learning_rate": 2.9663277247056923e-07,
+      "loss": 0.8218,
+      "step": 13954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0002064870516962,
+      "learning_rate": 2.9628117971471416e-07,
+      "loss": 0.6384,
+      "step": 13955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3548362214918015,
+      "learning_rate": 2.959297890881319e-07,
+      "loss": 0.6001,
+      "step": 13956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5495851016037885,
+      "learning_rate": 2.955786006059247e-07,
+      "loss": 1.0414,
+      "step": 13957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5266667741522375,
+      "learning_rate": 2.952276142831806e-07,
+      "loss": 0.6493,
+      "step": 13958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9376702990396395,
+      "learning_rate": 2.9487683013498523e-07,
+      "loss": 0.6675,
+      "step": 13959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9432199930062355,
+      "learning_rate": 2.9452624817641163e-07,
+      "loss": 0.8581,
+      "step": 13960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.637508771448213,
+      "learning_rate": 2.9417586842252375e-07,
+      "loss": 0.6807,
+      "step": 13961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7194236314972071,
+      "learning_rate": 2.9382569088837966e-07,
+      "loss": 0.7926,
+      "step": 13962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2267977036219497,
+      "learning_rate": 2.934757155890272e-07,
+      "loss": 0.6992,
+      "step": 13963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7970399534593953,
+      "learning_rate": 2.93125942539505e-07,
+      "loss": 0.7442,
+      "step": 13964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0170405088642536,
+      "learning_rate": 2.9277637175484376e-07,
+      "loss": 0.7521,
+      "step": 13965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.145141796920394,
+      "learning_rate": 2.924270032500648e-07,
+      "loss": 0.9373,
+      "step": 13966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.894135835810812,
+      "learning_rate": 2.9207783704018223e-07,
+      "loss": 0.7715,
+      "step": 13967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.731085054372693,
+      "learning_rate": 2.9172887314020017e-07,
+      "loss": 0.8215,
+      "step": 13968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0574331044691023,
+      "learning_rate": 2.913801115651144e-07,
+      "loss": 1.0112,
+      "step": 13969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4611370515679036,
+      "learning_rate": 2.910315523299123e-07,
+      "loss": 1.0031,
+      "step": 13970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.377519433163404,
+      "learning_rate": 2.906831954495715e-07,
+      "loss": 0.7434,
+      "step": 13971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.22415375333219,
+      "learning_rate": 2.9033504093906207e-07,
+      "loss": 0.5912,
+      "step": 13972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8610556535885734,
+      "learning_rate": 2.8998708881334437e-07,
+      "loss": 0.9362,
+      "step": 13973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8579565293392502,
+      "learning_rate": 2.896393390873714e-07,
+      "loss": 0.894,
+      "step": 13974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9144111934630916,
+      "learning_rate": 2.892917917760873e-07,
+      "loss": 0.8093,
+      "step": 13975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.148787615173334,
+      "learning_rate": 2.8894444689442526e-07,
+      "loss": 0.8802,
+      "step": 13976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3747627583214475,
+      "learning_rate": 2.885973044573126e-07,
+      "loss": 0.6877,
+      "step": 13977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9334662179194335,
+      "learning_rate": 2.8825036447966745e-07,
+      "loss": 0.764,
+      "step": 13978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3131732307535686,
+      "learning_rate": 2.8790362697639685e-07,
+      "loss": 0.8044,
+      "step": 13979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0806178734694165,
+      "learning_rate": 2.8755709196240264e-07,
+      "loss": 0.9185,
+      "step": 13980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.682916238464043,
+      "learning_rate": 2.8721075945257414e-07,
+      "loss": 0.6735,
+      "step": 13981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.142727016507669,
+      "learning_rate": 2.868646294617966e-07,
+      "loss": 0.7701,
+      "step": 13982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.321104835538316,
+      "learning_rate": 2.86518702004942e-07,
+      "loss": 0.7578,
+      "step": 13983
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2214901220438548,
+      "learning_rate": 2.8617297709687577e-07,
+      "loss": 0.6192,
+      "step": 13984
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.425868311997281,
+      "learning_rate": 2.858274547524559e-07,
+      "loss": 0.6239,
+      "step": 13985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9745936622234568,
+      "learning_rate": 2.854821349865289e-07,
+      "loss": 0.8177,
+      "step": 13986
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2183352727898638,
+      "learning_rate": 2.8513701781393456e-07,
+      "loss": 0.9872,
+      "step": 13987
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1057788583463504,
+      "learning_rate": 2.8479210324950266e-07,
+      "loss": 0.703,
+      "step": 13988
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0903004497130828,
+      "learning_rate": 2.8444739130805587e-07,
+      "loss": 0.6045,
+      "step": 13989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7737896470336776,
+      "learning_rate": 2.841028820044067e-07,
+      "loss": 0.7029,
+      "step": 13990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7854991582591493,
+      "learning_rate": 2.837585753533589e-07,
+      "loss": 0.7794,
+      "step": 13991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.146068475905639,
+      "learning_rate": 2.8341447136970954e-07,
+      "loss": 0.9227,
+      "step": 13992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.119176967883836,
+      "learning_rate": 2.8307057006824514e-07,
+      "loss": 0.6651,
+      "step": 13993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8943984264817852,
+      "learning_rate": 2.827268714637421e-07,
+      "loss": 0.7201,
+      "step": 13994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.098976921397279,
+      "learning_rate": 2.823833755709721e-07,
+      "loss": 0.9143,
+      "step": 13995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.933749552785001,
+      "learning_rate": 2.820400824046954e-07,
+      "loss": 0.8638,
+      "step": 13996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.161940175915366,
+      "learning_rate": 2.816969919796636e-07,
+      "loss": 0.6545,
+      "step": 13997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9935281606667228,
+      "learning_rate": 2.813541043106205e-07,
+      "loss": 0.6215,
+      "step": 13998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1101708518983795,
+      "learning_rate": 2.810114194122998e-07,
+      "loss": 0.8271,
+      "step": 13999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1896192632588223,
+      "learning_rate": 2.806689372994292e-07,
+      "loss": 0.8332,
+      "step": 14000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.149927146635097,
+      "learning_rate": 2.803266579867242e-07,
+      "loss": 0.7645,
+      "step": 14001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.067750272601552,
+      "learning_rate": 2.7998458148889354e-07,
+      "loss": 0.7893,
+      "step": 14002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6956327624202212,
+      "learning_rate": 2.796427078206382e-07,
+      "loss": 0.8909,
+      "step": 14003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.220547162975445,
+      "learning_rate": 2.793010369966487e-07,
+      "loss": 0.8459,
+      "step": 14004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.875417014075634,
+      "learning_rate": 2.789595690316066e-07,
+      "loss": 0.7848,
+      "step": 14005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8618333504614548,
+      "learning_rate": 2.7861830394018627e-07,
+      "loss": 0.9906,
+      "step": 14006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1607684852930342,
+      "learning_rate": 2.7827724173705273e-07,
+      "loss": 0.8189,
+      "step": 14007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5238267663673013,
+      "learning_rate": 2.779363824368619e-07,
+      "loss": 0.9383,
+      "step": 14008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.482584534517959,
+      "learning_rate": 2.7759572605426057e-07,
+      "loss": 0.8528,
+      "step": 14009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5539837880309295,
+      "learning_rate": 2.7725527260388905e-07,
+      "loss": 0.7223,
+      "step": 14010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3715035257014856,
+      "learning_rate": 2.7691502210037744e-07,
+      "loss": 0.7866,
+      "step": 14011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.516453697226107,
+      "learning_rate": 2.7657497455834457e-07,
+      "loss": 0.6509,
+      "step": 14012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.12690074722468,
+      "learning_rate": 2.762351299924054e-07,
+      "loss": 0.8962,
+      "step": 14013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5719888916798483,
+      "learning_rate": 2.7589548841716274e-07,
+      "loss": 0.8012,
+      "step": 14014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5621693827831518,
+      "learning_rate": 2.755560498472126e-07,
+      "loss": 0.8278,
+      "step": 14015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.969514934284264,
+      "learning_rate": 2.752168142971406e-07,
+      "loss": 0.6726,
+      "step": 14016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.113123839700502,
+      "learning_rate": 2.7487778178152457e-07,
+      "loss": 0.6952,
+      "step": 14017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2673719867032163,
+      "learning_rate": 2.7453895231493445e-07,
+      "loss": 0.5535,
+      "step": 14018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3721763855719975,
+      "learning_rate": 2.7420032591192856e-07,
+      "loss": 0.9654,
+      "step": 14019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0961604693523643,
+      "learning_rate": 2.7386190258706033e-07,
+      "loss": 0.7537,
+      "step": 14020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.067403662780701,
+      "learning_rate": 2.735236823548715e-07,
+      "loss": 0.8326,
+      "step": 14021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.576959444052205,
+      "learning_rate": 2.73185665229897e-07,
+      "loss": 0.8755,
+      "step": 14022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1988529577448708,
+      "learning_rate": 2.7284785122666136e-07,
+      "loss": 0.6035,
+      "step": 14023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.829277163931294,
+      "learning_rate": 2.7251024035968134e-07,
+      "loss": 0.7005,
+      "step": 14024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.117778441075035,
+      "learning_rate": 2.721728326434653e-07,
+      "loss": 0.7558,
+      "step": 14025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3889588089071534,
+      "learning_rate": 2.718356280925116e-07,
+      "loss": 0.6366,
+      "step": 14026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.236786519947908,
+      "learning_rate": 2.71498626721311e-07,
+      "loss": 0.8755,
+      "step": 14027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3487913958058466,
+      "learning_rate": 2.711618285443457e-07,
+      "loss": 0.8957,
+      "step": 14028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1826492741449666,
+      "learning_rate": 2.7082523357608856e-07,
+      "loss": 0.6926,
+      "step": 14029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0319725061647587,
+      "learning_rate": 2.7048884183100313e-07,
+      "loss": 0.8827,
+      "step": 14030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1139515004497482,
+      "learning_rate": 2.70152653323546e-07,
+      "loss": 0.6368,
+      "step": 14031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.308496655139665,
+      "learning_rate": 2.6981666806816185e-07,
+      "loss": 0.7195,
+      "step": 14032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.234901526504392,
+      "learning_rate": 2.694808860792919e-07,
+      "loss": 0.8462,
+      "step": 14033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.296823359916676,
+      "learning_rate": 2.6914530737136346e-07,
+      "loss": 0.8489,
+      "step": 14034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.080231986145147,
+      "learning_rate": 2.6880993195879614e-07,
+      "loss": 0.7387,
+      "step": 14035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.449014218360392,
+      "learning_rate": 2.684747598560045e-07,
+      "loss": 0.7136,
+      "step": 14036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2882184875171272,
+      "learning_rate": 2.6813979107738864e-07,
+      "loss": 0.5232,
+      "step": 14037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0576174684774755,
+      "learning_rate": 2.678050256373449e-07,
+      "loss": 0.7246,
+      "step": 14038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9901779641469459,
+      "learning_rate": 2.674704635502584e-07,
+      "loss": 0.7132,
+      "step": 14039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.216807258890418,
+      "learning_rate": 2.6713610483050654e-07,
+      "loss": 0.8203,
+      "step": 14040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1425075401241516,
+      "learning_rate": 2.668019494924562e-07,
+      "loss": 0.6686,
+      "step": 14041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.366460801295164,
+      "learning_rate": 2.6646799755046746e-07,
+      "loss": 0.9123,
+      "step": 14042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.257100456709912,
+      "learning_rate": 2.661342490188917e-07,
+      "loss": 0.6397,
+      "step": 14043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.276867586556014,
+      "learning_rate": 2.658007039120697e-07,
+      "loss": 0.8981,
+      "step": 14044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.045368553535986,
+      "learning_rate": 2.654673622443349e-07,
+      "loss": 0.5533,
+      "step": 14045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9922783563828494,
+      "learning_rate": 2.65134224030012e-07,
+      "loss": 0.9366,
+      "step": 14046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3216637352447127,
+      "learning_rate": 2.6480128928341675e-07,
+      "loss": 0.8122,
+      "step": 14047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.916528714226619,
+      "learning_rate": 2.6446855801885606e-07,
+      "loss": 0.7089,
+      "step": 14048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.049770485722744,
+      "learning_rate": 2.64136030250628e-07,
+      "loss": 0.6456,
+      "step": 14049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.11569934102198,
+      "learning_rate": 2.638037059930215e-07,
+      "loss": 0.617,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2597052317212056,
+      "learning_rate": 2.634715852603187e-07,
+      "loss": 0.9623,
+      "step": 14051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.30998685740304,
+      "learning_rate": 2.631396680667908e-07,
+      "loss": 0.6974,
+      "step": 14052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2702193255326502,
+      "learning_rate": 2.6280795442669925e-07,
+      "loss": 0.7241,
+      "step": 14053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6142174022600555,
+      "learning_rate": 2.6247644435430263e-07,
+      "loss": 0.8286,
+      "step": 14054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.351819782344634,
+      "learning_rate": 2.621451378638418e-07,
+      "loss": 0.8873,
+      "step": 14055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2783510270874046,
+      "learning_rate": 2.618140349695575e-07,
+      "loss": 0.8455,
+      "step": 14056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3701405435142258,
+      "learning_rate": 2.6148313568567674e-07,
+      "loss": 0.788,
+      "step": 14057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.02553839837445,
+      "learning_rate": 2.611524400264187e-07,
+      "loss": 0.7206,
+      "step": 14058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4649391173667943,
+      "learning_rate": 2.6082194800599424e-07,
+      "loss": 0.8392,
+      "step": 14059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.041545429880724,
+      "learning_rate": 2.6049165963860425e-07,
+      "loss": 0.7192,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.10834125866615,
+      "learning_rate": 2.601615749384451e-07,
+      "loss": 0.6371,
+      "step": 14061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2483721851852554,
+      "learning_rate": 2.5983169391969823e-07,
+      "loss": 0.7893,
+      "step": 14062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4809120922960144,
+      "learning_rate": 2.595020165965401e-07,
+      "loss": 0.6237,
+      "step": 14063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.277052065580937,
+      "learning_rate": 2.591725429831382e-07,
+      "loss": 0.9447,
+      "step": 14064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.70258400237596,
+      "learning_rate": 2.5884327309365074e-07,
+      "loss": 0.7537,
+      "step": 14065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3123510074123634,
+      "learning_rate": 2.5851420694222697e-07,
+      "loss": 0.5228,
+      "step": 14066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0953212279433417,
+      "learning_rate": 2.581853445430077e-07,
+      "loss": 0.7058,
+      "step": 14067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2878012913086745,
+      "learning_rate": 2.5785668591012393e-07,
+      "loss": 1.1124,
+      "step": 14068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1094847919350728,
+      "learning_rate": 2.57528231057701e-07,
+      "loss": 0.7363,
+      "step": 14069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0925119813929907,
+      "learning_rate": 2.571999799998509e-07,
+      "loss": 0.7058,
+      "step": 14070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0565332836355044,
+      "learning_rate": 2.568719327506808e-07,
+      "loss": 0.8427,
+      "step": 14071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.120739773441538,
+      "learning_rate": 2.565440893242882e-07,
+      "loss": 0.7144,
+      "step": 14072
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0062703038281424,
+      "learning_rate": 2.5621644973475914e-07,
+      "loss": 0.6944,
+      "step": 14073
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6373474966893298,
+      "learning_rate": 2.558890139961745e-07,
+      "loss": 0.6865,
+      "step": 14074
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1946999721783365,
+      "learning_rate": 2.5556178212260474e-07,
+      "loss": 0.7997,
+      "step": 14075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1131118631364163,
+      "learning_rate": 2.552347541281114e-07,
+      "loss": 0.6219,
+      "step": 14076
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9829539890447151,
+      "learning_rate": 2.549079300267482e-07,
+      "loss": 0.6237,
+      "step": 14077
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2472945184821786,
+      "learning_rate": 2.5458130983255835e-07,
+      "loss": 0.9522,
+      "step": 14078
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2427697530204012,
+      "learning_rate": 2.5425489355957956e-07,
+      "loss": 0.8004,
+      "step": 14079
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0935980938346215,
+      "learning_rate": 2.539286812218361e-07,
+      "loss": 0.8148,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1533144600449161,
+      "learning_rate": 2.5360267283334796e-07,
+      "loss": 0.7388,
+      "step": 14081
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7041909851796546,
+      "learning_rate": 2.5327686840812326e-07,
+      "loss": 0.729,
+      "step": 14082
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.303915578398551,
+      "learning_rate": 2.5295126796016366e-07,
+      "loss": 0.7606,
+      "step": 14083
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2131750976824818,
+      "learning_rate": 2.526258715034602e-07,
+      "loss": 0.6997,
+      "step": 14084
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.243008723636838,
+      "learning_rate": 2.523006790519955e-07,
+      "loss": 0.7661,
+      "step": 14085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.801004594062984,
+      "learning_rate": 2.5197569061974404e-07,
+      "loss": 0.7769,
+      "step": 14086
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3510865383286292,
+      "learning_rate": 2.5165090622067343e-07,
+      "loss": 0.9139,
+      "step": 14087
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4030000807351746,
+      "learning_rate": 2.513263258687365e-07,
+      "loss": 0.6793,
+      "step": 14088
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.310807147203609,
+      "learning_rate": 2.510019495778837e-07,
+      "loss": 0.8724,
+      "step": 14089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.282577451740699,
+      "learning_rate": 2.5067777736205444e-07,
+      "loss": 0.7416,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5960655467138056,
+      "learning_rate": 2.503538092351782e-07,
+      "loss": 0.7566,
+      "step": 14091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9515815487597008,
+      "learning_rate": 2.5003004521117646e-07,
+      "loss": 0.9095,
+      "step": 14092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4782242646669896,
+      "learning_rate": 2.4970648530396157e-07,
+      "loss": 0.7711,
+      "step": 14093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.415183576423074,
+      "learning_rate": 2.4938312952744016e-07,
+      "loss": 0.5721,
+      "step": 14094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2384583208803006,
+      "learning_rate": 2.490599778955055e-07,
+      "loss": 0.6574,
+      "step": 14095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2148688776622616,
+      "learning_rate": 2.4873703042204323e-07,
+      "loss": 1.0489,
+      "step": 14096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.064324581679482,
+      "learning_rate": 2.484142871209344e-07,
+      "loss": 0.8818,
+      "step": 14097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9942333275065877,
+      "learning_rate": 2.480917480060441e-07,
+      "loss": 0.5822,
+      "step": 14098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1357275540677072,
+      "learning_rate": 2.477694130912356e-07,
+      "loss": 0.5892,
+      "step": 14099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4980985140076903,
+      "learning_rate": 2.474472823903584e-07,
+      "loss": 0.8306,
+      "step": 14100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.082600880791511,
+      "learning_rate": 2.4712535591725706e-07,
+      "loss": 0.8762,
+      "step": 14101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.004608412602767,
+      "learning_rate": 2.4680363368576376e-07,
+      "loss": 0.6812,
+      "step": 14102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.790894826576746,
+      "learning_rate": 2.464821157097036e-07,
+      "loss": 0.7989,
+      "step": 14103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.548723644763108,
+      "learning_rate": 2.461608020028944e-07,
+      "loss": 0.7855,
+      "step": 14104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3471191847286208,
+      "learning_rate": 2.458396925791434e-07,
+      "loss": 0.5613,
+      "step": 14105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8846880819643552,
+      "learning_rate": 2.4551878745224745e-07,
+      "loss": 0.8188,
+      "step": 14106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.096915207301633,
+      "learning_rate": 2.4519808663599876e-07,
+      "loss": 0.8932,
+      "step": 14107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.137533594491639,
+      "learning_rate": 2.44877590144178e-07,
+      "loss": 0.8001,
+      "step": 14108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.386543118060049,
+      "learning_rate": 2.445572979905575e-07,
+      "loss": 0.8925,
+      "step": 14109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1899116228304836,
+      "learning_rate": 2.442372101889007e-07,
+      "loss": 0.8751,
+      "step": 14110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.287783368774466,
+      "learning_rate": 2.4391732675296207e-07,
+      "loss": 0.8142,
+      "step": 14111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9914682189064199,
+      "learning_rate": 2.4359764769648907e-07,
+      "loss": 0.5782,
+      "step": 14112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.855208795846365,
+      "learning_rate": 2.4327817303321786e-07,
+      "loss": 0.8888,
+      "step": 14113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.142137142277678,
+      "learning_rate": 2.4295890277687695e-07,
+      "loss": 0.7758,
+      "step": 14114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.316348278655279,
+      "learning_rate": 2.4263983694118754e-07,
+      "loss": 0.6731,
+      "step": 14115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5570458801792704,
+      "learning_rate": 2.423209755398587e-07,
+      "loss": 0.7162,
+      "step": 14116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1278766904083404,
+      "learning_rate": 2.4200231858659384e-07,
+      "loss": 0.8627,
+      "step": 14117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.475794823259438,
+      "learning_rate": 2.416838660950854e-07,
+      "loss": 0.7113,
+      "step": 14118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4677582461152205,
+      "learning_rate": 2.4136561807901916e-07,
+      "loss": 0.7987,
+      "step": 14119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.16965240939339,
+      "learning_rate": 2.410475745520702e-07,
+      "loss": 0.863,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8394537093672205,
+      "learning_rate": 2.4072973552790535e-07,
+      "loss": 0.8688,
+      "step": 14121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3726276022004074,
+      "learning_rate": 2.4041210102018375e-07,
+      "loss": 0.663,
+      "step": 14122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9116918206702214,
+      "learning_rate": 2.40094671042555e-07,
+      "loss": 0.9193,
+      "step": 14123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2818148336115738,
+      "learning_rate": 2.397774456086577e-07,
+      "loss": 0.6826,
+      "step": 14124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.484383982144624,
+      "learning_rate": 2.3946042473212526e-07,
+      "loss": 0.8572,
+      "step": 14125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.263903327257752,
+      "learning_rate": 2.391436084265814e-07,
+      "loss": 0.8029,
+      "step": 14126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0046154856574967,
+      "learning_rate": 2.3882699670563947e-07,
+      "loss": 0.7266,
+      "step": 14127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1074169440962187,
+      "learning_rate": 2.385105895829054e-07,
+      "loss": 0.9665,
+      "step": 14128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.5318464341746205,
+      "learning_rate": 2.3819438707197495e-07,
+      "loss": 0.8896,
+      "step": 14129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3183434582670905,
+      "learning_rate": 2.378783891864378e-07,
+      "loss": 0.8921,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9873486323692264,
+      "learning_rate": 2.3756259593987086e-07,
+      "loss": 1.0316,
+      "step": 14131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.130002366663911,
+      "learning_rate": 2.372470073458466e-07,
+      "loss": 0.7239,
+      "step": 14132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8158528659137647,
+      "learning_rate": 2.3693162341792532e-07,
+      "loss": 0.7024,
+      "step": 14133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.247636797595876,
+      "learning_rate": 2.3661644416966057e-07,
+      "loss": 0.9546,
+      "step": 14134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2163609331924454,
+      "learning_rate": 2.3630146961459543e-07,
+      "loss": 0.6993,
+      "step": 14135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1951946322826816,
+      "learning_rate": 2.3598669976626576e-07,
+      "loss": 0.6651,
+      "step": 14136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.156084956020802,
+      "learning_rate": 2.3567213463819738e-07,
+      "loss": 0.7315,
+      "step": 14137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.491173407866139,
+      "learning_rate": 2.353577742439084e-07,
+      "loss": 0.8391,
+      "step": 14138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.485565824406001,
+      "learning_rate": 2.3504361859690628e-07,
+      "loss": 0.8191,
+      "step": 14139
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.284525235048342,
+      "learning_rate": 2.347296677106925e-07,
+      "loss": 0.7888,
+      "step": 14140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5956921681617464,
+      "learning_rate": 2.344159215987585e-07,
+      "loss": 0.8657,
+      "step": 14141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4941203088155737,
+      "learning_rate": 2.341023802745851e-07,
+      "loss": 0.9823,
+      "step": 14142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.213462973960131,
+      "learning_rate": 2.3378904375164714e-07,
+      "loss": 0.8304,
+      "step": 14143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4123220966286913,
+      "learning_rate": 2.3347591204340881e-07,
+      "loss": 0.8977,
+      "step": 14144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.239962921832974,
+      "learning_rate": 2.3316298516332604e-07,
+      "loss": 0.8223,
+      "step": 14145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8589266991246576,
+      "learning_rate": 2.3285026312484583e-07,
+      "loss": 0.9424,
+      "step": 14146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4959773443210356,
+      "learning_rate": 2.3253774594140633e-07,
+      "loss": 0.7189,
+      "step": 14147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3471873599660573,
+      "learning_rate": 2.32225433626439e-07,
+      "loss": 0.9084,
+      "step": 14148
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4435952135591443,
+      "learning_rate": 2.3191332619336204e-07,
+      "loss": 0.7959,
+      "step": 14149
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8257184280572596,
+      "learning_rate": 2.3160142365558914e-07,
+      "loss": 0.7092,
+      "step": 14150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4008063531360033,
+      "learning_rate": 2.312897260265229e-07,
+      "loss": 0.8406,
+      "step": 14151
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4941270604549146,
+      "learning_rate": 2.309782333195576e-07,
+      "loss": 0.8977,
+      "step": 14152
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8267882967311535,
+      "learning_rate": 2.3066694554807867e-07,
+      "loss": 0.7663,
+      "step": 14153
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8035796497589351,
+      "learning_rate": 2.3035586272546207e-07,
+      "loss": 0.6633,
+      "step": 14154
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.058418431079745,
+      "learning_rate": 2.300449848650782e-07,
+      "loss": 0.6188,
+      "step": 14155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3756237858926466,
+      "learning_rate": 2.2973431198028417e-07,
+      "loss": 0.7489,
+      "step": 14156
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1001390934877167,
+      "learning_rate": 2.2942384408442986e-07,
+      "loss": 0.6556,
+      "step": 14157
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.218078394870811,
+      "learning_rate": 2.2911358119085792e-07,
+      "loss": 0.7788,
+      "step": 14158
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0626927786033913,
+      "learning_rate": 2.2880352331290102e-07,
+      "loss": 0.8188,
+      "step": 14159
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5702532411113066,
+      "learning_rate": 2.2849367046388294e-07,
+      "loss": 0.6404,
+      "step": 14160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.17136650233924,
+      "learning_rate": 2.2818402265711858e-07,
+      "loss": 0.8075,
+      "step": 14161
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.139831183069008,
+      "learning_rate": 2.27874579905914e-07,
+      "loss": 0.7989,
+      "step": 14162
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9759061200453742,
+      "learning_rate": 2.2756534222356687e-07,
+      "loss": 0.6151,
+      "step": 14163
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.800813924254703,
+      "learning_rate": 2.2725630962336542e-07,
+      "loss": 0.7799,
+      "step": 14164
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0065395433745747,
+      "learning_rate": 2.2694748211858964e-07,
+      "loss": 0.735,
+      "step": 14165
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1467021792165966,
+      "learning_rate": 2.2663885972251165e-07,
+      "loss": 0.918,
+      "step": 14166
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.825453843320932,
+      "learning_rate": 2.2633044244839143e-07,
+      "loss": 0.8285,
+      "step": 14167
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.224059864250879,
+      "learning_rate": 2.2602223030948445e-07,
+      "loss": 0.7788,
+      "step": 14168
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6284231242565297,
+      "learning_rate": 2.2571422331903458e-07,
+      "loss": 0.7334,
+      "step": 14169
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.598926799808844,
+      "learning_rate": 2.2540642149027735e-07,
+      "loss": 0.9162,
+      "step": 14170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.189982609171577,
+      "learning_rate": 2.2509882483643942e-07,
+      "loss": 0.691,
+      "step": 14171
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.0456188368335875,
+      "learning_rate": 2.2479143337073904e-07,
+      "loss": 0.7903,
+      "step": 14172
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.916225807502771,
+      "learning_rate": 2.2448424710638684e-07,
+      "loss": 0.7999,
+      "step": 14173
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9453831431607536,
+      "learning_rate": 2.2417726605658164e-07,
+      "loss": 0.8566,
+      "step": 14174
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.693665243859885,
+      "learning_rate": 2.2387049023451458e-07,
+      "loss": 0.7198,
+      "step": 14175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.198155518493491,
+      "learning_rate": 2.2356391965337064e-07,
+      "loss": 0.8796,
+      "step": 14176
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6529573038053975,
+      "learning_rate": 2.2325755432632267e-07,
+      "loss": 0.7109,
+      "step": 14177
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9550446584596495,
+      "learning_rate": 2.2295139426653622e-07,
+      "loss": 0.5719,
+      "step": 14178
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0976545432781424,
+      "learning_rate": 2.226454394871669e-07,
+      "loss": 0.7832,
+      "step": 14179
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8485457528042515,
+      "learning_rate": 2.2233969000136357e-07,
+      "loss": 0.8935,
+      "step": 14180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2907115233287128,
+      "learning_rate": 2.2203414582226357e-07,
+      "loss": 0.7,
+      "step": 14181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0373902243370865,
+      "learning_rate": 2.2172880696299692e-07,
+      "loss": 0.6534,
+      "step": 14182
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.418414411280986,
+      "learning_rate": 2.2142367343668592e-07,
+      "loss": 1.0537,
+      "step": 14183
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.154595869286078,
+      "learning_rate": 2.2111874525644228e-07,
+      "loss": 0.7272,
+      "step": 14184
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8319313857307769,
+      "learning_rate": 2.2081402243536833e-07,
+      "loss": 0.7082,
+      "step": 14185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.383912552979191,
+      "learning_rate": 2.2050950498656078e-07,
+      "loss": 0.8998,
+      "step": 14186
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8607424594288944,
+      "learning_rate": 2.2020519292310361e-07,
+      "loss": 0.7776,
+      "step": 14187
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3490554717617687,
+      "learning_rate": 2.199010862580747e-07,
+      "loss": 0.8817,
+      "step": 14188
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5523996305670145,
+      "learning_rate": 2.1959718500454196e-07,
+      "loss": 0.7591,
+      "step": 14189
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8823682049701724,
+      "learning_rate": 2.1929348917556436e-07,
+      "loss": 0.8065,
+      "step": 14190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3108913536144886,
+      "learning_rate": 2.189899987841937e-07,
+      "loss": 0.7949,
+      "step": 14191
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7636464421041964,
+      "learning_rate": 2.186867138434695e-07,
+      "loss": 0.992,
+      "step": 14192
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9467758130809583,
+      "learning_rate": 2.183836343664264e-07,
+      "loss": 0.8212,
+      "step": 14193
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.072735626813248,
+      "learning_rate": 2.1808076036608783e-07,
+      "loss": 0.7584,
+      "step": 14194
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9167670977245477,
+      "learning_rate": 2.1777809185546894e-07,
+      "loss": 0.8048,
+      "step": 14195
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.232426627203466,
+      "learning_rate": 2.17475628847576e-07,
+      "loss": 0.7976,
+      "step": 14196
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.150034322865171,
+      "learning_rate": 2.171733713554064e-07,
+      "loss": 0.9316,
+      "step": 14197
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5844450000736847,
+      "learning_rate": 2.1687131939194862e-07,
+      "loss": 0.7391,
+      "step": 14198
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6079837640449606,
+      "learning_rate": 2.165694729701834e-07,
+      "loss": 0.77,
+      "step": 14199
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.615267442156334,
+      "learning_rate": 2.1626783210308032e-07,
+      "loss": 0.901,
+      "step": 14200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.092659055352766,
+      "learning_rate": 2.15966396803603e-07,
+      "loss": 0.7728,
+      "step": 14201
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.484012634595588,
+      "learning_rate": 2.1566516708470487e-07,
+      "loss": 0.9379,
+      "step": 14202
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0435493243131915,
+      "learning_rate": 2.1536414295932896e-07,
+      "loss": 0.7746,
+      "step": 14203
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.016318325821326,
+      "learning_rate": 2.1506332444041212e-07,
+      "loss": 0.6448,
+      "step": 14204
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.685815439493679,
+      "learning_rate": 2.1476271154088013e-07,
+      "loss": 0.7946,
+      "step": 14205
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.063090791728641,
+      "learning_rate": 2.1446230427365323e-07,
+      "loss": 0.6237,
+      "step": 14206
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.155175852831767,
+      "learning_rate": 2.1416210265163828e-07,
+      "loss": 0.6258,
+      "step": 14207
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1284864024376704,
+      "learning_rate": 2.1386210668773554e-07,
+      "loss": 1.008,
+      "step": 14208
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8376611455395615,
+      "learning_rate": 2.1356231639483917e-07,
+      "loss": 0.7816,
+      "step": 14209
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.120050340346337,
+      "learning_rate": 2.1326273178582822e-07,
+      "loss": 0.849,
+      "step": 14210
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5600003375596376,
+      "learning_rate": 2.1296335287357918e-07,
+      "loss": 0.9006,
+      "step": 14211
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0645784280645234,
+      "learning_rate": 2.1266417967095665e-07,
+      "loss": 0.6057,
+      "step": 14212
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9028795585091336,
+      "learning_rate": 2.123652121908154e-07,
+      "loss": 0.7262,
+      "step": 14213
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.876819248949884,
+      "learning_rate": 2.1206645044600404e-07,
+      "loss": 0.8527,
+      "step": 14214
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.934797257398539,
+      "learning_rate": 2.1176789444936065e-07,
+      "loss": 0.692,
+      "step": 14215
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.084778305949087,
+      "learning_rate": 2.114695442137149e-07,
+      "loss": 0.6544,
+      "step": 14216
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.513308456661493,
+      "learning_rate": 2.1117139975188716e-07,
+      "loss": 0.7821,
+      "step": 14217
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3058852180187666,
+      "learning_rate": 2.1087346107668937e-07,
+      "loss": 1.0516,
+      "step": 14218
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9951234535938904,
+      "learning_rate": 2.1057572820092576e-07,
+      "loss": 0.6204,
+      "step": 14219
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2007882834148735,
+      "learning_rate": 2.1027820113738884e-07,
+      "loss": 0.8437,
+      "step": 14220
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.308244251009387,
+      "learning_rate": 2.099808798988656e-07,
+      "loss": 0.8832,
+      "step": 14221
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1534646419445354,
+      "learning_rate": 2.0968376449813198e-07,
+      "loss": 0.9772,
+      "step": 14222
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2755523806396165,
+      "learning_rate": 2.093868549479544e-07,
+      "loss": 0.8761,
+      "step": 14223
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7151464204558324,
+      "learning_rate": 2.0909015126109488e-07,
+      "loss": 0.8187,
+      "step": 14224
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4951743359639567,
+      "learning_rate": 2.0879365345030044e-07,
+      "loss": 0.7417,
+      "step": 14225
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0123799059732892,
+      "learning_rate": 2.0849736152831256e-07,
+      "loss": 0.6379,
+      "step": 14226
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.518920050037201,
+      "learning_rate": 2.0820127550786551e-07,
+      "loss": 0.7161,
+      "step": 14227
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8011178607275897,
+      "learning_rate": 2.0790539540168074e-07,
+      "loss": 0.6945,
+      "step": 14228
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9541648885280878,
+      "learning_rate": 2.0760972122247425e-07,
+      "loss": 0.7581,
+      "step": 14229
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5775280995367424,
+      "learning_rate": 2.0731425298295083e-07,
+      "loss": 0.8492,
+      "step": 14230
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8700988354744796,
+      "learning_rate": 2.070189906958081e-07,
+      "loss": 0.7451,
+      "step": 14231
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2096754199157913,
+      "learning_rate": 2.0672393437373372e-07,
+      "loss": 0.8276,
+      "step": 14232
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9204568863022164,
+      "learning_rate": 2.06429084029407e-07,
+      "loss": 0.6512,
+      "step": 14233
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1993118304368835,
+      "learning_rate": 2.061344396754994e-07,
+      "loss": 0.8457,
+      "step": 14234
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.939467109309061,
+      "learning_rate": 2.058400013246703e-07,
+      "loss": 1.0817,
+      "step": 14235
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.228450373093403,
+      "learning_rate": 2.0554576898957346e-07,
+      "loss": 0.8985,
+      "step": 14236
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0901431531671477,
+      "learning_rate": 2.052517426828532e-07,
+      "loss": 0.63,
+      "step": 14237
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1089547138954037,
+      "learning_rate": 2.0495792241714386e-07,
+      "loss": 0.6084,
+      "step": 14238
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8675171748441257,
+      "learning_rate": 2.04664308205072e-07,
+      "loss": 0.8476,
+      "step": 14239
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.664312778668487,
+      "learning_rate": 2.0437090005925475e-07,
+      "loss": 0.7842,
+      "step": 14240
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2957085246759967,
+      "learning_rate": 2.0407769799229925e-07,
+      "loss": 0.6966,
+      "step": 14241
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7123330474880043,
+      "learning_rate": 2.0378470201680767e-07,
+      "loss": 0.8087,
+      "step": 14242
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4277012789010937,
+      "learning_rate": 2.0349191214536767e-07,
+      "loss": 0.808,
+      "step": 14243
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0619453305462925,
+      "learning_rate": 2.0319932839056365e-07,
+      "loss": 0.7557,
+      "step": 14244
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3331282757893814,
+      "learning_rate": 2.029069507649678e-07,
+      "loss": 0.7931,
+      "step": 14245
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8958426027579809,
+      "learning_rate": 2.0261477928114282e-07,
+      "loss": 0.8481,
+      "step": 14246
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.005452001184977,
+      "learning_rate": 2.0232281395164587e-07,
+      "loss": 0.756,
+      "step": 14247
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7368624785087223,
+      "learning_rate": 2.0203105478902252e-07,
+      "loss": 0.6467,
+      "step": 14248
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3059577103704978,
+      "learning_rate": 2.0173950180581047e-07,
+      "loss": 0.6508,
+      "step": 14249
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2395710271114382,
+      "learning_rate": 2.0144815501453863e-07,
+      "loss": 0.7154,
+      "step": 14250
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0954429354766924,
+      "learning_rate": 2.0115701442772528e-07,
+      "loss": 0.6564,
+      "step": 14251
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.619818141312856,
+      "learning_rate": 2.0086608005788376e-07,
+      "loss": 0.8194,
+      "step": 14252
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.542458735329269,
+      "learning_rate": 2.0057535191751465e-07,
+      "loss": 0.7549,
+      "step": 14253
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6785066361962264,
+      "learning_rate": 2.002848300191118e-07,
+      "loss": 0.6195,
+      "step": 14254
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1042397503688615,
+      "learning_rate": 1.9999451437515915e-07,
+      "loss": 0.6,
+      "step": 14255
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3046503042663207,
+      "learning_rate": 1.997044049981328e-07,
+      "loss": 0.8709,
+      "step": 14256
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2073698863546345,
+      "learning_rate": 1.994145019004984e-07,
+      "loss": 0.8762,
+      "step": 14257
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.925397683609044,
+      "learning_rate": 1.9912480509471476e-07,
+      "loss": 0.9445,
+      "step": 14258
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.128611788503351,
+      "learning_rate": 1.988353145932298e-07,
+      "loss": 0.7592,
+      "step": 14259
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1860072527721084,
+      "learning_rate": 1.985460304084852e-07,
+      "loss": 0.6931,
+      "step": 14260
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.589659356281897,
+      "learning_rate": 1.9825695255290934e-07,
+      "loss": 1.0164,
+      "step": 14261
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0904096149853415,
+      "learning_rate": 1.9796808103892783e-07,
+      "loss": 0.7007,
+      "step": 14262
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2923804053355936,
+      "learning_rate": 1.9767941587895246e-07,
+      "loss": 0.7745,
+      "step": 14263
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0366032080164898,
+      "learning_rate": 1.9739095708538714e-07,
+      "loss": 0.7168,
+      "step": 14264
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4977984035332494,
+      "learning_rate": 1.9710270467062865e-07,
+      "loss": 0.798,
+      "step": 14265
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.000030978937679,
+      "learning_rate": 1.9681465864706372e-07,
+      "loss": 0.9058,
+      "step": 14266
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2042653279782587,
+      "learning_rate": 1.9652681902707028e-07,
+      "loss": 0.6215,
+      "step": 14267
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4031997390588902,
+      "learning_rate": 1.9623918582301726e-07,
+      "loss": 0.7711,
+      "step": 14268
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4745750152458865,
+      "learning_rate": 1.9595175904726481e-07,
+      "loss": 0.7231,
+      "step": 14269
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0399220601396313,
+      "learning_rate": 1.9566453871216417e-07,
+      "loss": 0.8337,
+      "step": 14270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0971030206431234,
+      "learning_rate": 1.953775248300588e-07,
+      "loss": 0.7308,
+      "step": 14271
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0467446549156407,
+      "learning_rate": 1.9509071741328155e-07,
+      "loss": 0.9294,
+      "step": 14272
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4300476635255728,
+      "learning_rate": 1.9480411647415708e-07,
+      "loss": 0.8118,
+      "step": 14273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.758939098020287,
+      "learning_rate": 1.9451772202500163e-07,
+      "loss": 0.6708,
+      "step": 14274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2670988441052855,
+      "learning_rate": 1.94231534078122e-07,
+      "loss": 0.7836,
+      "step": 14275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6695590480807896,
+      "learning_rate": 1.9394555264581672e-07,
+      "loss": 0.8217,
+      "step": 14276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2025543202413354,
+      "learning_rate": 1.936597777403737e-07,
+      "loss": 0.608,
+      "step": 14277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4088607037222176,
+      "learning_rate": 1.9337420937407536e-07,
+      "loss": 0.8296,
+      "step": 14278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.437030020712099,
+      "learning_rate": 1.9308884755919132e-07,
+      "loss": 0.733,
+      "step": 14279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.54067928765091,
+      "learning_rate": 1.9280369230798568e-07,
+      "loss": 0.8526,
+      "step": 14280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6414674333312136,
+      "learning_rate": 1.925187436327114e-07,
+      "loss": 0.6817,
+      "step": 14281
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7774748372254026,
+      "learning_rate": 1.9223400154561313e-07,
+      "loss": 0.6866,
+      "step": 14282
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5248346999347873,
+      "learning_rate": 1.9194946605892773e-07,
+      "loss": 0.8436,
+      "step": 14283
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.849646274444947,
+      "learning_rate": 1.9166513718488155e-07,
+      "loss": 0.6515,
+      "step": 14284
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7582907753038506,
+      "learning_rate": 1.9138101493569428e-07,
+      "loss": 0.8074,
+      "step": 14285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6766168652153133,
+      "learning_rate": 1.9109709932357333e-07,
+      "loss": 0.7941,
+      "step": 14286
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6110252087085533,
+      "learning_rate": 1.9081339036071956e-07,
+      "loss": 0.7549,
+      "step": 14287
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1392366364194935,
+      "learning_rate": 1.9052988805932538e-07,
+      "loss": 0.8905,
+      "step": 14288
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.438904723539214,
+      "learning_rate": 1.902465924315733e-07,
+      "loss": 0.9683,
+      "step": 14289
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.562146709808586,
+      "learning_rate": 1.899635034896369e-07,
+      "loss": 0.7122,
+      "step": 14290
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2368230842413808,
+      "learning_rate": 1.8968062124568142e-07,
+      "loss": 0.7934,
+      "step": 14291
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2592690145587997,
+      "learning_rate": 1.8939794571186332e-07,
+      "loss": 0.6657,
+      "step": 14292
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9456489594123647,
+      "learning_rate": 1.8911547690032838e-07,
+      "loss": 0.676,
+      "step": 14293
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5015795882952307,
+      "learning_rate": 1.8883321482321583e-07,
+      "loss": 0.8202,
+      "step": 14294
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1094296805925743,
+      "learning_rate": 1.8855115949265535e-07,
+      "loss": 0.7692,
+      "step": 14295
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0861370432820694,
+      "learning_rate": 1.882693109207684e-07,
+      "loss": 0.8189,
+      "step": 14296
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.366450286353285,
+      "learning_rate": 1.879876691196636e-07,
+      "loss": 0.7108,
+      "step": 14297
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1563184580444985,
+      "learning_rate": 1.8770623410144628e-07,
+      "loss": 0.8428,
+      "step": 14298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.80736678474706,
+      "learning_rate": 1.8742500587820955e-07,
+      "loss": 0.7494,
+      "step": 14299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1687087985327045,
+      "learning_rate": 1.8714398446203873e-07,
+      "loss": 0.7368,
+      "step": 14300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.086870010255106,
+      "learning_rate": 1.8686316986500974e-07,
+      "loss": 0.7233,
+      "step": 14301
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9463842157371596,
+      "learning_rate": 1.86582562099189e-07,
+      "loss": 0.7287,
+      "step": 14302
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.3586328118538935,
+      "learning_rate": 1.8630216117663747e-07,
+      "loss": 0.8335,
+      "step": 14303
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6896869966285495,
+      "learning_rate": 1.86021967109401e-07,
+      "loss": 0.7233,
+      "step": 14304
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8064977162636606,
+      "learning_rate": 1.8574197990952225e-07,
+      "loss": 0.8089,
+      "step": 14305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2293119012102496,
+      "learning_rate": 1.8546219958903267e-07,
+      "loss": 0.7985,
+      "step": 14306
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.33802070221608,
+      "learning_rate": 1.8518262615995541e-07,
+      "loss": 0.8443,
+      "step": 14307
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.10648700936325,
+      "learning_rate": 1.8490325963430368e-07,
+      "loss": 0.7644,
+      "step": 14308
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8942413535326257,
+      "learning_rate": 1.8462410002408228e-07,
+      "loss": 0.8468,
+      "step": 14309
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.943708620339912,
+      "learning_rate": 1.843451473412877e-07,
+      "loss": 0.8649,
+      "step": 14310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9407672668977138,
+      "learning_rate": 1.8406640159790767e-07,
+      "loss": 0.7207,
+      "step": 14311
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2318446115060806,
+      "learning_rate": 1.837878628059192e-07,
+      "loss": 0.7103,
+      "step": 14312
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.005721202981332,
+      "learning_rate": 1.835095309772933e-07,
+      "loss": 0.7884,
+      "step": 14313
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1411166886389403,
+      "learning_rate": 1.8323140612399038e-07,
+      "loss": 0.7968,
+      "step": 14314
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8931073740622373,
+      "learning_rate": 1.829534882579598e-07,
+      "loss": 0.7094,
+      "step": 14315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1636109132084456,
+      "learning_rate": 1.8267577739114695e-07,
+      "loss": 0.7002,
+      "step": 14316
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1950726882321754,
+      "learning_rate": 1.8239827353548457e-07,
+      "loss": 0.6567,
+      "step": 14317
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9304385673554325,
+      "learning_rate": 1.8212097670289807e-07,
+      "loss": 0.7807,
+      "step": 14318
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.80337339965314,
+      "learning_rate": 1.8184388690530242e-07,
+      "loss": 0.7955,
+      "step": 14319
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8657571821482506,
+      "learning_rate": 1.8156700415460582e-07,
+      "loss": 0.7168,
+      "step": 14320
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.266113113012416,
+      "learning_rate": 1.812903284627071e-07,
+      "loss": 0.7219,
+      "step": 14321
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6046499223985538,
+      "learning_rate": 1.8101385984149343e-07,
+      "loss": 0.9436,
+      "step": 14322
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.837029269315071,
+      "learning_rate": 1.8073759830284754e-07,
+      "loss": 0.6335,
+      "step": 14323
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8934605600683363,
+      "learning_rate": 1.804615438586399e-07,
+      "loss": 0.9219,
+      "step": 14324
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.44056870604854,
+      "learning_rate": 1.801856965207338e-07,
+      "loss": 0.9167,
+      "step": 14325
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.17839744301854,
+      "learning_rate": 1.79910056300982e-07,
+      "loss": 0.6201,
+      "step": 14326
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.791201102569017,
+      "learning_rate": 1.796346232112306e-07,
+      "loss": 0.7267,
+      "step": 14327
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.075946436657794,
+      "learning_rate": 1.7935939726331453e-07,
+      "loss": 0.7993,
+      "step": 14328
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3390424904557063,
+      "learning_rate": 1.7908437846906158e-07,
+      "loss": 0.8309,
+      "step": 14329
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1146331731015167,
+      "learning_rate": 1.7880956684028894e-07,
+      "loss": 0.9439,
+      "step": 14330
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2017634136915112,
+      "learning_rate": 1.7853496238880718e-07,
+      "loss": 0.6477,
+      "step": 14331
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0623765466327977,
+      "learning_rate": 1.7826056512641633e-07,
+      "loss": 0.719,
+      "step": 14332
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0923157019929708,
+      "learning_rate": 1.7798637506490745e-07,
+      "loss": 0.6057,
+      "step": 14333
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3245656532205072,
+      "learning_rate": 1.7771239221606285e-07,
+      "loss": 0.6627,
+      "step": 14334
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0726177724862334,
+      "learning_rate": 1.7743861659165695e-07,
+      "loss": 0.8995,
+      "step": 14335
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2016399598647243,
+      "learning_rate": 1.7716504820345427e-07,
+      "loss": 0.8507,
+      "step": 14336
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.118151648121919,
+      "learning_rate": 1.768916870632098e-07,
+      "loss": 0.8261,
+      "step": 14337
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0909223076085897,
+      "learning_rate": 1.7661853318267142e-07,
+      "loss": 0.6993,
+      "step": 14338
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1067520495924406,
+      "learning_rate": 1.7634558657357748e-07,
+      "loss": 0.822,
+      "step": 14339
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.28220865745263,
+      "learning_rate": 1.7607284724765583e-07,
+      "loss": 0.7542,
+      "step": 14340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.652823553106424,
+      "learning_rate": 1.7580031521662765e-07,
+      "loss": 0.746,
+      "step": 14341
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5222675931942726,
+      "learning_rate": 1.7552799049220415e-07,
+      "loss": 0.6954,
+      "step": 14342
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.944071409606013,
+      "learning_rate": 1.752558730860876e-07,
+      "loss": 0.9025,
+      "step": 14343
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.315770057785506,
+      "learning_rate": 1.7498396300997146e-07,
+      "loss": 0.7689,
+      "step": 14344
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7938092938975072,
+      "learning_rate": 1.747122602755391e-07,
+      "loss": 0.8203,
+      "step": 14345
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.305862806737502,
+      "learning_rate": 1.7444076489446958e-07,
+      "loss": 0.6888,
+      "step": 14346
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8419387227818143,
+      "learning_rate": 1.7416947687842633e-07,
+      "loss": 0.9945,
+      "step": 14347
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.742821226467196,
+      "learning_rate": 1.738983962390678e-07,
+      "loss": 0.6939,
+      "step": 14348
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0051272318012945,
+      "learning_rate": 1.736275229880441e-07,
+      "loss": 0.6845,
+      "step": 14349
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7434404304398887,
+      "learning_rate": 1.733568571369948e-07,
+      "loss": 0.9573,
+      "step": 14350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.116313196483196,
+      "learning_rate": 1.7308639869755062e-07,
+      "loss": 0.9431,
+      "step": 14351
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.407468239162837,
+      "learning_rate": 1.728161476813345e-07,
+      "loss": 0.8348,
+      "step": 14352
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9202435067856682,
+      "learning_rate": 1.7254610409995875e-07,
+      "loss": 0.9067,
+      "step": 14353
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.211298892959236,
+      "learning_rate": 1.7227626796502807e-07,
+      "loss": 0.7349,
+      "step": 14354
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7729194040124123,
+      "learning_rate": 1.720066392881381e-07,
+      "loss": 0.7429,
+      "step": 14355
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.447259896103425,
+      "learning_rate": 1.7173721808087573e-07,
+      "loss": 0.7338,
+      "step": 14356
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1384883252366347,
+      "learning_rate": 1.7146800435481837e-07,
+      "loss": 0.8929,
+      "step": 14357
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8306098296933717,
+      "learning_rate": 1.7119899812153394e-07,
+      "loss": 0.6507,
+      "step": 14358
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.207178187058072,
+      "learning_rate": 1.7093019939258327e-07,
+      "loss": 0.7047,
+      "step": 14359
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9236564082598155,
+      "learning_rate": 1.706616081795165e-07,
+      "loss": 0.7705,
+      "step": 14360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.795785727888032,
+      "learning_rate": 1.7039322449387664e-07,
+      "loss": 0.7536,
+      "step": 14361
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4346839964069886,
+      "learning_rate": 1.7012504834719556e-07,
+      "loss": 0.8448,
+      "step": 14362
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2678500187269033,
+      "learning_rate": 1.6985707975099742e-07,
+      "loss": 0.6759,
+      "step": 14363
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.28404107972384,
+      "learning_rate": 1.6958931871679908e-07,
+      "loss": 0.9263,
+      "step": 14364
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1007165213009036,
+      "learning_rate": 1.6932176525610465e-07,
+      "loss": 0.8801,
+      "step": 14365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1704718408211239,
+      "learning_rate": 1.6905441938041222e-07,
+      "loss": 0.7381,
+      "step": 14366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0951120271271435,
+      "learning_rate": 1.687872811012109e-07,
+      "loss": 0.7476,
+      "step": 14367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.66696782194646,
+      "learning_rate": 1.685203504299804e-07,
+      "loss": 0.5688,
+      "step": 14368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.872448622400664,
+      "learning_rate": 1.6825362737818985e-07,
+      "loss": 0.734,
+      "step": 14369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.692157718753607,
+      "learning_rate": 1.6798711195730233e-07,
+      "loss": 0.8039,
+      "step": 14370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.895946521747711,
+      "learning_rate": 1.677208041787698e-07,
+      "loss": 0.7737,
+      "step": 14371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9467801625767907,
+      "learning_rate": 1.6745470405403697e-07,
+      "loss": 0.6897,
+      "step": 14372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.825861650556877,
+      "learning_rate": 1.6718881159453693e-07,
+      "loss": 0.6924,
+      "step": 14373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5551678290351125,
+      "learning_rate": 1.6692312681169775e-07,
+      "loss": 0.6779,
+      "step": 14374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2623497158955606,
+      "learning_rate": 1.6665764971693587e-07,
+      "loss": 0.729,
+      "step": 14375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1269500335557434,
+      "learning_rate": 1.6639238032165883e-07,
+      "loss": 0.8111,
+      "step": 14376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.096306498708458,
+      "learning_rate": 1.6612731863726638e-07,
+      "loss": 0.7412,
+      "step": 14377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9109510551409277,
+      "learning_rate": 1.6586246467514833e-07,
+      "loss": 0.6736,
+      "step": 14378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8858635074793628,
+      "learning_rate": 1.6559781844668666e-07,
+      "loss": 0.7421,
+      "step": 14379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.247079188321499,
+      "learning_rate": 1.6533337996325337e-07,
+      "loss": 0.8119,
+      "step": 14380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.393012542058732,
+      "learning_rate": 1.650691492362122e-07,
+      "loss": 0.9715,
+      "step": 14381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8164665720603916,
+      "learning_rate": 1.6480512627691792e-07,
+      "loss": 0.7773,
+      "step": 14382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0357012845310773,
+      "learning_rate": 1.6454131109671534e-07,
+      "loss": 0.7915,
+      "step": 14383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9332163010362062,
+      "learning_rate": 1.6427770370694208e-07,
+      "loss": 0.579,
+      "step": 14384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0430607266948944,
+      "learning_rate": 1.6401430411892572e-07,
+      "loss": 0.6086,
+      "step": 14385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2719174897008925,
+      "learning_rate": 1.63751112343985e-07,
+      "loss": 0.7723,
+      "step": 14386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.253810867977371,
+      "learning_rate": 1.6348812839342976e-07,
+      "loss": 0.7985,
+      "step": 14387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.319110338363763,
+      "learning_rate": 1.6322535227856152e-07,
+      "loss": 0.8829,
+      "step": 14388
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1986932576687064,
+      "learning_rate": 1.6296278401067122e-07,
+      "loss": 0.8247,
+      "step": 14389
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5109400591066957,
+      "learning_rate": 1.6270042360104377e-07,
+      "loss": 0.6379,
+      "step": 14390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5274861210043107,
+      "learning_rate": 1.624382710609512e-07,
+      "loss": 0.8327,
+      "step": 14391
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8483560723397494,
+      "learning_rate": 1.621763264016607e-07,
+      "loss": 0.6432,
+      "step": 14392
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.523135019168041,
+      "learning_rate": 1.6191458963442762e-07,
+      "loss": 0.8888,
+      "step": 14393
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.407808216012434,
+      "learning_rate": 1.6165306077049969e-07,
+      "loss": 0.7696,
+      "step": 14394
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.413677230688246,
+      "learning_rate": 1.6139173982111567e-07,
+      "loss": 0.8482,
+      "step": 14395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.921119568442313,
+      "learning_rate": 1.611306267975038e-07,
+      "loss": 0.6675,
+      "step": 14396
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0038504505610835,
+      "learning_rate": 1.6086972171088678e-07,
+      "loss": 0.8807,
+      "step": 14397
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7616714565237896,
+      "learning_rate": 1.606090245724745e-07,
+      "loss": 0.7772,
+      "step": 14398
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.152060520337453,
+      "learning_rate": 1.603485353934703e-07,
+      "loss": 0.6538,
+      "step": 14399
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.354396082193376,
+      "learning_rate": 1.6008825418506846e-07,
+      "loss": 0.8572,
+      "step": 14400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.00520745350259,
+      "learning_rate": 1.5982818095845286e-07,
+      "loss": 0.6917,
+      "step": 14401
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4563119094520136,
+      "learning_rate": 1.5956831572480013e-07,
+      "loss": 1.0004,
+      "step": 14402
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9228368470297865,
+      "learning_rate": 1.5930865849527744e-07,
+      "loss": 0.7451,
+      "step": 14403
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.587131511367415,
+      "learning_rate": 1.5904920928104196e-07,
+      "loss": 0.7761,
+      "step": 14404
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.089925280023264,
+      "learning_rate": 1.5878996809324366e-07,
+      "loss": 0.5744,
+      "step": 14405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2127878466354747,
+      "learning_rate": 1.5853093494302195e-07,
+      "loss": 0.8515,
+      "step": 14406
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.11180974028508,
+      "learning_rate": 1.582721098415091e-07,
+      "loss": 0.5947,
+      "step": 14407
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.130295549668118,
+      "learning_rate": 1.5801349279982726e-07,
+      "loss": 0.6561,
+      "step": 14408
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5093475606361646,
+      "learning_rate": 1.577550838290881e-07,
+      "loss": 0.8835,
+      "step": 14409
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9090407368991442,
+      "learning_rate": 1.5749688294039778e-07,
+      "loss": 0.7184,
+      "step": 14410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.025486937484922,
+      "learning_rate": 1.5723889014485128e-07,
+      "loss": 0.8712,
+      "step": 14411
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.215207118164194,
+      "learning_rate": 1.5698110545353474e-07,
+      "loss": 0.8839,
+      "step": 14412
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.943797399648688,
+      "learning_rate": 1.567235288775265e-07,
+      "loss": 0.8041,
+      "step": 14413
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2760118337261046,
+      "learning_rate": 1.564661604278944e-07,
+      "loss": 0.8796,
+      "step": 14414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2867961136360155,
+      "learning_rate": 1.5620900011569906e-07,
+      "loss": 0.8919,
+      "step": 14415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9460274269197084,
+      "learning_rate": 1.5595204795199047e-07,
+      "loss": 0.7246,
+      "step": 14416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8086503400241645,
+      "learning_rate": 1.5569530394780986e-07,
+      "loss": 0.7867,
+      "step": 14417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4213536766954076,
+      "learning_rate": 1.5543876811419222e-07,
+      "loss": 0.5088,
+      "step": 14418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.006039489075549,
+      "learning_rate": 1.5518244046215936e-07,
+      "loss": 0.6183,
+      "step": 14419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.133875214266454,
+      "learning_rate": 1.5492632100272686e-07,
+      "loss": 0.8058,
+      "step": 14420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.466608942409519,
+      "learning_rate": 1.546704097469015e-07,
+      "loss": 0.7382,
+      "step": 14421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0593433730478448,
+      "learning_rate": 1.544147067056795e-07,
+      "loss": 0.8029,
+      "step": 14422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9111298674432795,
+      "learning_rate": 1.5415921189004922e-07,
+      "loss": 0.7545,
+      "step": 14423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9387198259491765,
+      "learning_rate": 1.539039253109892e-07,
+      "loss": 0.8155,
+      "step": 14424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1659448323149972,
+      "learning_rate": 1.5364884697947113e-07,
+      "loss": 0.6561,
+      "step": 14425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.992188820338209,
+      "learning_rate": 1.533939769064563e-07,
+      "loss": 0.7648,
+      "step": 14426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7846922908631038,
+      "learning_rate": 1.5313931510289482e-07,
+      "loss": 0.8503,
+      "step": 14427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9411572599523645,
+      "learning_rate": 1.528848615797318e-07,
+      "loss": 0.609,
+      "step": 14428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4108542512893383,
+      "learning_rate": 1.526306163479019e-07,
+      "loss": 0.9199,
+      "step": 14429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.255700375846419,
+      "learning_rate": 1.523765794183296e-07,
+      "loss": 0.7802,
+      "step": 14430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.162151813172094,
+      "learning_rate": 1.521227508019324e-07,
+      "loss": 1.0171,
+      "step": 14431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9587168146239287,
+      "learning_rate": 1.518691305096165e-07,
+      "loss": 0.7463,
+      "step": 14432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.04535033781878,
+      "learning_rate": 1.5161571855228264e-07,
+      "loss": 0.8531,
+      "step": 14433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.113508490490645,
+      "learning_rate": 1.5136251494081822e-07,
+      "loss": 1.0202,
+      "step": 14434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9836953651202176,
+      "learning_rate": 1.5110951968610564e-07,
+      "loss": 0.6622,
+      "step": 14435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0970802785546354,
+      "learning_rate": 1.5085673279901626e-07,
+      "loss": 0.6474,
+      "step": 14436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.406854996458001,
+      "learning_rate": 1.5060415429041186e-07,
+      "loss": 0.8291,
+      "step": 14437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.908439575749106,
+      "learning_rate": 1.503517841711477e-07,
+      "loss": 0.7446,
+      "step": 14438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8118846818804974,
+      "learning_rate": 1.5009962245206845e-07,
+      "loss": 0.7746,
+      "step": 14439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5717879682783495,
+      "learning_rate": 1.4984766914400928e-07,
+      "loss": 0.8015,
+      "step": 14440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.981409234557674,
+      "learning_rate": 1.4959592425779768e-07,
+      "loss": 0.8456,
+      "step": 14441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9890024469175855,
+      "learning_rate": 1.4934438780425108e-07,
+      "loss": 0.8172,
+      "step": 14442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8134795211189494,
+      "learning_rate": 1.4909305979417976e-07,
+      "loss": 0.915,
+      "step": 14443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4506387775157914,
+      "learning_rate": 1.488419402383834e-07,
+      "loss": 0.6264,
+      "step": 14444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.40419889835605,
+      "learning_rate": 1.4859102914765278e-07,
+      "loss": 0.8503,
+      "step": 14445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8824667403196464,
+      "learning_rate": 1.4834032653276986e-07,
+      "loss": 0.6734,
+      "step": 14446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9179112472193376,
+      "learning_rate": 1.4808983240450881e-07,
+      "loss": 0.8497,
+      "step": 14447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7978830581187348,
+      "learning_rate": 1.4783954677363376e-07,
+      "loss": 0.7294,
+      "step": 14448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.862310861682938,
+      "learning_rate": 1.4758946965089894e-07,
+      "loss": 0.4963,
+      "step": 14449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.992417013028821,
+      "learning_rate": 1.473396010470518e-07,
+      "loss": 0.7615,
+      "step": 14450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3683858493314265,
+      "learning_rate": 1.470899409728299e-07,
+      "loss": 0.783,
+      "step": 14451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0591020548703423,
+      "learning_rate": 1.4684048943896022e-07,
+      "loss": 0.54,
+      "step": 14452
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7732862237661298,
+      "learning_rate": 1.4659124645616418e-07,
+      "loss": 0.9037,
+      "step": 14453
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2484539833716355,
+      "learning_rate": 1.4634221203515097e-07,
+      "loss": 0.7393,
+      "step": 14454
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0469684613083265,
+      "learning_rate": 1.4609338618662318e-07,
+      "loss": 0.6268,
+      "step": 14455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.723235323410724,
+      "learning_rate": 1.458447689212722e-07,
+      "loss": 0.6159,
+      "step": 14456
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.17773878144325,
+      "learning_rate": 1.455963602497823e-07,
+      "loss": 0.698,
+      "step": 14457
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.455522427211079,
+      "learning_rate": 1.4534816018282882e-07,
+      "loss": 0.7495,
+      "step": 14458
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.275652556543144,
+      "learning_rate": 1.4510016873107657e-07,
+      "loss": 0.8844,
+      "step": 14459
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9507023563707326,
+      "learning_rate": 1.4485238590518147e-07,
+      "loss": 0.8859,
+      "step": 14460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.450185399978002,
+      "learning_rate": 1.4460481171579334e-07,
+      "loss": 0.937,
+      "step": 14461
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9770027404892403,
+      "learning_rate": 1.4435744617354975e-07,
+      "loss": 0.987,
+      "step": 14462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7602069057018812,
+      "learning_rate": 1.4411028928908054e-07,
+      "loss": 0.8258,
+      "step": 14463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3038315036486234,
+      "learning_rate": 1.4386334107300727e-07,
+      "loss": 0.8132,
+      "step": 14464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.768089169743287,
+      "learning_rate": 1.4361660153594137e-07,
+      "loss": 0.8999,
+      "step": 14465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4512173142598748,
+      "learning_rate": 1.4337007068848607e-07,
+      "loss": 0.8268,
+      "step": 14466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2859731070978246,
+      "learning_rate": 1.4312374854123456e-07,
+      "loss": 0.9006,
+      "step": 14467
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8030334872924976,
+      "learning_rate": 1.4287763510477225e-07,
+      "loss": 0.7653,
+      "step": 14468
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1920471038490095,
+      "learning_rate": 1.4263173038967627e-07,
+      "loss": 0.8271,
+      "step": 14469
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6542142168483718,
+      "learning_rate": 1.42386034406512e-07,
+      "loss": 0.8134,
+      "step": 14470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3262450261486283,
+      "learning_rate": 1.421405471658388e-07,
+      "loss": 0.9405,
+      "step": 14471
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6240598167964073,
+      "learning_rate": 1.4189526867820491e-07,
+      "loss": 0.8194,
+      "step": 14472
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9295327004081038,
+      "learning_rate": 1.4165019895415133e-07,
+      "loss": 0.7412,
+      "step": 14473
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9717043074979301,
+      "learning_rate": 1.4140533800420853e-07,
+      "loss": 0.9988,
+      "step": 14474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7716490511346863,
+      "learning_rate": 1.4116068583889865e-07,
+      "loss": 0.6709,
+      "step": 14475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0746488051873988,
+      "learning_rate": 1.409162424687366e-07,
+      "loss": 0.7162,
+      "step": 14476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3939843821327087,
+      "learning_rate": 1.4067200790422454e-07,
+      "loss": 0.8212,
+      "step": 14477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9326256824138155,
+      "learning_rate": 1.4042798215585796e-07,
+      "loss": 0.7321,
+      "step": 14478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.053224546016285,
+      "learning_rate": 1.401841652341246e-07,
+      "loss": 0.77,
+      "step": 14479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1480217966871724,
+      "learning_rate": 1.3994055714950105e-07,
+      "loss": 0.6263,
+      "step": 14480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9316413690168077,
+      "learning_rate": 1.3969715791245565e-07,
+      "loss": 0.7039,
+      "step": 14481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.042452998392673,
+      "learning_rate": 1.3945396753344776e-07,
+      "loss": 0.9223,
+      "step": 14482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.219804713377953,
+      "learning_rate": 1.3921098602292793e-07,
+      "loss": 0.7022,
+      "step": 14483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0214251937272403,
+      "learning_rate": 1.389682133913378e-07,
+      "loss": 0.6691,
+      "step": 14484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6847536924365345,
+      "learning_rate": 1.3872564964910905e-07,
+      "loss": 0.7554,
+      "step": 14485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0733914778767537,
+      "learning_rate": 1.384832948066661e-07,
+      "loss": 0.7257,
+      "step": 14486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.06260558700967,
+      "learning_rate": 1.382411488744234e-07,
+      "loss": 0.7758,
+      "step": 14487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.385199716457109,
+      "learning_rate": 1.3799921186278597e-07,
+      "loss": 0.6064,
+      "step": 14488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0518671209757424,
+      "learning_rate": 1.3775748378215047e-07,
+      "loss": 0.806,
+      "step": 14489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.167515257070062,
+      "learning_rate": 1.3751596464290529e-07,
+      "loss": 0.7014,
+      "step": 14490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2427077367372723,
+      "learning_rate": 1.372746544554282e-07,
+      "loss": 0.8629,
+      "step": 14491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0520153451552363,
+      "learning_rate": 1.3703355323008872e-07,
+      "loss": 0.8044,
+      "step": 14492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.662913143346792,
+      "learning_rate": 1.3679266097724797e-07,
+      "loss": 0.9246,
+      "step": 14493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.681951794737808,
+      "learning_rate": 1.3655197770725826e-07,
+      "loss": 0.73,
+      "step": 14494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5079047284679934,
+      "learning_rate": 1.363115034304602e-07,
+      "loss": 0.9492,
+      "step": 14495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.37459947757917,
+      "learning_rate": 1.3607123815718993e-07,
+      "loss": 0.6091,
+      "step": 14496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.224314838209512,
+      "learning_rate": 1.358311818977709e-07,
+      "loss": 0.8621,
+      "step": 14497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8611688648771403,
+      "learning_rate": 1.3559133466251874e-07,
+      "loss": 0.986,
+      "step": 14498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.725471114822992,
+      "learning_rate": 1.3535169646174073e-07,
+      "loss": 0.7538,
+      "step": 14499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.595920389160128,
+      "learning_rate": 1.3511226730573478e-07,
+      "loss": 1.0341,
+      "step": 14500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.03429365630583,
+      "learning_rate": 1.3487304720478933e-07,
+      "loss": 0.9364,
+      "step": 14501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9367200132616773,
+      "learning_rate": 1.346340361691839e-07,
+      "loss": 0.7347,
+      "step": 14502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.762894031357296,
+      "learning_rate": 1.343952342091892e-07,
+      "loss": 0.7805,
+      "step": 14503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.409731470362435,
+      "learning_rate": 1.3415664133506812e-07,
+      "loss": 0.7145,
+      "step": 14504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.224788468318283,
+      "learning_rate": 1.3391825755707356e-07,
+      "loss": 0.7844,
+      "step": 14505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8568460731694392,
+      "learning_rate": 1.3368008288544843e-07,
+      "loss": 0.8927,
+      "step": 14506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.255023033748131,
+      "learning_rate": 1.3344211733042788e-07,
+      "loss": 0.8888,
+      "step": 14507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7079788496000718,
+      "learning_rate": 1.3320436090223765e-07,
+      "loss": 0.6847,
+      "step": 14508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4323106354233603,
+      "learning_rate": 1.3296681361109564e-07,
+      "loss": 0.919,
+      "step": 14509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.195056957687636,
+      "learning_rate": 1.327294754672087e-07,
+      "loss": 0.7873,
+      "step": 14510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8376678105230155,
+      "learning_rate": 1.324923464807759e-07,
+      "loss": 0.8423,
+      "step": 14511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.171907599426582,
+      "learning_rate": 1.3225542666198854e-07,
+      "loss": 0.8503,
+      "step": 14512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1028086637021879,
+      "learning_rate": 1.3201871602102512e-07,
+      "loss": 0.6527,
+      "step": 14513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3561111437420372,
+      "learning_rate": 1.3178221456806028e-07,
+      "loss": 0.8706,
+      "step": 14514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3371330842133164,
+      "learning_rate": 1.3154592231325535e-07,
+      "loss": 0.869,
+      "step": 14515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4599772949998537,
+      "learning_rate": 1.3130983926676445e-07,
+      "loss": 0.7334,
+      "step": 14516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3422173832721254,
+      "learning_rate": 1.310739654387333e-07,
+      "loss": 0.6774,
+      "step": 14517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9838200616548294,
+      "learning_rate": 1.308383008392977e-07,
+      "loss": 0.7921,
+      "step": 14518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.982259938119308,
+      "learning_rate": 1.3060284547858403e-07,
+      "loss": 0.7093,
+      "step": 14519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8551827314954652,
+      "learning_rate": 1.3036759936671084e-07,
+      "loss": 0.6876,
+      "step": 14520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9299949782928034,
+      "learning_rate": 1.3013256251378725e-07,
+      "loss": 0.8246,
+      "step": 14521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.113387125351299,
+      "learning_rate": 1.2989773492991298e-07,
+      "loss": 0.802,
+      "step": 14522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5112047326649103,
+      "learning_rate": 1.2966311662517994e-07,
+      "loss": 0.5825,
+      "step": 14523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.048208664908258,
+      "learning_rate": 1.2942870760966952e-07,
+      "loss": 0.8154,
+      "step": 14524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7375930715654295,
+      "learning_rate": 1.2919450789345477e-07,
+      "loss": 0.6992,
+      "step": 14525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7873420549989274,
+      "learning_rate": 1.289605174865993e-07,
+      "loss": 0.7789,
+      "step": 14526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0615399866826216,
+      "learning_rate": 1.2872673639915955e-07,
+      "loss": 0.6557,
+      "step": 14527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9874378321870843,
+      "learning_rate": 1.2849316464118078e-07,
+      "loss": 0.7062,
+      "step": 14528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.30082879874245,
+      "learning_rate": 1.282598022226994e-07,
+      "loss": 0.8199,
+      "step": 14529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6258260166968963,
+      "learning_rate": 1.2802664915374575e-07,
+      "loss": 0.8961,
+      "step": 14530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7614062234710364,
+      "learning_rate": 1.2779370544433567e-07,
+      "loss": 0.9098,
+      "step": 14531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.183840105366804,
+      "learning_rate": 1.275609711044823e-07,
+      "loss": 0.8396,
+      "step": 14532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3997965346695436,
+      "learning_rate": 1.2732844614418482e-07,
+      "loss": 0.7545,
+      "step": 14533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0807143358052962,
+      "learning_rate": 1.270961305734364e-07,
+      "loss": 0.6355,
+      "step": 14534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.085275005309126,
+      "learning_rate": 1.2686402440222013e-07,
+      "loss": 0.8364,
+      "step": 14535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.644492725935907,
+      "learning_rate": 1.2663212764050858e-07,
+      "loss": 0.8532,
+      "step": 14536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.093270120419849,
+      "learning_rate": 1.2640044029826937e-07,
+      "loss": 0.8204,
+      "step": 14537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0668791899383887,
+      "learning_rate": 1.261689623854573e-07,
+      "loss": 0.8933,
+      "step": 14538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2107766234695942,
+      "learning_rate": 1.2593769391201827e-07,
+      "loss": 0.6847,
+      "step": 14539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3493368076193626,
+      "learning_rate": 1.257066348878927e-07,
+      "loss": 0.6397,
+      "step": 14540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9791019284839173,
+      "learning_rate": 1.254757853230082e-07,
+      "loss": 0.671,
+      "step": 14541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.029759013483875,
+      "learning_rate": 1.252451452272857e-07,
+      "loss": 0.7805,
+      "step": 14542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1623474577129094,
+      "learning_rate": 1.2501471461063564e-07,
+      "loss": 0.9634,
+      "step": 14543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0954313120126207,
+      "learning_rate": 1.247844934829606e-07,
+      "loss": 0.714,
+      "step": 14544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.102942207475603,
+      "learning_rate": 1.2455448185415386e-07,
+      "loss": 0.8526,
+      "step": 14545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.030025786586306,
+      "learning_rate": 1.2432467973409857e-07,
+      "loss": 0.6584,
+      "step": 14546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.535636938922849,
+      "learning_rate": 1.2409508713267072e-07,
+      "loss": 0.6268,
+      "step": 14547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.144836465335095,
+      "learning_rate": 1.238657040597374e-07,
+      "loss": 0.6832,
+      "step": 14548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9003154022651163,
+      "learning_rate": 1.2363653052515302e-07,
+      "loss": 0.7126,
+      "step": 14549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6223906130786516,
+      "learning_rate": 1.234075665387674e-07,
+      "loss": 0.7095,
+      "step": 14550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2074279156268206,
+      "learning_rate": 1.2317881211041993e-07,
+      "loss": 0.8754,
+      "step": 14551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.379077673282439,
+      "learning_rate": 1.2295026724993998e-07,
+      "loss": 0.6446,
+      "step": 14552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.182435954681758,
+      "learning_rate": 1.2272193196714854e-07,
+      "loss": 0.7181,
+      "step": 14553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2605627328795275,
+      "learning_rate": 1.2249380627185781e-07,
+      "loss": 0.8325,
+      "step": 14554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7443815339272952,
+      "learning_rate": 1.2226589017387158e-07,
+      "loss": 0.7041,
+      "step": 14555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0486446757680756,
+      "learning_rate": 1.2203818368298315e-07,
+      "loss": 0.6874,
+      "step": 14556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4022325741063777,
+      "learning_rate": 1.2181068680897745e-07,
+      "loss": 0.713,
+      "step": 14557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.169093874462297,
+      "learning_rate": 1.2158339956163113e-07,
+      "loss": 0.9692,
+      "step": 14558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.269991661848138,
+      "learning_rate": 1.2135632195071133e-07,
+      "loss": 0.7239,
+      "step": 14559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.06662105647848,
+      "learning_rate": 1.211294539859753e-07,
+      "loss": 0.7587,
+      "step": 14560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6664530463631713,
+      "learning_rate": 1.2090279567717246e-07,
+      "loss": 0.9086,
+      "step": 14561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.395397508668774,
+      "learning_rate": 1.2067634703404218e-07,
+      "loss": 0.9162,
+      "step": 14562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.069581898429705,
+      "learning_rate": 1.204501080663173e-07,
+      "loss": 0.634,
+      "step": 14563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9708542773440725,
+      "learning_rate": 1.202240787837178e-07,
+      "loss": 0.6875,
+      "step": 14564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.082919933336952,
+      "learning_rate": 1.1999825919595808e-07,
+      "loss": 0.6552,
+      "step": 14565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9312137627575519,
+      "learning_rate": 1.1977264931274158e-07,
+      "loss": 0.7901,
+      "step": 14566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3433691200507565,
+      "learning_rate": 1.1954724914376215e-07,
+      "loss": 0.8892,
+      "step": 14567
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.485522032830346,
+      "learning_rate": 1.1932205869870762e-07,
+      "loss": 0.7763,
+      "step": 14568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7037795133379943,
+      "learning_rate": 1.1909707798725412e-07,
+      "loss": 0.9357,
+      "step": 14569
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0835875886433555,
+      "learning_rate": 1.1887230701906893e-07,
+      "loss": 0.6258,
+      "step": 14570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0892869540381005,
+      "learning_rate": 1.186477458038121e-07,
+      "loss": 0.711,
+      "step": 14571
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.212076750147803,
+      "learning_rate": 1.1842339435113259e-07,
+      "loss": 0.7854,
+      "step": 14572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.058685603148361,
+      "learning_rate": 1.1819925267067211e-07,
+      "loss": 0.7842,
+      "step": 14573
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.580510447668976,
+      "learning_rate": 1.1797532077206187e-07,
+      "loss": 0.883,
+      "step": 14574
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3054368467117565,
+      "learning_rate": 1.177515986649247e-07,
+      "loss": 0.8117,
+      "step": 14575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.482088649931666,
+      "learning_rate": 1.1752808635887514e-07,
+      "loss": 0.9124,
+      "step": 14576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2761550028369486,
+      "learning_rate": 1.1730478386351719e-07,
+      "loss": 0.7492,
+      "step": 14577
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.245747390158776,
+      "learning_rate": 1.1708169118844759e-07,
+      "loss": 0.8698,
+      "step": 14578
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1358802934856302,
+      "learning_rate": 1.1685880834325203e-07,
+      "loss": 0.5321,
+      "step": 14579
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.056677229847286,
+      "learning_rate": 1.1663613533750839e-07,
+      "loss": 0.7908,
+      "step": 14580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9237744655268008,
+      "learning_rate": 1.1641367218078736e-07,
+      "loss": 0.6837,
+      "step": 14581
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5294562409270673,
+      "learning_rate": 1.1619141888264574e-07,
+      "loss": 0.872,
+      "step": 14582
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7148773994924027,
+      "learning_rate": 1.1596937545263587e-07,
+      "loss": 0.8921,
+      "step": 14583
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1529051533736645,
+      "learning_rate": 1.1574754190030014e-07,
+      "loss": 0.852,
+      "step": 14584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.592253724244335,
+      "learning_rate": 1.1552591823516978e-07,
+      "loss": 0.8826,
+      "step": 14585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1051886941335058,
+      "learning_rate": 1.1530450446676889e-07,
+      "loss": 0.632,
+      "step": 14586
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1711485955799925,
+      "learning_rate": 1.1508330060461204e-07,
+      "loss": 0.7445,
+      "step": 14587
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1382602278541394,
+      "learning_rate": 1.1486230665820552e-07,
+      "loss": 0.5212,
+      "step": 14588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2037119723671426,
+      "learning_rate": 1.1464152263704565e-07,
+      "loss": 0.8144,
+      "step": 14589
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0293607985968416,
+      "learning_rate": 1.1442094855061925e-07,
+      "loss": 0.7669,
+      "step": 14590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4112877750397455,
+      "learning_rate": 1.1420058440840598e-07,
+      "loss": 0.7997,
+      "step": 14591
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.477672987324862,
+      "learning_rate": 1.1398043021987438e-07,
+      "loss": 0.6651,
+      "step": 14592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.41400604037286,
+      "learning_rate": 1.1376048599448519e-07,
+      "loss": 0.6587,
+      "step": 14593
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7852234866520031,
+      "learning_rate": 1.1354075174169088e-07,
+      "loss": 0.8553,
+      "step": 14594
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7596787534454164,
+      "learning_rate": 1.1332122747093277e-07,
+      "loss": 0.8469,
+      "step": 14595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.401230706935137,
+      "learning_rate": 1.1310191319164498e-07,
+      "loss": 0.9322,
+      "step": 14596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7496741473163682,
+      "learning_rate": 1.1288280891325165e-07,
+      "loss": 0.8528,
+      "step": 14597
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.086369245675055,
+      "learning_rate": 1.1266391464516802e-07,
+      "loss": 0.865,
+      "step": 14598
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8920849060036349,
+      "learning_rate": 1.12445230396801e-07,
+      "loss": 0.7682,
+      "step": 14599
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.028639198614091,
+      "learning_rate": 1.1222675617754752e-07,
+      "loss": 0.9221,
+      "step": 14600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.388414720003739,
+      "learning_rate": 1.1200849199679564e-07,
+      "loss": 0.757,
+      "step": 14601
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7884813946184583,
+      "learning_rate": 1.1179043786392507e-07,
+      "loss": 0.6422,
+      "step": 14602
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2264172375831044,
+      "learning_rate": 1.1157259378830665e-07,
+      "loss": 0.7641,
+      "step": 14603
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9449213274808959,
+      "learning_rate": 1.1135495977930011e-07,
+      "loss": 0.6383,
+      "step": 14604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2459833620213114,
+      "learning_rate": 1.1113753584625853e-07,
+      "loss": 0.8068,
+      "step": 14605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8748140836916034,
+      "learning_rate": 1.1092032199852609e-07,
+      "loss": 0.8078,
+      "step": 14606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7904192298374517,
+      "learning_rate": 1.1070331824543534e-07,
+      "loss": 0.6259,
+      "step": 14607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3285027928478637,
+      "learning_rate": 1.1048652459631216e-07,
+      "loss": 0.9516,
+      "step": 14608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.138471406235783,
+      "learning_rate": 1.1026994106047296e-07,
+      "loss": 0.8748,
+      "step": 14609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1634160181335567,
+      "learning_rate": 1.1005356764722364e-07,
+      "loss": 0.7247,
+      "step": 14610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7852171404598833,
+      "learning_rate": 1.0983740436586343e-07,
+      "loss": 0.6184,
+      "step": 14611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.457848765847274,
+      "learning_rate": 1.0962145122568046e-07,
+      "loss": 0.64,
+      "step": 14612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6964576152975392,
+      "learning_rate": 1.0940570823595565e-07,
+      "loss": 0.7921,
+      "step": 14613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.21064948405917,
+      "learning_rate": 1.0919017540595933e-07,
+      "loss": 0.9792,
+      "step": 14614
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1090085852705132,
+      "learning_rate": 1.0897485274495357e-07,
+      "loss": 0.702,
+      "step": 14615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2021419685280028,
+      "learning_rate": 1.0875974026219149e-07,
+      "loss": 0.6862,
+      "step": 14616
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0582071529350319,
+      "learning_rate": 1.0854483796691629e-07,
+      "loss": 0.748,
+      "step": 14617
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2450322773104303,
+      "learning_rate": 1.0833014586836388e-07,
+      "loss": 0.8265,
+      "step": 14618
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8472371267521055,
+      "learning_rate": 1.0811566397575912e-07,
+      "loss": 0.7226,
+      "step": 14619
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1882358708892955,
+      "learning_rate": 1.0790139229831909e-07,
+      "loss": 0.7694,
+      "step": 14620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1604718767229687,
+      "learning_rate": 1.0768733084525196e-07,
+      "loss": 0.7734,
+      "step": 14621
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9818906389313946,
+      "learning_rate": 1.0747347962575539e-07,
+      "loss": 0.7275,
+      "step": 14622
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.281974843549214,
+      "learning_rate": 1.0725983864901978e-07,
+      "loss": 0.8953,
+      "step": 14623
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.379585185872317,
+      "learning_rate": 1.0704640792422616e-07,
+      "loss": 0.7364,
+      "step": 14624
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1386642556153634,
+      "learning_rate": 1.0683318746054439e-07,
+      "loss": 0.9251,
+      "step": 14625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0375098807682055,
+      "learning_rate": 1.066201772671388e-07,
+      "loss": 0.9127,
+      "step": 14626
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.915284384319432,
+      "learning_rate": 1.0640737735316264e-07,
+      "loss": 0.8537,
+      "step": 14627
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.818161188801461,
+      "learning_rate": 1.0619478772776027e-07,
+      "loss": 0.9196,
+      "step": 14628
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 40.69011328870871,
+      "learning_rate": 1.0598240840006658e-07,
+      "loss": 0.6525,
+      "step": 14629
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.053851119368602,
+      "learning_rate": 1.0577023937920816e-07,
+      "loss": 0.8561,
+      "step": 14630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5653742325018776,
+      "learning_rate": 1.055582806743033e-07,
+      "loss": 0.9031,
+      "step": 14631
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0608882167173928,
+      "learning_rate": 1.0534653229445913e-07,
+      "loss": 0.6466,
+      "step": 14632
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2763688879586628,
+      "learning_rate": 1.0513499424877505e-07,
+      "loss": 0.5478,
+      "step": 14633
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7774251082229673,
+      "learning_rate": 1.0492366654634211e-07,
+      "loss": 0.6306,
+      "step": 14634
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.846958885949423,
+      "learning_rate": 1.0471254919624141e-07,
+      "loss": 0.6602,
+      "step": 14635
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.082845649731511,
+      "learning_rate": 1.0450164220754456e-07,
+      "loss": 1.0038,
+      "step": 14636
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.192534456693,
+      "learning_rate": 1.0429094558931485e-07,
+      "loss": 0.8183,
+      "step": 14637
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2004234134602694,
+      "learning_rate": 1.0408045935060674e-07,
+      "loss": 0.8,
+      "step": 14638
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4141046878031953,
+      "learning_rate": 1.0387018350046519e-07,
+      "loss": 1.0074,
+      "step": 14639
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.389724278554273,
+      "learning_rate": 1.0366011804792631e-07,
+      "loss": 0.6685,
+      "step": 14640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.102974818082443,
+      "learning_rate": 1.0345026300201622e-07,
+      "loss": 0.9752,
+      "step": 14641
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.224752555226548,
+      "learning_rate": 1.0324061837175436e-07,
+      "loss": 0.7455,
+      "step": 14642
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4797785103650383,
+      "learning_rate": 1.0303118416614798e-07,
+      "loss": 0.8057,
+      "step": 14643
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.657821487732203,
+      "learning_rate": 1.0282196039419823e-07,
+      "loss": 0.8607,
+      "step": 14644
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.283426476938918,
+      "learning_rate": 1.0261294706489511e-07,
+      "loss": 0.8121,
+      "step": 14645
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.292483045964111,
+      "learning_rate": 1.0240414418722088e-07,
+      "loss": 0.7783,
+      "step": 14646
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.698489763175208,
+      "learning_rate": 1.021955517701484e-07,
+      "loss": 0.7952,
+      "step": 14647
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5483673528724258,
+      "learning_rate": 1.0198716982264045e-07,
+      "loss": 0.7987,
+      "step": 14648
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.80185065200097,
+      "learning_rate": 1.0177899835365323e-07,
+      "loss": 0.7932,
+      "step": 14649
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7628248691464234,
+      "learning_rate": 1.015710373721307e-07,
+      "loss": 0.8898,
+      "step": 14650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0424017507574033,
+      "learning_rate": 1.0136328688700958e-07,
+      "loss": 0.6588,
+      "step": 14651
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4921506199203347,
+      "learning_rate": 1.0115574690721886e-07,
+      "loss": 0.714,
+      "step": 14652
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.934958892934565,
+      "learning_rate": 1.0094841744167583e-07,
+      "loss": 0.7839,
+      "step": 14653
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4768378804746316,
+      "learning_rate": 1.0074129849928948e-07,
+      "loss": 0.839,
+      "step": 14654
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6084855418910395,
+      "learning_rate": 1.005343900889616e-07,
+      "loss": 0.78,
+      "step": 14655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0824423123129525,
+      "learning_rate": 1.0032769221958227e-07,
+      "loss": 0.6674,
+      "step": 14656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.99544782045559,
+      "learning_rate": 1.0012120490003385e-07,
+      "loss": 0.7805,
+      "step": 14657
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1797423161890763,
+      "learning_rate": 9.99149281391898e-08,
+      "loss": 0.7014,
+      "step": 14658
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.458418155863635,
+      "learning_rate": 9.970886194591467e-08,
+      "loss": 0.7076,
+      "step": 14659
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2569203763783,
+      "learning_rate": 9.950300632906363e-08,
+      "loss": 0.7661,
+      "step": 14660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2478748373947226,
+      "learning_rate": 9.92973612974818e-08,
+      "loss": 0.6403,
+      "step": 14661
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3425441931993327,
+      "learning_rate": 9.909192686000713e-08,
+      "loss": 0.7313,
+      "step": 14662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.282529260982578,
+      "learning_rate": 9.888670302546754e-08,
+      "loss": 0.6863,
+      "step": 14663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1023836913885785,
+      "learning_rate": 9.8681689802681e-08,
+      "loss": 0.7508,
+      "step": 14664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4472917129113148,
+      "learning_rate": 9.847688720045878e-08,
+      "loss": 0.7827,
+      "step": 14665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.309734339714846,
+      "learning_rate": 9.827229522759996e-08,
+      "loss": 0.7926,
+      "step": 14666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9342298122576989,
+      "learning_rate": 9.806791389289861e-08,
+      "loss": 0.7342,
+      "step": 14667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2176327139426575,
+      "learning_rate": 9.786374320513547e-08,
+      "loss": 0.7465,
+      "step": 14668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1983613022478785,
+      "learning_rate": 9.765978317308522e-08,
+      "loss": 0.6914,
+      "step": 14669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.142518239352071,
+      "learning_rate": 9.745603380551194e-08,
+      "loss": 0.9323,
+      "step": 14670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.838407085644712,
+      "learning_rate": 9.72524951111714e-08,
+      "loss": 0.8037,
+      "step": 14671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8271092450663566,
+      "learning_rate": 9.704916709881052e-08,
+      "loss": 0.5844,
+      "step": 14672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9554847903096964,
+      "learning_rate": 9.684604977716616e-08,
+      "loss": 0.8404,
+      "step": 14673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3190988210992223,
+      "learning_rate": 9.664314315496692e-08,
+      "loss": 0.7824,
+      "step": 14674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0010409802147775,
+      "learning_rate": 9.644044724093249e-08,
+      "loss": 0.608,
+      "step": 14675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.630881505729617,
+      "learning_rate": 9.6237962043772e-08,
+      "loss": 0.7975,
+      "step": 14676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8591809855923223,
+      "learning_rate": 9.603568757218795e-08,
+      "loss": 0.6962,
+      "step": 14677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0193859385189596,
+      "learning_rate": 9.583362383487283e-08,
+      "loss": 0.7108,
+      "step": 14678
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9014558805596438,
+      "learning_rate": 9.5631770840508e-08,
+      "loss": 0.7442,
+      "step": 14679
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9650540827868095,
+      "learning_rate": 9.543012859776934e-08,
+      "loss": 0.8149,
+      "step": 14680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4910071698113248,
+      "learning_rate": 9.5228697115321e-08,
+      "loss": 0.8488,
+      "step": 14681
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8836359210344096,
+      "learning_rate": 9.502747640181887e-08,
+      "loss": 0.8559,
+      "step": 14682
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.670688988904194,
+      "learning_rate": 9.482646646591042e-08,
+      "loss": 0.7725,
+      "step": 14683
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.469187789386504,
+      "learning_rate": 9.462566731623213e-08,
+      "loss": 0.8382,
+      "step": 14684
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.061234538806662,
+      "learning_rate": 9.442507896141484e-08,
+      "loss": 0.9619,
+      "step": 14685
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.374394352138526,
+      "learning_rate": 9.422470141007667e-08,
+      "loss": 0.7102,
+      "step": 14686
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8453375467972357,
+      "learning_rate": 9.402453467082906e-08,
+      "loss": 0.6757,
+      "step": 14687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.430869863848779,
+      "learning_rate": 9.382457875227346e-08,
+      "loss": 0.9205,
+      "step": 14688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.038723564865097,
+      "learning_rate": 9.36248336630019e-08,
+      "loss": 0.6824,
+      "step": 14689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0032976958442323,
+      "learning_rate": 9.342529941159862e-08,
+      "loss": 0.6296,
+      "step": 14690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9964256480285033,
+      "learning_rate": 9.322597600663785e-08,
+      "loss": 0.6794,
+      "step": 14691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.150662767789757,
+      "learning_rate": 9.3026863456685e-08,
+      "loss": 0.8186,
+      "step": 14692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8904070461478544,
+      "learning_rate": 9.282796177029596e-08,
+      "loss": 0.7975,
+      "step": 14693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6765620290627927,
+      "learning_rate": 9.262927095601782e-08,
+      "loss": 0.879,
+      "step": 14694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1596998245100605,
+      "learning_rate": 9.243079102238983e-08,
+      "loss": 0.6583,
+      "step": 14695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.148391088294645,
+      "learning_rate": 9.22325219779402e-08,
+      "loss": 0.9603,
+      "step": 14696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0533526245509375,
+      "learning_rate": 9.20344638311893e-08,
+      "loss": 0.7654,
+      "step": 14697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1416321283373358,
+      "learning_rate": 9.183661659064869e-08,
+      "loss": 0.661,
+      "step": 14698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8355437211596757,
+      "learning_rate": 9.163898026481876e-08,
+      "loss": 0.8679,
+      "step": 14699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.274699302235375,
+      "learning_rate": 9.144155486219442e-08,
+      "loss": 0.64,
+      "step": 14700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9686262238304466,
+      "learning_rate": 9.124434039125774e-08,
+      "loss": 0.8942,
+      "step": 14701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9466190368851457,
+      "learning_rate": 9.104733686048417e-08,
+      "loss": 0.6683,
+      "step": 14702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.832810105581837,
+      "learning_rate": 9.085054427833972e-08,
+      "loss": 0.7555,
+      "step": 14703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.164938331149863,
+      "learning_rate": 9.065396265327986e-08,
+      "loss": 0.7882,
+      "step": 14704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1494746803900258,
+      "learning_rate": 9.045759199375393e-08,
+      "loss": 0.7889,
+      "step": 14705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.183676373842787,
+      "learning_rate": 9.02614323081985e-08,
+      "loss": 0.6976,
+      "step": 14706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7227113103692993,
+      "learning_rate": 9.006548360504463e-08,
+      "loss": 0.7899,
+      "step": 14707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9955561501297178,
+      "learning_rate": 8.986974589271169e-08,
+      "loss": 0.5957,
+      "step": 14708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.933887502427927,
+      "learning_rate": 8.967421917961072e-08,
+      "loss": 0.7926,
+      "step": 14709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.253240102326998,
+      "learning_rate": 8.947890347414556e-08,
+      "loss": 0.8912,
+      "step": 14710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.199958956353751,
+      "learning_rate": 8.928379878470727e-08,
+      "loss": 0.7588,
+      "step": 14711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.316261870457978,
+      "learning_rate": 8.908890511968083e-08,
+      "loss": 0.9536,
+      "step": 14712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0066587103350058,
+      "learning_rate": 8.889422248744173e-08,
+      "loss": 0.7533,
+      "step": 14713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1228332586894325,
+      "learning_rate": 8.869975089635552e-08,
+      "loss": 0.8957,
+      "step": 14714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.373315723185109,
+      "learning_rate": 8.85054903547794e-08,
+      "loss": 0.6762,
+      "step": 14715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.230286632784557,
+      "learning_rate": 8.831144087106003e-08,
+      "loss": 0.8152,
+      "step": 14716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.295931416956223,
+      "learning_rate": 8.811760245353739e-08,
+      "loss": 0.8735,
+      "step": 14717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.701674535223032,
+      "learning_rate": 8.792397511054097e-08,
+      "loss": 0.6113,
+      "step": 14718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.229417033582901,
+      "learning_rate": 8.773055885039072e-08,
+      "loss": 0.7652,
+      "step": 14719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3659688183162186,
+      "learning_rate": 8.753735368139838e-08,
+      "loss": 0.6694,
+      "step": 14720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3105323604188426,
+      "learning_rate": 8.734435961186782e-08,
+      "loss": 0.7438,
+      "step": 14721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8128315192180104,
+      "learning_rate": 8.715157665008967e-08,
+      "loss": 0.6816,
+      "step": 14722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.544016137696951,
+      "learning_rate": 8.69590048043506e-08,
+      "loss": 0.6745,
+      "step": 14723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7260651110460985,
+      "learning_rate": 8.676664408292457e-08,
+      "loss": 0.6987,
+      "step": 14724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2010564616568669,
+      "learning_rate": 8.657449449407884e-08,
+      "loss": 0.7298,
+      "step": 14725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1614772049829267,
+      "learning_rate": 8.63825560460696e-08,
+      "loss": 0.6534,
+      "step": 14726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0079885602138883,
+      "learning_rate": 8.619082874714413e-08,
+      "loss": 0.6366,
+      "step": 14727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.546695686802159,
+      "learning_rate": 8.599931260554417e-08,
+      "loss": 0.7947,
+      "step": 14728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.799150522377019,
+      "learning_rate": 8.580800762949704e-08,
+      "loss": 0.8014,
+      "step": 14729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9892183450619453,
+      "learning_rate": 8.561691382722337e-08,
+      "loss": 0.5545,
+      "step": 14730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6309582237280207,
+      "learning_rate": 8.542603120693716e-08,
+      "loss": 0.9651,
+      "step": 14731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0684260673334014,
+      "learning_rate": 8.523535977683905e-08,
+      "loss": 0.7075,
+      "step": 14732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.722707209532799,
+      "learning_rate": 8.504489954512363e-08,
+      "loss": 0.7225,
+      "step": 14733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.87496725821515,
+      "learning_rate": 8.485465051997488e-08,
+      "loss": 0.8135,
+      "step": 14734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7009122238784071,
+      "learning_rate": 8.466461270956794e-08,
+      "loss": 0.8248,
+      "step": 14735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.255850947369831,
+      "learning_rate": 8.447478612207072e-08,
+      "loss": 0.7749,
+      "step": 14736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8549344170986317,
+      "learning_rate": 8.428517076563836e-08,
+      "loss": 0.5827,
+      "step": 14737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2921013925357565,
+      "learning_rate": 8.409576664842045e-08,
+      "loss": 0.8913,
+      "step": 14738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6570511250517392,
+      "learning_rate": 8.39065737785566e-08,
+      "loss": 0.7242,
+      "step": 14739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7994191588999038,
+      "learning_rate": 8.371759216417474e-08,
+      "loss": 0.8131,
+      "step": 14740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1276475096238054,
+      "learning_rate": 8.352882181339783e-08,
+      "loss": 0.9603,
+      "step": 14741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6007587745643983,
+      "learning_rate": 8.334026273433659e-08,
+      "loss": 0.7431,
+      "step": 14742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0949419864619794,
+      "learning_rate": 8.315191493509456e-08,
+      "loss": 0.6691,
+      "step": 14743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.45763301190977,
+      "learning_rate": 8.296377842376524e-08,
+      "loss": 0.797,
+      "step": 14744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3423248894733946,
+      "learning_rate": 8.277585320843273e-08,
+      "loss": 0.6882,
+      "step": 14745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1184998201885805,
+      "learning_rate": 8.258813929717391e-08,
+      "loss": 0.823,
+      "step": 14746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.73989111406678,
+      "learning_rate": 8.240063669805343e-08,
+      "loss": 0.7703,
+      "step": 14747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1217172207076946,
+      "learning_rate": 8.221334541912984e-08,
+      "loss": 0.6686,
+      "step": 14748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6601057917537383,
+      "learning_rate": 8.202626546845172e-08,
+      "loss": 0.7643,
+      "step": 14749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0630047212052183,
+      "learning_rate": 8.183939685405762e-08,
+      "loss": 0.6632,
+      "step": 14750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4599716697557708,
+      "learning_rate": 8.165273958397835e-08,
+      "loss": 1.0136,
+      "step": 14751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.101843054356013,
+      "learning_rate": 8.146629366623415e-08,
+      "loss": 0.6268,
+      "step": 14752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0793355800901394,
+      "learning_rate": 8.128005910883751e-08,
+      "loss": 0.6904,
+      "step": 14753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.714799234272722,
+      "learning_rate": 8.109403591979148e-08,
+      "loss": 0.8136,
+      "step": 14754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.037049858417971,
+      "learning_rate": 8.090822410708965e-08,
+      "loss": 0.7844,
+      "step": 14755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6254121053817725,
+      "learning_rate": 8.072262367871675e-08,
+      "loss": 0.9422,
+      "step": 14756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.733812290533967,
+      "learning_rate": 8.053723464264863e-08,
+      "loss": 0.6846,
+      "step": 14757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.507594002598345,
+      "learning_rate": 8.035205700685167e-08,
+      "loss": 0.6947,
+      "step": 14758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.29791344352315,
+      "learning_rate": 8.016709077928397e-08,
+      "loss": 0.7656,
+      "step": 14759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9661087188575359,
+      "learning_rate": 7.998233596789251e-08,
+      "loss": 0.584,
+      "step": 14760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.437821976831371,
+      "learning_rate": 7.97977925806187e-08,
+      "loss": 0.7696,
+      "step": 14761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5714628494651555,
+      "learning_rate": 7.961346062539066e-08,
+      "loss": 0.6398,
+      "step": 14762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1700030148094966,
+      "learning_rate": 7.942934011013037e-08,
+      "loss": 0.8164,
+      "step": 14763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9193319933896744,
+      "learning_rate": 7.924543104275095e-08,
+      "loss": 0.6728,
+      "step": 14764
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9692197774771056,
+      "learning_rate": 7.906173343115386e-08,
+      "loss": 0.7839,
+      "step": 14765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1283455207779447,
+      "learning_rate": 7.88782472832339e-08,
+      "loss": 0.7428,
+      "step": 14766
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8220132279974965,
+      "learning_rate": 7.86949726068753e-08,
+      "loss": 0.6884,
+      "step": 14767
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.092081657508197,
+      "learning_rate": 7.851190940995402e-08,
+      "loss": 0.6719,
+      "step": 14768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9823557092866015,
+      "learning_rate": 7.832905770033705e-08,
+      "loss": 0.7101,
+      "step": 14769
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3285381408639974,
+      "learning_rate": 7.814641748588148e-08,
+      "loss": 0.8468,
+      "step": 14770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0304817104426856,
+      "learning_rate": 7.796398877443545e-08,
+      "loss": 0.6205,
+      "step": 14771
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2363243770431347,
+      "learning_rate": 7.778177157383993e-08,
+      "loss": 0.8498,
+      "step": 14772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9693997121371405,
+      "learning_rate": 7.759976589192309e-08,
+      "loss": 0.6671,
+      "step": 14773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4062819351948663,
+      "learning_rate": 7.7417971736507e-08,
+      "loss": 0.732,
+      "step": 14774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1354530124973055,
+      "learning_rate": 7.723638911540432e-08,
+      "loss": 0.8584,
+      "step": 14775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1787925007112723,
+      "learning_rate": 7.705501803641769e-08,
+      "loss": 0.736,
+      "step": 14776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.020801290160734,
+      "learning_rate": 7.687385850734086e-08,
+      "loss": 0.77,
+      "step": 14777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.133874037689054,
+      "learning_rate": 7.669291053595818e-08,
+      "loss": 0.9429,
+      "step": 14778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.239529212121544,
+      "learning_rate": 7.651217413004674e-08,
+      "loss": 0.6939,
+      "step": 14779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.81996588576532,
+      "learning_rate": 7.633164929737202e-08,
+      "loss": 0.6351,
+      "step": 14780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6307792622322297,
+      "learning_rate": 7.615133604569113e-08,
+      "loss": 0.7206,
+      "step": 14781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2888893542616766,
+      "learning_rate": 7.597123438275399e-08,
+      "loss": 0.8302,
+      "step": 14782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.499457747285492,
+      "learning_rate": 7.579134431629942e-08,
+      "loss": 0.7258,
+      "step": 14783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0831428848987317,
+      "learning_rate": 7.561166585405789e-08,
+      "loss": 0.873,
+      "step": 14784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0294343814558773,
+      "learning_rate": 7.543219900374987e-08,
+      "loss": 0.8706,
+      "step": 14785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2069638913141834,
+      "learning_rate": 7.525294377308812e-08,
+      "loss": 0.6862,
+      "step": 14786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.648346179088674,
+      "learning_rate": 7.507390016977478e-08,
+      "loss": 0.7976,
+      "step": 14787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1658468177124806,
+      "learning_rate": 7.489506820150483e-08,
+      "loss": 0.6965,
+      "step": 14788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4576937846857816,
+      "learning_rate": 7.47164478759621e-08,
+      "loss": 0.8108,
+      "step": 14789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0867831280189946,
+      "learning_rate": 7.453803920082381e-08,
+      "loss": 0.8449,
+      "step": 14790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1159888479041062,
+      "learning_rate": 7.435984218375436e-08,
+      "loss": 0.6735,
+      "step": 14791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2868696743702466,
+      "learning_rate": 7.418185683241319e-08,
+      "loss": 0.8041,
+      "step": 14792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8953998782742765,
+      "learning_rate": 7.400408315444807e-08,
+      "loss": 0.6933,
+      "step": 14793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.459606074476635,
+      "learning_rate": 7.382652115749789e-08,
+      "loss": 0.8828,
+      "step": 14794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7111001347063413,
+      "learning_rate": 7.364917084919376e-08,
+      "loss": 0.8184,
+      "step": 14795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2910760223809405,
+      "learning_rate": 7.347203223715571e-08,
+      "loss": 0.917,
+      "step": 14796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.452888103888609,
+      "learning_rate": 7.329510532899708e-08,
+      "loss": 0.8308,
+      "step": 14797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.022359991042292,
+      "learning_rate": 7.311839013231959e-08,
+      "loss": 0.7021,
+      "step": 14798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9591288448130313,
+      "learning_rate": 7.294188665471769e-08,
+      "loss": 0.6911,
+      "step": 14799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8561048575265873,
+      "learning_rate": 7.276559490377644e-08,
+      "loss": 0.6979,
+      "step": 14800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.692353387742182,
+      "learning_rate": 7.258951488707089e-08,
+      "loss": 0.8172,
+      "step": 14801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.68493177792059,
+      "learning_rate": 7.241364661216776e-08,
+      "loss": 0.6764,
+      "step": 14802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0227056341466665,
+      "learning_rate": 7.223799008662491e-08,
+      "loss": 0.9408,
+      "step": 14803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7806523624106587,
+      "learning_rate": 7.206254531799018e-08,
+      "loss": 0.9385,
+      "step": 14804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.666251126843226,
+      "learning_rate": 7.188731231380253e-08,
+      "loss": 0.8162,
+      "step": 14805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4988040613797113,
+      "learning_rate": 7.171229108159262e-08,
+      "loss": 0.8886,
+      "step": 14806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.193803332677129,
+      "learning_rate": 7.153748162888164e-08,
+      "loss": 0.8787,
+      "step": 14807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0804764383849967,
+      "learning_rate": 7.136288396318192e-08,
+      "loss": 0.6521,
+      "step": 14808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0969514072374378,
+      "learning_rate": 7.118849809199524e-08,
+      "loss": 0.7168,
+      "step": 14809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6438928101380825,
+      "learning_rate": 7.10143240228156e-08,
+      "loss": 0.9378,
+      "step": 14810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.452166225401381,
+      "learning_rate": 7.084036176312814e-08,
+      "loss": 0.9145,
+      "step": 14811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.652967411336388,
+      "learning_rate": 7.066661132040853e-08,
+      "loss": 0.813,
+      "step": 14812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3849688756172902,
+      "learning_rate": 7.049307270212246e-08,
+      "loss": 0.6116,
+      "step": 14813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1825334756204626,
+      "learning_rate": 7.031974591572732e-08,
+      "loss": 0.6241,
+      "step": 14814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0029153075601616,
+      "learning_rate": 7.014663096867214e-08,
+      "loss": 0.6518,
+      "step": 14815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.883006497248617,
+      "learning_rate": 6.99737278683954e-08,
+      "loss": 0.8565,
+      "step": 14816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9216200790311275,
+      "learning_rate": 6.980103662232728e-08,
+      "loss": 0.9841,
+      "step": 14817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2873970078591617,
+      "learning_rate": 6.962855723788909e-08,
+      "loss": 0.9144,
+      "step": 14818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0563071364174315,
+      "learning_rate": 6.945628972249208e-08,
+      "loss": 0.7107,
+      "step": 14819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.084337173875153,
+      "learning_rate": 6.928423408353979e-08,
+      "loss": 0.833,
+      "step": 14820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1810622279030845,
+      "learning_rate": 6.91123903284241e-08,
+      "loss": 0.6827,
+      "step": 14821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0051738033010542,
+      "learning_rate": 6.894075846453185e-08,
+      "loss": 0.6442,
+      "step": 14822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.090843066621107,
+      "learning_rate": 6.876933849923662e-08,
+      "loss": 0.7757,
+      "step": 14823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0298200355543066,
+      "learning_rate": 6.859813043990526e-08,
+      "loss": 0.813,
+      "step": 14824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9565973787157027,
+      "learning_rate": 6.842713429389525e-08,
+      "loss": 0.7102,
+      "step": 14825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.392242482449344,
+      "learning_rate": 6.825635006855458e-08,
+      "loss": 0.5583,
+      "step": 14826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.157890466216829,
+      "learning_rate": 6.808577777122183e-08,
+      "loss": 0.9068,
+      "step": 14827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0708922146033621,
+      "learning_rate": 6.79154174092278e-08,
+      "loss": 0.6593,
+      "step": 14828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.07734807860389,
+      "learning_rate": 6.77452689898922e-08,
+      "loss": 0.6169,
+      "step": 14829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1078254506736553,
+      "learning_rate": 6.757533252052695e-08,
+      "loss": 0.6519,
+      "step": 14830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.367319398146255,
+      "learning_rate": 6.74056080084351e-08,
+      "loss": 0.8632,
+      "step": 14831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.191308966319733,
+      "learning_rate": 6.723609546090915e-08,
+      "loss": 0.6815,
+      "step": 14832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.804699372158293,
+      "learning_rate": 6.706679488523494e-08,
+      "loss": 0.8514,
+      "step": 14833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6870995877627957,
+      "learning_rate": 6.689770628868609e-08,
+      "loss": 0.7456,
+      "step": 14834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1992848442789965,
+      "learning_rate": 6.672882967852956e-08,
+      "loss": 0.9781,
+      "step": 14835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.229644829798378,
+      "learning_rate": 6.656016506202234e-08,
+      "loss": 0.8746,
+      "step": 14836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8395521606606904,
+      "learning_rate": 6.639171244641252e-08,
+      "loss": 0.8377,
+      "step": 14837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.220853066032424,
+      "learning_rate": 6.622347183893818e-08,
+      "loss": 1.0256,
+      "step": 14838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.389927652425577,
+      "learning_rate": 6.605544324682855e-08,
+      "loss": 0.8634,
+      "step": 14839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4661925068252533,
+      "learning_rate": 6.58876266773062e-08,
+      "loss": 0.7485,
+      "step": 14840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7941556687058933,
+      "learning_rate": 6.572002213758144e-08,
+      "loss": 0.6318,
+      "step": 14841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8777446281206696,
+      "learning_rate": 6.555262963485576e-08,
+      "loss": 0.6531,
+      "step": 14842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4790018306159864,
+      "learning_rate": 6.53854491763234e-08,
+      "loss": 0.7913,
+      "step": 14843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4270107996526256,
+      "learning_rate": 6.521848076916859e-08,
+      "loss": 0.7733,
+      "step": 14844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.472936362517218,
+      "learning_rate": 6.505172442056562e-08,
+      "loss": 0.5562,
+      "step": 14845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4539648016721314,
+      "learning_rate": 6.488518013768096e-08,
+      "loss": 0.8856,
+      "step": 14846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1398433782250548,
+      "learning_rate": 6.471884792767169e-08,
+      "loss": 0.7956,
+      "step": 14847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.291698938939092,
+      "learning_rate": 6.455272779768429e-08,
+      "loss": 0.8476,
+      "step": 14848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.768355687962881,
+      "learning_rate": 6.438681975485805e-08,
+      "loss": 0.7966,
+      "step": 14849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.354179650346063,
+      "learning_rate": 6.422112380632228e-08,
+      "loss": 0.5994,
+      "step": 14850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.80492077216177,
+      "learning_rate": 6.40556399591985e-08,
+      "loss": 0.8668,
+      "step": 14851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1854555758861027,
+      "learning_rate": 6.389036822059546e-08,
+      "loss": 0.5902,
+      "step": 14852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3111570285244007,
+      "learning_rate": 6.372530859761694e-08,
+      "loss": 0.9035,
+      "step": 14853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.943376664467429,
+      "learning_rate": 6.356046109735614e-08,
+      "loss": 0.7376,
+      "step": 14854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2436998697396056,
+      "learning_rate": 6.339582572689573e-08,
+      "loss": 0.975,
+      "step": 14855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.039986446798096,
+      "learning_rate": 6.323140249331116e-08,
+      "loss": 0.8754,
+      "step": 14856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.009732596522635,
+      "learning_rate": 6.306719140366791e-08,
+      "loss": 0.6985,
+      "step": 14857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9162263722261597,
+      "learning_rate": 6.290319246502363e-08,
+      "loss": 0.881,
+      "step": 14858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3444674256694933,
+      "learning_rate": 6.273940568442327e-08,
+      "loss": 0.9598,
+      "step": 14859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9415068740616515,
+      "learning_rate": 6.25758310689073e-08,
+      "loss": 0.8633,
+      "step": 14860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.329974511719613,
+      "learning_rate": 6.241246862550398e-08,
+      "loss": 0.8981,
+      "step": 14861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.326784968287418,
+      "learning_rate": 6.224931836123327e-08,
+      "loss": 0.7533,
+      "step": 14862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1379101620080614,
+      "learning_rate": 6.208638028310677e-08,
+      "loss": 0.9025,
+      "step": 14863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9738332054028858,
+      "learning_rate": 6.192365439812553e-08,
+      "loss": 0.6984,
+      "step": 14864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8999489495764634,
+      "learning_rate": 6.176114071328287e-08,
+      "loss": 0.9759,
+      "step": 14865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3068494069605703,
+      "learning_rate": 6.159883923556154e-08,
+      "loss": 0.797,
+      "step": 14866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5383302062442272,
+      "learning_rate": 6.143674997193649e-08,
+      "loss": 0.643,
+      "step": 14867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1723690964267064,
+      "learning_rate": 6.127487292937328e-08,
+      "loss": 0.883,
+      "step": 14868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7533428097544208,
+      "learning_rate": 6.111320811482802e-08,
+      "loss": 0.6487,
+      "step": 14869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4144519754829616,
+      "learning_rate": 6.095175553524735e-08,
+      "loss": 0.7903,
+      "step": 14870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2277990358493414,
+      "learning_rate": 6.079051519757018e-08,
+      "loss": 0.817,
+      "step": 14871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.290825429245867,
+      "learning_rate": 6.062948710872373e-08,
+      "loss": 0.821,
+      "step": 14872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0229461750001456,
+      "learning_rate": 6.046867127563027e-08,
+      "loss": 0.812,
+      "step": 14873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.855003271328163,
+      "learning_rate": 6.030806770519815e-08,
+      "loss": 0.9441,
+      "step": 14874
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.187383676833998,
+      "learning_rate": 6.014767640432905e-08,
+      "loss": 0.8438,
+      "step": 14875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7127359466609913,
+      "learning_rate": 5.998749737991693e-08,
+      "loss": 0.7154,
+      "step": 14876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.445125962940918,
+      "learning_rate": 5.98275306388435e-08,
+      "loss": 0.6984,
+      "step": 14877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8671803819418087,
+      "learning_rate": 5.96677761879838e-08,
+      "loss": 0.8774,
+      "step": 14878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.114362176198585,
+      "learning_rate": 5.9508234034202364e-08,
+      "loss": 0.8551,
+      "step": 14879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2478710647826183,
+      "learning_rate": 5.9348904184354795e-08,
+      "loss": 0.8612,
+      "step": 14880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7727464617223512,
+      "learning_rate": 5.918978664528841e-08,
+      "loss": 0.9017,
+      "step": 14881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1947102638600673,
+      "learning_rate": 5.903088142384106e-08,
+      "loss": 0.8443,
+      "step": 14882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.222778471376077,
+      "learning_rate": 5.887218852684007e-08,
+      "loss": 0.8746,
+      "step": 14883
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1972951554828777,
+      "learning_rate": 5.871370796110665e-08,
+      "loss": 0.853,
+      "step": 14884
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2495390228620176,
+      "learning_rate": 5.855543973344924e-08,
+      "loss": 0.6935,
+      "step": 14885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.368738733374209,
+      "learning_rate": 5.8397383850670176e-08,
+      "loss": 0.8316,
+      "step": 14886
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9395474149524672,
+      "learning_rate": 5.823954031956069e-08,
+      "loss": 0.8636,
+      "step": 14887
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0941494081850465,
+      "learning_rate": 5.808190914690426e-08,
+      "loss": 0.9301,
+      "step": 14888
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5298419337105895,
+      "learning_rate": 5.7924490339474335e-08,
+      "loss": 0.956,
+      "step": 14889
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2641833773721243,
+      "learning_rate": 5.776728390403552e-08,
+      "loss": 0.8099,
+      "step": 14890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7693851927683713,
+      "learning_rate": 5.761028984734407e-08,
+      "loss": 0.8747,
+      "step": 14891
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1749738988417706,
+      "learning_rate": 5.74535081761457e-08,
+      "loss": 0.5476,
+      "step": 14892
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2741590061041452,
+      "learning_rate": 5.7296938897177245e-08,
+      "loss": 0.8025,
+      "step": 14893
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0912365866021907,
+      "learning_rate": 5.7140582017167764e-08,
+      "loss": 0.698,
+      "step": 14894
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8996620384637297,
+      "learning_rate": 5.6984437542835225e-08,
+      "loss": 0.8453,
+      "step": 14895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.077258083480282,
+      "learning_rate": 5.682850548089036e-08,
+      "loss": 0.8849,
+      "step": 14896
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.183966195497629,
+      "learning_rate": 5.6672785838033925e-08,
+      "loss": 0.8789,
+      "step": 14897
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.775130285463277,
+      "learning_rate": 5.651727862095668e-08,
+      "loss": 0.7262,
+      "step": 14898
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.772215545341934,
+      "learning_rate": 5.636198383634217e-08,
+      "loss": 0.8808,
+      "step": 14899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9198653248102118,
+      "learning_rate": 5.620690149086283e-08,
+      "loss": 0.7353,
+      "step": 14900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5720973047228486,
+      "learning_rate": 5.6052031591183884e-08,
+      "loss": 0.7334,
+      "step": 14901
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0489569496971844,
+      "learning_rate": 5.5897374143960013e-08,
+      "loss": 0.9118,
+      "step": 14902
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1427860856859597,
+      "learning_rate": 5.574292915583646e-08,
+      "loss": 1.068,
+      "step": 14903
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9184755857627527,
+      "learning_rate": 5.558869663345123e-08,
+      "loss": 0.7453,
+      "step": 14904
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1100556760450546,
+      "learning_rate": 5.543467658343127e-08,
+      "loss": 0.8709,
+      "step": 14905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.592501324611749,
+      "learning_rate": 5.5280869012395154e-08,
+      "loss": 0.6873,
+      "step": 14906
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1360154235735966,
+      "learning_rate": 5.5127273926953165e-08,
+      "loss": 0.8744,
+      "step": 14907
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.401908804869561,
+      "learning_rate": 5.4973891333704453e-08,
+      "loss": 0.8798,
+      "step": 14908
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7475940966266506,
+      "learning_rate": 5.482072123924098e-08,
+      "loss": 0.7847,
+      "step": 14909
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.483729750343021,
+      "learning_rate": 5.46677636501447e-08,
+      "loss": 0.6987,
+      "step": 14910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8648373275409393,
+      "learning_rate": 5.4515018572988687e-08,
+      "loss": 0.9014,
+      "step": 14911
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8266082524958793,
+      "learning_rate": 5.436248601433658e-08,
+      "loss": 0.7585,
+      "step": 14912
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1726258923611816,
+      "learning_rate": 5.4210165980742025e-08,
+      "loss": 0.7594,
+      "step": 14913
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4667982205699417,
+      "learning_rate": 5.405805847875256e-08,
+      "loss": 0.7453,
+      "step": 14914
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.546594902148417,
+      "learning_rate": 5.3906163514902966e-08,
+      "loss": 0.9304,
+      "step": 14915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.516473645128542,
+      "learning_rate": 5.3754481095721344e-08,
+      "loss": 0.6867,
+      "step": 14916
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.056083275975385,
+      "learning_rate": 5.3603011227725265e-08,
+      "loss": 0.9162,
+      "step": 14917
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0982147390864814,
+      "learning_rate": 5.3451753917423964e-08,
+      "loss": 0.7556,
+      "step": 14918
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.104720347257029,
+      "learning_rate": 5.330070917131724e-08,
+      "loss": 0.9017,
+      "step": 14919
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0526536968293556,
+      "learning_rate": 5.3149876995896575e-08,
+      "loss": 0.6961,
+      "step": 14920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.14984531954375,
+      "learning_rate": 5.299925739764233e-08,
+      "loss": 0.7799,
+      "step": 14921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3166440004194557,
+      "learning_rate": 5.284885038302823e-08,
+      "loss": 0.7841,
+      "step": 14922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0502099014277597,
+      "learning_rate": 5.269865595851631e-08,
+      "loss": 0.7916,
+      "step": 14923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.456318696795234,
+      "learning_rate": 5.2548674130561974e-08,
+      "loss": 0.7709,
+      "step": 14924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5601427617124277,
+      "learning_rate": 5.239890490560895e-08,
+      "loss": 0.6682,
+      "step": 14925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3006049694331403,
+      "learning_rate": 5.224934829009376e-08,
+      "loss": 0.9309,
+      "step": 14926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.991883202165902,
+      "learning_rate": 5.21000042904446e-08,
+      "loss": 0.9027,
+      "step": 14927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8318422368975473,
+      "learning_rate": 5.195087291307632e-08,
+      "loss": 0.7799,
+      "step": 14928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5211119950149787,
+      "learning_rate": 5.1801954164399925e-08,
+      "loss": 0.9735,
+      "step": 14929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2324106058524524,
+      "learning_rate": 5.165324805081362e-08,
+      "loss": 0.9614,
+      "step": 14930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.002328323856843,
+      "learning_rate": 5.1504754578707294e-08,
+      "loss": 0.9594,
+      "step": 14931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7408885350568726,
+      "learning_rate": 5.135647375446307e-08,
+      "loss": 0.6471,
+      "step": 14932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1937875739622923,
+      "learning_rate": 5.120840558445195e-08,
+      "loss": 0.7144,
+      "step": 14933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.340741687887237,
+      "learning_rate": 5.106055007503774e-08,
+      "loss": 0.8294,
+      "step": 14934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6813158909058794,
+      "learning_rate": 5.091290723257314e-08,
+      "loss": 0.8047,
+      "step": 14935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.781457639431416,
+      "learning_rate": 5.076547706340307e-08,
+      "loss": 0.7219,
+      "step": 14936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4273414495348136,
+      "learning_rate": 5.061825957386302e-08,
+      "loss": 0.8837,
+      "step": 14937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8416633237654174,
+      "learning_rate": 5.047125477027959e-08,
+      "loss": 0.7259,
+      "step": 14938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1564091274322887,
+      "learning_rate": 5.0324462658969395e-08,
+      "loss": 0.8141,
+      "step": 14939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.337035087704094,
+      "learning_rate": 5.017788324624073e-08,
+      "loss": 0.5967,
+      "step": 14940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4658268959742022,
+      "learning_rate": 5.003151653839189e-08,
+      "loss": 0.8749,
+      "step": 14941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0218495507229397,
+      "learning_rate": 4.988536254171284e-08,
+      "loss": 0.7364,
+      "step": 14942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.335019114419379,
+      "learning_rate": 4.9739421262484656e-08,
+      "loss": 0.7348,
+      "step": 14943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2871290420795716,
+      "learning_rate": 4.959369270697789e-08,
+      "loss": 0.6431,
+      "step": 14944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.195419775272606,
+      "learning_rate": 4.944817688145642e-08,
+      "loss": 0.7239,
+      "step": 14945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2346805147274025,
+      "learning_rate": 4.93028737921708e-08,
+      "loss": 0.7581,
+      "step": 14946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.653236364706129,
+      "learning_rate": 4.915778344536715e-08,
+      "loss": 0.7197,
+      "step": 14947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.887486288864111,
+      "learning_rate": 4.9012905847279913e-08,
+      "loss": 0.8755,
+      "step": 14948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3578871196417412,
+      "learning_rate": 4.886824100413412e-08,
+      "loss": 0.8001,
+      "step": 14949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.305953850159673,
+      "learning_rate": 4.872378892214702e-08,
+      "loss": 0.873,
+      "step": 14950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.783891385213295,
+      "learning_rate": 4.8579549607525866e-08,
+      "loss": 0.5633,
+      "step": 14951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9976292639271485,
+      "learning_rate": 4.843552306646904e-08,
+      "loss": 0.7825,
+      "step": 14952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6160446025030684,
+      "learning_rate": 4.829170930516491e-08,
+      "loss": 0.6872,
+      "step": 14953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2084593988928547,
+      "learning_rate": 4.814810832979411e-08,
+      "loss": 0.6776,
+      "step": 14954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.091570063606529,
+      "learning_rate": 4.8004720146527795e-08,
+      "loss": 0.8217,
+      "step": 14955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.817687559661137,
+      "learning_rate": 4.786154476152716e-08,
+      "loss": 0.8708,
+      "step": 14956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0976303401635135,
+      "learning_rate": 4.771858218094505e-08,
+      "loss": 0.8084,
+      "step": 14957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5152371674738356,
+      "learning_rate": 4.7575832410924893e-08,
+      "loss": 0.9531,
+      "step": 14958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.024859303632297,
+      "learning_rate": 4.743329545760122e-08,
+      "loss": 0.6597,
+      "step": 14959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.470718030358735,
+      "learning_rate": 4.7290971327098035e-08,
+      "loss": 0.8082,
+      "step": 14960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.216898235840108,
+      "learning_rate": 4.71488600255321e-08,
+      "loss": 0.8812,
+      "step": 14961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9681283273247305,
+      "learning_rate": 4.7006961559010745e-08,
+      "loss": 0.8962,
+      "step": 14962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6434071924481364,
+      "learning_rate": 4.686527593363133e-08,
+      "loss": 0.7456,
+      "step": 14963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3823895076676012,
+      "learning_rate": 4.67238031554812e-08,
+      "loss": 0.6985,
+      "step": 14964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.048809116080571,
+      "learning_rate": 4.6582543230641596e-08,
+      "loss": 0.805,
+      "step": 14965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.144863966470096,
+      "learning_rate": 4.644149616518212e-08,
+      "loss": 0.7327,
+      "step": 14966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9071590145460484,
+      "learning_rate": 4.630066196516347e-08,
+      "loss": 0.7936,
+      "step": 14967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.37792173025886,
+      "learning_rate": 4.6160040636637484e-08,
+      "loss": 0.8989,
+      "step": 14968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.955982871172612,
+      "learning_rate": 4.6019632185647645e-08,
+      "loss": 0.9643,
+      "step": 14969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.819616985005159,
+      "learning_rate": 4.587943661822747e-08,
+      "loss": 0.74,
+      "step": 14970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0629659337146755,
+      "learning_rate": 4.5739453940401024e-08,
+      "loss": 0.6788,
+      "step": 14971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6558358719915578,
+      "learning_rate": 4.5599684158184055e-08,
+      "loss": 0.7778,
+      "step": 14972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8813998329735693,
+      "learning_rate": 4.5460127277582863e-08,
+      "loss": 0.8085,
+      "step": 14973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.463910199045444,
+      "learning_rate": 4.532078330459433e-08,
+      "loss": 0.9524,
+      "step": 14974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.270637535164914,
+      "learning_rate": 4.518165224520643e-08,
+      "loss": 0.9511,
+      "step": 14975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5183440708818785,
+      "learning_rate": 4.504273410539772e-08,
+      "loss": 0.7217,
+      "step": 14976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0183243509742224,
+      "learning_rate": 4.490402889113787e-08,
+      "loss": 0.6979,
+      "step": 14977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3852619816097844,
+      "learning_rate": 4.4765536608388227e-08,
+      "loss": 0.6594,
+      "step": 14978
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.127701186110304,
+      "learning_rate": 4.4627257263098465e-08,
+      "loss": 0.8487,
+      "step": 14979
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.728585537494874,
+      "learning_rate": 4.448919086121217e-08,
+      "loss": 0.7671,
+      "step": 14980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2553119836390552,
+      "learning_rate": 4.435133740866182e-08,
+      "loss": 0.7324,
+      "step": 14981
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3126053620249145,
+      "learning_rate": 4.4213696911371006e-08,
+      "loss": 0.8752,
+      "step": 14982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8042678314354843,
+      "learning_rate": 4.4076269375255e-08,
+      "loss": 0.65,
+      "step": 14983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.354507229337422,
+      "learning_rate": 4.393905480621907e-08,
+      "loss": 0.7135,
+      "step": 14984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2689264156267988,
+      "learning_rate": 4.380205321016018e-08,
+      "loss": 0.7616,
+      "step": 14985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7164037349874506,
+      "learning_rate": 4.366526459296472e-08,
+      "loss": 0.6729,
+      "step": 14986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0124176229129849,
+      "learning_rate": 4.352868896051077e-08,
+      "loss": 0.6319,
+      "step": 14987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.179361599029844,
+      "learning_rate": 4.3392326318668076e-08,
+      "loss": 0.5791,
+      "step": 14988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.197136594955526,
+      "learning_rate": 4.3256176673295846e-08,
+      "loss": 0.8288,
+      "step": 14989
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1562152238759777,
+      "learning_rate": 4.312024003024551e-08,
+      "loss": 0.7378,
+      "step": 14990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1109916833261448,
+      "learning_rate": 4.2984516395357944e-08,
+      "loss": 0.6292,
+      "step": 14991
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.076403288329859,
+      "learning_rate": 4.284900577446516e-08,
+      "loss": 0.6899,
+      "step": 14992
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.87450399725029,
+      "learning_rate": 4.271370817339082e-08,
+      "loss": 0.8028,
+      "step": 14993
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8913193430624502,
+      "learning_rate": 4.2578623597949174e-08,
+      "loss": 0.7685,
+      "step": 14994
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.653358225974742,
+      "learning_rate": 4.244375205394502e-08,
+      "loss": 0.8774,
+      "step": 14995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.052626147224303,
+      "learning_rate": 4.230909354717372e-08,
+      "loss": 0.7723,
+      "step": 14996
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.850923343017498,
+      "learning_rate": 4.217464808342175e-08,
+      "loss": 0.9113,
+      "step": 14997
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.31015230406616,
+      "learning_rate": 4.204041566846673e-08,
+      "loss": 0.943,
+      "step": 14998
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.298734001088268,
+      "learning_rate": 4.1906396308077356e-08,
+      "loss": 0.8905,
+      "step": 14999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.838777654843412,
+      "learning_rate": 4.177259000801237e-08,
+      "loss": 0.9108,
+      "step": 15000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.635488604659508,
+      "learning_rate": 4.163899677402161e-08,
+      "loss": 0.5785,
+      "step": 15001
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9254509655414132,
+      "learning_rate": 4.150561661184604e-08,
+      "loss": 0.9202,
+      "step": 15002
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.538178563223243,
+      "learning_rate": 4.1372449527217754e-08,
+      "loss": 0.7673,
+      "step": 15003
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.899348942865919,
+      "learning_rate": 4.123949552585826e-08,
+      "loss": 0.7317,
+      "step": 15004
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3700371045032282,
+      "learning_rate": 4.1106754613481347e-08,
+      "loss": 0.8914,
+      "step": 15005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3511292626482234,
+      "learning_rate": 4.097422679579188e-08,
+      "loss": 0.7254,
+      "step": 15006
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3428932717633004,
+      "learning_rate": 4.0841912078483645e-08,
+      "loss": 0.5611,
+      "step": 15007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.491344927784078,
+      "learning_rate": 4.0709810467243204e-08,
+      "loss": 0.8036,
+      "step": 15008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.238506632993291,
+      "learning_rate": 4.0577921967747126e-08,
+      "loss": 0.8586,
+      "step": 15009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.292600664686858,
+      "learning_rate": 4.0446246585663093e-08,
+      "loss": 0.85,
+      "step": 15010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6088663788567117,
+      "learning_rate": 4.031478432664937e-08,
+      "loss": 0.9196,
+      "step": 15011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.682667797484833,
+      "learning_rate": 4.018353519635476e-08,
+      "loss": 0.7417,
+      "step": 15012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.113316507863263,
+      "learning_rate": 4.0052499200420314e-08,
+      "loss": 1.0303,
+      "step": 15013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1344918811284557,
+      "learning_rate": 3.9921676344475966e-08,
+      "loss": 0.6023,
+      "step": 15014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1302993740923415,
+      "learning_rate": 3.979106663414389e-08,
+      "loss": 0.8313,
+      "step": 15015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.743909602060734,
+      "learning_rate": 3.9660670075036823e-08,
+      "loss": 0.6399,
+      "step": 15016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0180601088817562,
+      "learning_rate": 3.953048667275805e-08,
+      "loss": 0.6721,
+      "step": 15017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2245743863211622,
+      "learning_rate": 3.9400516432901434e-08,
+      "loss": 0.8789,
+      "step": 15018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4616239272214218,
+      "learning_rate": 3.927075936105307e-08,
+      "loss": 0.8884,
+      "step": 15019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0435184150233474,
+      "learning_rate": 3.914121546278793e-08,
+      "loss": 0.776,
+      "step": 15020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.951712704558825,
+      "learning_rate": 3.901188474367323e-08,
+      "loss": 0.9621,
+      "step": 15021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.541750497673153,
+      "learning_rate": 3.8882767209266756e-08,
+      "loss": 0.7023,
+      "step": 15022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9132980091623086,
+      "learning_rate": 3.875386286511684e-08,
+      "loss": 0.6928,
+      "step": 15023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0240926984840488,
+      "learning_rate": 3.8625171716762385e-08,
+      "loss": 0.6592,
+      "step": 15024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.951452731479069,
+      "learning_rate": 3.849669376973397e-08,
+      "loss": 0.6363,
+      "step": 15025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.316084568659973,
+      "learning_rate": 3.836842902955273e-08,
+      "loss": 0.5773,
+      "step": 15026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.371779060361091,
+      "learning_rate": 3.8240377501729817e-08,
+      "loss": 0.6941,
+      "step": 15027
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.351330903122871,
+      "learning_rate": 3.8112539191768604e-08,
+      "loss": 0.7694,
+      "step": 15028
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9770611620490678,
+      "learning_rate": 3.7984914105162474e-08,
+      "loss": 0.6653,
+      "step": 15029
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.226951760920081,
+      "learning_rate": 3.785750224739482e-08,
+      "loss": 0.653,
+      "step": 15030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.556733357120608,
+      "learning_rate": 3.7730303623942364e-08,
+      "loss": 0.6981,
+      "step": 15031
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9010150424392416,
+      "learning_rate": 3.760331824027019e-08,
+      "loss": 0.8242,
+      "step": 15032
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1293275216256238,
+      "learning_rate": 3.7476546101835045e-08,
+      "loss": 0.6267,
+      "step": 15033
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0978937092809824,
+      "learning_rate": 3.7349987214084784e-08,
+      "loss": 0.6442,
+      "step": 15034
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.027508327129086,
+      "learning_rate": 3.722364158245839e-08,
+      "loss": 0.8655,
+      "step": 15035
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4323011463621698,
+      "learning_rate": 3.709750921238486e-08,
+      "loss": 1.0803,
+      "step": 15036
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1127341919049836,
+      "learning_rate": 3.697159010928375e-08,
+      "loss": 0.7012,
+      "step": 15037
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9157041107786164,
+      "learning_rate": 3.6845884278566836e-08,
+      "loss": 0.6629,
+      "step": 15038
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.695645044956028,
+      "learning_rate": 3.672039172563646e-08,
+      "loss": 0.6809,
+      "step": 15039
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.009308435450318,
+      "learning_rate": 3.659511245588387e-08,
+      "loss": 0.9142,
+      "step": 15040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3067070740817233,
+      "learning_rate": 3.647004647469365e-08,
+      "loss": 0.8345,
+      "step": 15041
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6540653116623516,
+      "learning_rate": 3.634519378744039e-08,
+      "loss": 0.8534,
+      "step": 15042
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1646864473020129,
+      "learning_rate": 3.622055439948813e-08,
+      "loss": 0.6586,
+      "step": 15043
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9570518952727204,
+      "learning_rate": 3.609612831619369e-08,
+      "loss": 0.5597,
+      "step": 15044
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4362802355173567,
+      "learning_rate": 3.597191554290391e-08,
+      "loss": 0.8461,
+      "step": 15045
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1894951093658186,
+      "learning_rate": 3.584791608495675e-08,
+      "loss": 0.8581,
+      "step": 15046
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1278204035304267,
+      "learning_rate": 3.5724129947680156e-08,
+      "loss": 0.708,
+      "step": 15047
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.038911791726629,
+      "learning_rate": 3.560055713639321e-08,
+      "loss": 0.7448,
+      "step": 15048
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.073348944002343,
+      "learning_rate": 3.547719765640778e-08,
+      "loss": 0.8902,
+      "step": 15049
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2071235034848875,
+      "learning_rate": 3.5354051513022405e-08,
+      "loss": 0.665,
+      "step": 15050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9505023793438314,
+      "learning_rate": 3.523111871153118e-08,
+      "loss": 0.7701,
+      "step": 15051
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4732492837553772,
+      "learning_rate": 3.510839925721543e-08,
+      "loss": 0.6447,
+      "step": 15052
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.622050427289953,
+      "learning_rate": 3.4985893155349835e-08,
+      "loss": 0.7417,
+      "step": 15053
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9242907385377475,
+      "learning_rate": 3.4863600411197404e-08,
+      "loss": 0.7257,
+      "step": 15054
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9354243141337324,
+      "learning_rate": 3.474152103001449e-08,
+      "loss": 0.8575,
+      "step": 15055
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1115337632635836,
+      "learning_rate": 3.461965501704634e-08,
+      "loss": 0.6438,
+      "step": 15056
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.864079051240804,
+      "learning_rate": 3.449800237753043e-08,
+      "loss": 0.8942,
+      "step": 15057
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.304604408779942,
+      "learning_rate": 3.4376563116693685e-08,
+      "loss": 0.9402,
+      "step": 15058
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.51732690507747,
+      "learning_rate": 3.425533723975527e-08,
+      "loss": 0.8852,
+      "step": 15059
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0476612908800256,
+      "learning_rate": 3.413432475192435e-08,
+      "loss": 0.9239,
+      "step": 15060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7063460594262816,
+      "learning_rate": 3.401352565840121e-08,
+      "loss": 0.8423,
+      "step": 15061
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0776084471162526,
+      "learning_rate": 3.3892939964376704e-08,
+      "loss": 0.6795,
+      "step": 15062
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4122005264266617,
+      "learning_rate": 3.377256767503223e-08,
+      "loss": 0.7909,
+      "step": 15063
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1797544987557984,
+      "learning_rate": 3.365240879554144e-08,
+      "loss": 0.8034,
+      "step": 15064
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6156161899337427,
+      "learning_rate": 3.353246333106741e-08,
+      "loss": 0.9461,
+      "step": 15065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.075304989236389,
+      "learning_rate": 3.3412731286763256e-08,
+      "loss": 0.7901,
+      "step": 15066
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3907218753242416,
+      "learning_rate": 3.329321266777652e-08,
+      "loss": 0.8706,
+      "step": 15067
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3636273454298413,
+      "learning_rate": 3.3173907479240875e-08,
+      "loss": 0.8036,
+      "step": 15068
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1760921554415646,
+      "learning_rate": 3.3054815726285e-08,
+      "loss": 0.6899,
+      "step": 15069
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1426727513280506,
+      "learning_rate": 3.2935937414025345e-08,
+      "loss": 0.697,
+      "step": 15070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.332762300386892,
+      "learning_rate": 3.281727254757061e-08,
+      "loss": 0.8147,
+      "step": 15071
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.283121361868221,
+      "learning_rate": 3.269882113202005e-08,
+      "loss": 1.0286,
+      "step": 15072
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.021623514593908,
+      "learning_rate": 3.258058317246404e-08,
+      "loss": 0.7779,
+      "step": 15073
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.013525105123009,
+      "learning_rate": 3.2462558673983516e-08,
+      "loss": 0.9508,
+      "step": 15074
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2380967255677833,
+      "learning_rate": 3.234474764165052e-08,
+      "loss": 0.8136,
+      "step": 15075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6273696898483334,
+      "learning_rate": 3.222715008052657e-08,
+      "loss": 0.96,
+      "step": 15076
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1408256101841745,
+      "learning_rate": 3.2109765995665957e-08,
+      "loss": 0.667,
+      "step": 15077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9531465699267716,
+      "learning_rate": 3.1992595392112966e-08,
+      "loss": 0.6685,
+      "step": 15078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8495333661287,
+      "learning_rate": 3.1875638274902476e-08,
+      "loss": 0.9666,
+      "step": 15079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1382557473488364,
+      "learning_rate": 3.1758894649060456e-08,
+      "loss": 0.6767,
+      "step": 15080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1267837165944803,
+      "learning_rate": 3.164236451960345e-08,
+      "loss": 0.8469,
+      "step": 15081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3150966242564923,
+      "learning_rate": 3.152604789153912e-08,
+      "loss": 0.7167,
+      "step": 15082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9530618598055756,
+      "learning_rate": 3.140994476986625e-08,
+      "loss": 0.7922,
+      "step": 15083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2914916629563273,
+      "learning_rate": 3.129405515957307e-08,
+      "loss": 0.9288,
+      "step": 15084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3119785251356797,
+      "learning_rate": 3.117837906564114e-08,
+      "loss": 0.6394,
+      "step": 15085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9183689473938688,
+      "learning_rate": 3.106291649303983e-08,
+      "loss": 0.8199,
+      "step": 15086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.172204129904611,
+      "learning_rate": 3.094766744673128e-08,
+      "loss": 0.9008,
+      "step": 15087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.527580984168911,
+      "learning_rate": 3.083263193166819e-08,
+      "loss": 0.6501,
+      "step": 15088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0470436139699664,
+      "learning_rate": 3.071780995279439e-08,
+      "loss": 0.7427,
+      "step": 15089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1363524003550527,
+      "learning_rate": 3.060320151504315e-08,
+      "loss": 0.6664,
+      "step": 15090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9616652549310634,
+      "learning_rate": 3.0488806623339414e-08,
+      "loss": 0.678,
+      "step": 15091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0953553667618143,
+      "learning_rate": 3.0374625282599826e-08,
+      "loss": 0.6474,
+      "step": 15092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1621097672711698,
+      "learning_rate": 3.026065749773044e-08,
+      "loss": 0.8637,
+      "step": 15093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6434554240038675,
+      "learning_rate": 3.014690327362846e-08,
+      "loss": 0.9316,
+      "step": 15094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1991525617742607,
+      "learning_rate": 3.003336261518275e-08,
+      "loss": 0.92,
+      "step": 15095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9266102806699195,
+      "learning_rate": 2.992003552727219e-08,
+      "loss": 0.9307,
+      "step": 15096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.840765067336326,
+      "learning_rate": 2.980692201476676e-08,
+      "loss": 0.7643,
+      "step": 15097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6038926085311074,
+      "learning_rate": 2.969402208252703e-08,
+      "loss": 0.9656,
+      "step": 15098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4637536720574578,
+      "learning_rate": 2.9581335735404672e-08,
+      "loss": 0.9874,
+      "step": 15099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.615372231715072,
+      "learning_rate": 2.9468862978241917e-08,
+      "loss": 0.8389,
+      "step": 15100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3615394055269165,
+      "learning_rate": 2.9356603815872132e-08,
+      "loss": 0.8417,
+      "step": 15101
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4081556318212085,
+      "learning_rate": 2.9244558253119226e-08,
+      "loss": 0.8594,
+      "step": 15102
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.022537508313022,
+      "learning_rate": 2.9132726294798242e-08,
+      "loss": 0.8392,
+      "step": 15103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9834102424615088,
+      "learning_rate": 2.9021107945714777e-08,
+      "loss": 0.6833,
+      "step": 15104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4595162746297285,
+      "learning_rate": 2.8909703210664995e-08,
+      "loss": 0.9108,
+      "step": 15105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.186371794822528,
+      "learning_rate": 2.8798512094436738e-08,
+      "loss": 0.8603,
+      "step": 15106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2187153758103206,
+      "learning_rate": 2.8687534601807842e-08,
+      "loss": 0.7827,
+      "step": 15107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1352092604488844,
+      "learning_rate": 2.8576770737547276e-08,
+      "loss": 0.6989,
+      "step": 15108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.545550962243347,
+      "learning_rate": 2.8466220506414565e-08,
+      "loss": 0.8679,
+      "step": 15109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1152675142671598,
+      "learning_rate": 2.8355883913160907e-08,
+      "loss": 1.0036,
+      "step": 15110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.261031341884196,
+      "learning_rate": 2.8245760962527514e-08,
+      "loss": 0.7333,
+      "step": 15111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8553803605773287,
+      "learning_rate": 2.813585165924615e-08,
+      "loss": 0.6247,
+      "step": 15112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2061812526346443,
+      "learning_rate": 2.802615600804026e-08,
+      "loss": 0.7728,
+      "step": 15113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3808572468142835,
+      "learning_rate": 2.79166740136233e-08,
+      "loss": 0.8356,
+      "step": 15114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3501724242243585,
+      "learning_rate": 2.780740568070095e-08,
+      "loss": 0.7641,
+      "step": 15115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7716623981430748,
+      "learning_rate": 2.769835101396723e-08,
+      "loss": 0.7894,
+      "step": 15116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.91041316119356,
+      "learning_rate": 2.75895100181095e-08,
+      "loss": 0.7755,
+      "step": 15117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3137523553564896,
+      "learning_rate": 2.7480882697805133e-08,
+      "loss": 0.8997,
+      "step": 15118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.210147280407061,
+      "learning_rate": 2.7372469057721506e-08,
+      "loss": 0.9768,
+      "step": 15119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2118814257902577,
+      "learning_rate": 2.7264269102517117e-08,
+      "loss": 0.8808,
+      "step": 15120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0424824917918714,
+      "learning_rate": 2.7156282836842683e-08,
+      "loss": 0.7102,
+      "step": 15121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9187849788470939,
+      "learning_rate": 2.704851026533728e-08,
+      "loss": 0.5902,
+      "step": 15122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.322385707480836,
+      "learning_rate": 2.6940951392632753e-08,
+      "loss": 0.8736,
+      "step": 15123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9306945790843721,
+      "learning_rate": 2.6833606223351515e-08,
+      "loss": 0.7613,
+      "step": 15124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0306925048204398,
+      "learning_rate": 2.6726474762105992e-08,
+      "loss": 0.5652,
+      "step": 15125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.158367632192167,
+      "learning_rate": 2.661955701349972e-08,
+      "loss": 0.8171,
+      "step": 15126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3017328241203854,
+      "learning_rate": 2.6512852982127357e-08,
+      "loss": 0.6937,
+      "step": 15127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2556304793066275,
+      "learning_rate": 2.6406362672574125e-08,
+      "loss": 0.6146,
+      "step": 15128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.018347120534909,
+      "learning_rate": 2.6300086089416366e-08,
+      "loss": 0.8461,
+      "step": 15129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.580767555675418,
+      "learning_rate": 2.6194023237221533e-08,
+      "loss": 0.4961,
+      "step": 15130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.145299717594839,
+      "learning_rate": 2.6088174120545985e-08,
+      "loss": 0.7164,
+      "step": 15131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2583839841761595,
+      "learning_rate": 2.5982538743939413e-08,
+      "loss": 0.8833,
+      "step": 15132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.76856094597163,
+      "learning_rate": 2.5877117111940964e-08,
+      "loss": 0.5561,
+      "step": 15133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.704571697346991,
+      "learning_rate": 2.577190922908035e-08,
+      "loss": 0.7055,
+      "step": 15134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.208510805872625,
+      "learning_rate": 2.5666915099879508e-08,
+      "loss": 0.7527,
+      "step": 15135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0719721829619053,
+      "learning_rate": 2.556213472884983e-08,
+      "loss": 0.6234,
+      "step": 15136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2856494767101085,
+      "learning_rate": 2.5457568120493825e-08,
+      "loss": 0.7358,
+      "step": 15137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5311670626568787,
+      "learning_rate": 2.5353215279304566e-08,
+      "loss": 1.028,
+      "step": 15138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9818111798506164,
+      "learning_rate": 2.5249076209767353e-08,
+      "loss": 0.8943,
+      "step": 15139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.219371172323477,
+      "learning_rate": 2.5145150916356386e-08,
+      "loss": 0.7826,
+      "step": 15140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7849322624142245,
+      "learning_rate": 2.5041439403537537e-08,
+      "loss": 0.8732,
+      "step": 15141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9814235198109902,
+      "learning_rate": 2.493794167576835e-08,
+      "loss": 0.805,
+      "step": 15142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.96175214724821,
+      "learning_rate": 2.4834657737495828e-08,
+      "loss": 0.7955,
+      "step": 15143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.095946145331698,
+      "learning_rate": 2.473158759315808e-08,
+      "loss": 0.6387,
+      "step": 15144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0095125090249435,
+      "learning_rate": 2.4628731247184902e-08,
+      "loss": 0.7844,
+      "step": 15145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.186066824965035,
+      "learning_rate": 2.452608870399553e-08,
+      "loss": 0.6624,
+      "step": 15146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0063637009633495,
+      "learning_rate": 2.4423659968001446e-08,
+      "loss": 0.7216,
+      "step": 15147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1103901737836224,
+      "learning_rate": 2.4321445043603565e-08,
+      "loss": 0.8032,
+      "step": 15148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2075708432183814,
+      "learning_rate": 2.421944393519504e-08,
+      "loss": 0.7561,
+      "step": 15149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.271309678785854,
+      "learning_rate": 2.411765664715793e-08,
+      "loss": 1.0275,
+      "step": 15150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.087566562894347,
+      "learning_rate": 2.401608318386761e-08,
+      "loss": 0.5791,
+      "step": 15151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.861535006265934,
+      "learning_rate": 2.3914723549687825e-08,
+      "loss": 0.9567,
+      "step": 15152
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2555891707293947,
+      "learning_rate": 2.381357774897508e-08,
+      "loss": 0.6941,
+      "step": 15153
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.697017030294973,
+      "learning_rate": 2.3712645786075905e-08,
+      "loss": 0.9141,
+      "step": 15154
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.057782202259724,
+      "learning_rate": 2.3611927665326275e-08,
+      "loss": 0.8763,
+      "step": 15155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1289168088457258,
+      "learning_rate": 2.3511423391055498e-08,
+      "loss": 0.8023,
+      "step": 15156
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3130818911283813,
+      "learning_rate": 2.341113296758235e-08,
+      "loss": 0.7249,
+      "step": 15157
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.362466967442084,
+      "learning_rate": 2.3311056399216158e-08,
+      "loss": 0.8709,
+      "step": 15158
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6975907351381636,
+      "learning_rate": 2.3211193690257373e-08,
+      "loss": 0.819,
+      "step": 15159
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.618968199477126,
+      "learning_rate": 2.3111544844997003e-08,
+      "loss": 0.7567,
+      "step": 15160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.575633770584881,
+      "learning_rate": 2.3012109867718845e-08,
+      "loss": 0.7442,
+      "step": 15161
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.481464475555427,
+      "learning_rate": 2.291288876269393e-08,
+      "loss": 0.913,
+      "step": 15162
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.126297833179703,
+      "learning_rate": 2.2813881534186622e-08,
+      "loss": 0.6423,
+      "step": 15163
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0831100907345617,
+      "learning_rate": 2.271508818645185e-08,
+      "loss": 0.7333,
+      "step": 15164
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1042137989464944,
+      "learning_rate": 2.2616508723734555e-08,
+      "loss": 0.8802,
+      "step": 15165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1802825013233003,
+      "learning_rate": 2.2518143150271344e-08,
+      "loss": 0.7873,
+      "step": 15166
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0521066856685608,
+      "learning_rate": 2.2419991470289393e-08,
+      "loss": 0.7102,
+      "step": 15167
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.038922848856478,
+      "learning_rate": 2.232205368800533e-08,
+      "loss": 0.8653,
+      "step": 15168
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.170448837238704,
+      "learning_rate": 2.222432980762912e-08,
+      "loss": 0.7077,
+      "step": 15169
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.559744936719443,
+      "learning_rate": 2.2126819833358516e-08,
+      "loss": 0.7417,
+      "step": 15170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4411232582845375,
+      "learning_rate": 2.2029523769385718e-08,
+      "loss": 0.6754,
+      "step": 15171
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0518319894561157,
+      "learning_rate": 2.1932441619890722e-08,
+      "loss": 0.7435,
+      "step": 15172
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.132708262464266,
+      "learning_rate": 2.1835573389045185e-08,
+      "loss": 0.8621,
+      "step": 15173
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.487353182951768,
+      "learning_rate": 2.1738919081012446e-08,
+      "loss": 0.7201,
+      "step": 15174
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4371926705695794,
+      "learning_rate": 2.1642478699944736e-08,
+      "loss": 0.7886,
+      "step": 15175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3601815459283157,
+      "learning_rate": 2.1546252249988186e-08,
+      "loss": 0.7693,
+      "step": 15176
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.056924621032793,
+      "learning_rate": 2.1450239735276158e-08,
+      "loss": 0.7197,
+      "step": 15177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7712764232243914,
+      "learning_rate": 2.135444115993479e-08,
+      "loss": 0.8501,
+      "step": 15178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.435900802877139,
+      "learning_rate": 2.1258856528081906e-08,
+      "loss": 0.6749,
+      "step": 15179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.630445174314557,
+      "learning_rate": 2.116348584382366e-08,
+      "loss": 0.7814,
+      "step": 15180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0303237701848937,
+      "learning_rate": 2.1068329111259e-08,
+      "loss": 0.6997,
+      "step": 15181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.28089193099214,
+      "learning_rate": 2.097338633447743e-08,
+      "loss": 0.9833,
+      "step": 15182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0201377015122732,
+      "learning_rate": 2.087865751755791e-08,
+      "loss": 0.8145,
+      "step": 15183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.847994569842109,
+      "learning_rate": 2.0784142664571626e-08,
+      "loss": 0.7035,
+      "step": 15184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.35544565288684,
+      "learning_rate": 2.0689841779579777e-08,
+      "loss": 0.8516,
+      "step": 15185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6676194693129447,
+      "learning_rate": 2.059575486663523e-08,
+      "loss": 0.7312,
+      "step": 15186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.923513271131177,
+      "learning_rate": 2.0501881929780865e-08,
+      "loss": 0.7303,
+      "step": 15187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.074366773736389,
+      "learning_rate": 2.040822297305012e-08,
+      "loss": 1.0706,
+      "step": 15188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6048614716889955,
+      "learning_rate": 2.031477800046866e-08,
+      "loss": 0.7441,
+      "step": 15189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.904680644102253,
+      "learning_rate": 2.0221547016051614e-08,
+      "loss": 1.0182,
+      "step": 15190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2679127735690954,
+      "learning_rate": 2.012853002380466e-08,
+      "loss": 0.8589,
+      "step": 15191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8990170979224508,
+      "learning_rate": 2.003572702772627e-08,
+      "loss": 0.6612,
+      "step": 15192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.364533682939999,
+      "learning_rate": 1.9943138031803255e-08,
+      "loss": 0.6718,
+      "step": 15193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7415333625712446,
+      "learning_rate": 1.9850763040014654e-08,
+      "loss": 0.711,
+      "step": 15194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3872326850720262,
+      "learning_rate": 1.9758602056330623e-08,
+      "loss": 0.8435,
+      "step": 15195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.199197067241146,
+      "learning_rate": 1.9666655084710772e-08,
+      "loss": 0.8883,
+      "step": 15196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2638166782356846,
+      "learning_rate": 1.957492212910639e-08,
+      "loss": 0.7054,
+      "step": 15197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1896639049030804,
+      "learning_rate": 1.9483403193459873e-08,
+      "loss": 0.8856,
+      "step": 15198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.137466826635431,
+      "learning_rate": 1.939209828170363e-08,
+      "loss": 0.8252,
+      "step": 15199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.317758084718472,
+      "learning_rate": 1.9301007397761197e-08,
+      "loss": 0.6762,
+      "step": 15200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.612808858101036,
+      "learning_rate": 1.9210130545547213e-08,
+      "loss": 0.9796,
+      "step": 15201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0603596533853503,
+      "learning_rate": 1.9119467728966888e-08,
+      "loss": 0.6789,
+      "step": 15202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.902098454744301,
+      "learning_rate": 1.9029018951915446e-08,
+      "loss": 0.7214,
+      "step": 15203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9842431426850777,
+      "learning_rate": 1.8938784218281435e-08,
+      "loss": 0.7733,
+      "step": 15204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.577636188806338,
+      "learning_rate": 1.884876353194065e-08,
+      "loss": 0.8826,
+      "step": 15205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0057582987543774,
+      "learning_rate": 1.875895689676166e-08,
+      "loss": 0.6445,
+      "step": 15206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1983932705912832,
+      "learning_rate": 1.866936431660471e-08,
+      "loss": 0.7578,
+      "step": 15207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.668301416577075,
+      "learning_rate": 1.8579985795318946e-08,
+      "loss": 0.6225,
+      "step": 15208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1714485126508003,
+      "learning_rate": 1.849082133674518e-08,
+      "loss": 0.8619,
+      "step": 15209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.597344896528796,
+      "learning_rate": 1.8401870944715903e-08,
+      "loss": 0.805,
+      "step": 15210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.013001692037484,
+      "learning_rate": 1.8313134623051955e-08,
+      "loss": 0.7122,
+      "step": 15211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1788792884864736,
+      "learning_rate": 1.822461237556805e-08,
+      "loss": 1.0383,
+      "step": 15212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3761926857559055,
+      "learning_rate": 1.813630420606727e-08,
+      "loss": 0.8022,
+      "step": 15213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.092159501169834,
+      "learning_rate": 1.80482101183449e-08,
+      "loss": 0.7906,
+      "step": 15214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0685412105526735,
+      "learning_rate": 1.7960330116186807e-08,
+      "loss": 0.6367,
+      "step": 15215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3175558575193067,
+      "learning_rate": 1.7872664203368305e-08,
+      "loss": 0.6202,
+      "step": 15216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.673166661020861,
+      "learning_rate": 1.7785212383656934e-08,
+      "loss": 0.8437,
+      "step": 15217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1032078095607134,
+      "learning_rate": 1.7697974660811357e-08,
+      "loss": 0.7017,
+      "step": 15218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.1800225558237765,
+      "learning_rate": 1.761095103858024e-08,
+      "loss": 0.7716,
+      "step": 15219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0378098549508676,
+      "learning_rate": 1.7524141520702808e-08,
+      "loss": 0.683,
+      "step": 15220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.115668710883678,
+      "learning_rate": 1.743754611090942e-08,
+      "loss": 0.8657,
+      "step": 15221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3585789580267824,
+      "learning_rate": 1.7351164812921538e-08,
+      "loss": 0.7761,
+      "step": 15222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9741361659957473,
+      "learning_rate": 1.726499763045064e-08,
+      "loss": 0.8214,
+      "step": 15223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4645232714312852,
+      "learning_rate": 1.717904456720043e-08,
+      "loss": 0.8071,
+      "step": 15224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.176168259115762,
+      "learning_rate": 1.7093305626864065e-08,
+      "loss": 0.814,
+      "step": 15225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.957695360837071,
+      "learning_rate": 1.7007780813125817e-08,
+      "loss": 0.8135,
+      "step": 15226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4330987094478593,
+      "learning_rate": 1.692247012966053e-08,
+      "loss": 0.7293,
+      "step": 15227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0011070399497997,
+      "learning_rate": 1.6837373580135265e-08,
+      "loss": 0.7684,
+      "step": 15228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.041496146408462,
+      "learning_rate": 1.675249116820543e-08,
+      "loss": 0.6335,
+      "step": 15229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.927098896319453,
+      "learning_rate": 1.666782289751978e-08,
+      "loss": 0.9224,
+      "step": 15230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8540668321986247,
+      "learning_rate": 1.6583368771716513e-08,
+      "loss": 0.7722,
+      "step": 15231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0458872514226765,
+      "learning_rate": 1.6499128794423836e-08,
+      "loss": 0.7194,
+      "step": 15232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9783792125855715,
+      "learning_rate": 1.6415102969263297e-08,
+      "loss": 0.9037,
+      "step": 15233
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2246360288541895,
+      "learning_rate": 1.6331291299844233e-08,
+      "loss": 0.6936,
+      "step": 15234
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.007179677335679,
+      "learning_rate": 1.6247693789768758e-08,
+      "loss": 0.563,
+      "step": 15235
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1752513440614694,
+      "learning_rate": 1.616431044262956e-08,
+      "loss": 0.566,
+      "step": 15236
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4343323908409493,
+      "learning_rate": 1.6081141262009326e-08,
+      "loss": 0.876,
+      "step": 15237
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2912689712311052,
+      "learning_rate": 1.5998186251482416e-08,
+      "loss": 0.823,
+      "step": 15238
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2370209718622647,
+      "learning_rate": 1.5915445414613208e-08,
+      "loss": 0.7797,
+      "step": 15239
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.279467790364896,
+      "learning_rate": 1.5832918754958293e-08,
+      "loss": 0.8503,
+      "step": 15240
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2391779619910337,
+      "learning_rate": 1.5750606276062063e-08,
+      "loss": 0.8771,
+      "step": 15241
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.595482645442872,
+      "learning_rate": 1.5668507981463354e-08,
+      "loss": 0.7002,
+      "step": 15242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1103217726831685,
+      "learning_rate": 1.5586623874689345e-08,
+      "loss": 0.7823,
+      "step": 15243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7964399149627663,
+      "learning_rate": 1.550495395925944e-08,
+      "loss": 0.8492,
+      "step": 15244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7081141682883312,
+      "learning_rate": 1.54234982386825e-08,
+      "loss": 0.6755,
+      "step": 15245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1855453661785145,
+      "learning_rate": 1.5342256716459058e-08,
+      "loss": 0.7519,
+      "step": 15246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.867452396811167,
+      "learning_rate": 1.5261229396079657e-08,
+      "loss": 0.777,
+      "step": 15247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3843883546385585,
+      "learning_rate": 1.518041628102762e-08,
+      "loss": 0.8222,
+      "step": 15248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.189002342828366,
+      "learning_rate": 1.5099817374774615e-08,
+      "loss": 0.673,
+      "step": 15249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.554377737767728,
+      "learning_rate": 1.501943268078454e-08,
+      "loss": 0.8356,
+      "step": 15250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.25477211592098,
+      "learning_rate": 1.4939262202511297e-08,
+      "loss": 0.938,
+      "step": 15251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3650654214898474,
+      "learning_rate": 1.4859305943400459e-08,
+      "loss": 0.8772,
+      "step": 15252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.235226824861276,
+      "learning_rate": 1.4779563906888172e-08,
+      "loss": 0.5642,
+      "step": 15253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1504500092770575,
+      "learning_rate": 1.4700036096400028e-08,
+      "loss": 0.6047,
+      "step": 15254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.399941085469091,
+      "learning_rate": 1.4620722515354957e-08,
+      "loss": 0.8724,
+      "step": 15255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2792558874754323,
+      "learning_rate": 1.454162316715968e-08,
+      "loss": 0.8812,
+      "step": 15256
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8787715262365352,
+      "learning_rate": 1.4462738055214254e-08,
+      "loss": 0.6842,
+      "step": 15257
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3007948442228843,
+      "learning_rate": 1.4384067182908746e-08,
+      "loss": 0.7902,
+      "step": 15258
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1084532038478976,
+      "learning_rate": 1.4305610553623228e-08,
+      "loss": 0.8324,
+      "step": 15259
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1960165744090174,
+      "learning_rate": 1.4227368170728894e-08,
+      "loss": 0.7638,
+      "step": 15260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.001503152774662,
+      "learning_rate": 1.4149340037589166e-08,
+      "loss": 0.8471,
+      "step": 15261
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.897590216665811,
+      "learning_rate": 1.4071526157556358e-08,
+      "loss": 0.6661,
+      "step": 15262
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1975567060917363,
+      "learning_rate": 1.3993926533973912e-08,
+      "loss": 0.9078,
+      "step": 15263
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3226668244199806,
+      "learning_rate": 1.3916541170176934e-08,
+      "loss": 0.7851,
+      "step": 15264
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.216067299477574,
+      "learning_rate": 1.3839370069490543e-08,
+      "loss": 0.705,
+      "step": 15265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2421332534865788,
+      "learning_rate": 1.3762413235232086e-08,
+      "loss": 0.7414,
+      "step": 15266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.326509318019678,
+      "learning_rate": 1.3685670670706697e-08,
+      "loss": 0.8637,
+      "step": 15267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.198728087592819,
+      "learning_rate": 1.3609142379213957e-08,
+      "loss": 0.7489,
+      "step": 15268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.046086161836423,
+      "learning_rate": 1.3532828364041239e-08,
+      "loss": 0.7321,
+      "step": 15269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.0706204529675025,
+      "learning_rate": 1.3456728628468141e-08,
+      "loss": 0.7916,
+      "step": 15270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3230431588196068,
+      "learning_rate": 1.3380843175765379e-08,
+      "loss": 0.8817,
+      "step": 15271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4152443440117954,
+      "learning_rate": 1.3305172009193678e-08,
+      "loss": 0.8652,
+      "step": 15272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.589109365146526,
+      "learning_rate": 1.3229715132004328e-08,
+      "loss": 0.7748,
+      "step": 15273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.122305759048926,
+      "learning_rate": 1.3154472547440289e-08,
+      "loss": 0.7854,
+      "step": 15274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2637274615760594,
+      "learning_rate": 1.3079444258735086e-08,
+      "loss": 0.7079,
+      "step": 15275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2176081600746986,
+      "learning_rate": 1.3004630269112806e-08,
+      "loss": 0.7673,
+      "step": 15276
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5868764733717535,
+      "learning_rate": 1.29300305817881e-08,
+      "loss": 1.0159,
+      "step": 15277
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4039741912247585,
+      "learning_rate": 1.2855645199966738e-08,
+      "loss": 0.7007,
+      "step": 15278
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2093386520149645,
+      "learning_rate": 1.2781474126845051e-08,
+      "loss": 0.941,
+      "step": 15279
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0505845255581288,
+      "learning_rate": 1.270751736561049e-08,
+      "loss": 0.7811,
+      "step": 15280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.955579892908,
+      "learning_rate": 1.2633774919441622e-08,
+      "loss": 0.6682,
+      "step": 15281
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5574072520021476,
+      "learning_rate": 1.256024679150647e-08,
+      "loss": 0.8289,
+      "step": 15282
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.469206544211453,
+      "learning_rate": 1.2486932984965282e-08,
+      "loss": 0.5806,
+      "step": 15283
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6706415519354088,
+      "learning_rate": 1.241383350296832e-08,
+      "loss": 0.6786,
+      "step": 15284
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3003025189134203,
+      "learning_rate": 1.2340948348656955e-08,
+      "loss": 0.6511,
+      "step": 15285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0974707399358459,
+      "learning_rate": 1.2268277525163685e-08,
+      "loss": 0.6568,
+      "step": 15286
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4648444270074945,
+      "learning_rate": 1.21958210356099e-08,
+      "loss": 0.8094,
+      "step": 15287
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8478119153662278,
+      "learning_rate": 1.2123578883110887e-08,
+      "loss": 0.8246,
+      "step": 15288
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0761126774752285,
+      "learning_rate": 1.2051551070769719e-08,
+      "loss": 0.8869,
+      "step": 15289
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1535267796464534,
+      "learning_rate": 1.1979737601682252e-08,
+      "loss": 0.7483,
+      "step": 15290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8879034308789622,
+      "learning_rate": 1.1908138478934905e-08,
+      "loss": 0.87,
+      "step": 15291
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.457951963975998,
+      "learning_rate": 1.1836753705602998e-08,
+      "loss": 0.8881,
+      "step": 15292
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.385857068322229,
+      "learning_rate": 1.1765583284755743e-08,
+      "loss": 0.7132,
+      "step": 15293
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3158341473586246,
+      "learning_rate": 1.1694627219450694e-08,
+      "loss": 0.7168,
+      "step": 15294
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8857314126753326,
+      "learning_rate": 1.1623885512737076e-08,
+      "loss": 0.7371,
+      "step": 15295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4086639483691554,
+      "learning_rate": 1.1553358167654683e-08,
+      "loss": 0.6752,
+      "step": 15296
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.008329726860844,
+      "learning_rate": 1.1483045187233866e-08,
+      "loss": 0.7164,
+      "step": 15297
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8821407314334289,
+      "learning_rate": 1.1412946574497208e-08,
+      "loss": 0.653,
+      "step": 15298
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0243136147955902,
+      "learning_rate": 1.134306233245619e-08,
+      "loss": 0.9102,
+      "step": 15299
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.092205737111152,
+      "learning_rate": 1.1273392464114518e-08,
+      "loss": 0.6683,
+      "step": 15300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8345173300351796,
+      "learning_rate": 1.1203936972465356e-08,
+      "loss": 0.8945,
+      "step": 15301
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.390601529433297,
+      "learning_rate": 1.1134695860493539e-08,
+      "loss": 0.8157,
+      "step": 15302
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1549368994366023,
+      "learning_rate": 1.1065669131175016e-08,
+      "loss": 0.7044,
+      "step": 15303
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.303940244319169,
+      "learning_rate": 1.0996856787475197e-08,
+      "loss": 0.7164,
+      "step": 15304
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0763626818726673,
+      "learning_rate": 1.0928258832351712e-08,
+      "loss": 0.7566,
+      "step": 15305
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9116681588916589,
+      "learning_rate": 1.0859875268752206e-08,
+      "loss": 0.7565,
+      "step": 15306
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.200135311403055,
+      "learning_rate": 1.079170609961544e-08,
+      "loss": 0.7761,
+      "step": 15307
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3211977471622933,
+      "learning_rate": 1.072375132787018e-08,
+      "loss": 0.6182,
+      "step": 15308
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9730291958828654,
+      "learning_rate": 1.0656010956437979e-08,
+      "loss": 0.8649,
+      "step": 15309
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8379262520186725,
+      "learning_rate": 1.0588484988228175e-08,
+      "loss": 0.8042,
+      "step": 15310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.803657979273643,
+      "learning_rate": 1.0521173426142894e-08,
+      "loss": 0.7107,
+      "step": 15311
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.216087388610408,
+      "learning_rate": 1.0454076273075375e-08,
+      "loss": 0.7579,
+      "step": 15312
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0917298612261583,
+      "learning_rate": 1.0387193531908313e-08,
+      "loss": 0.8943,
+      "step": 15313
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.200992191925836,
+      "learning_rate": 1.0320525205516629e-08,
+      "loss": 0.7863,
+      "step": 15314
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2324262713601457,
+      "learning_rate": 1.025407129676359e-08,
+      "loss": 0.8508,
+      "step": 15315
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.07216940146783,
+      "learning_rate": 1.018783180850691e-08,
+      "loss": 0.6269,
+      "step": 15316
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3390276369044845,
+      "learning_rate": 1.0121806743591534e-08,
+      "loss": 0.7186,
+      "step": 15317
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.93847090872733,
+      "learning_rate": 1.0055996104854637e-08,
+      "loss": 0.7287,
+      "step": 15318
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4548812051271742,
+      "learning_rate": 9.990399895125624e-09,
+      "loss": 0.7325,
+      "step": 15319
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2722202604813315,
+      "learning_rate": 9.925018117221686e-09,
+      "loss": 0.7372,
+      "step": 15320
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.632495937237905,
+      "learning_rate": 9.859850773953905e-09,
+      "loss": 0.7594,
+      "step": 15321
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9945712872126427,
+      "learning_rate": 9.794897868121711e-09,
+      "loss": 0.852,
+      "step": 15322
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2027518260092642,
+      "learning_rate": 9.73015940251676e-09,
+      "loss": 0.8946,
+      "step": 15323
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0938171754188377,
+      "learning_rate": 9.665635379920157e-09,
+      "loss": 0.6569,
+      "step": 15324
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.206913325983744,
+      "learning_rate": 9.6013258031058e-09,
+      "loss": 0.6803,
+      "step": 15325
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0965143805635615,
+      "learning_rate": 9.537230674836473e-09,
+      "loss": 0.8028,
+      "step": 15326
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.088581816372171,
+      "learning_rate": 9.473349997867198e-09,
+      "loss": 0.7992,
+      "step": 15327
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3401473255714325,
+      "learning_rate": 9.409683774941891e-09,
+      "loss": 0.7356,
+      "step": 15328
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1518816156581835,
+      "learning_rate": 9.346232008797252e-09,
+      "loss": 0.7723,
+      "step": 15329
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.188064665183558,
+      "learning_rate": 9.282994702159986e-09,
+      "loss": 0.7567,
+      "step": 15330
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9518686200652893,
+      "learning_rate": 9.219971857747368e-09,
+      "loss": 0.7618,
+      "step": 15331
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5078597050497002,
+      "learning_rate": 9.157163478267783e-09,
+      "loss": 0.839,
+      "step": 15332
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.274668104709732,
+      "learning_rate": 9.094569566419632e-09,
+      "loss": 0.6243,
+      "step": 15333
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1524107372871675,
+      "learning_rate": 9.032190124893536e-09,
+      "loss": 0.7326,
+      "step": 15334
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0173119769222696,
+      "learning_rate": 8.970025156369578e-09,
+      "loss": 0.5842,
+      "step": 15335
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2326320345190687,
+      "learning_rate": 8.908074663519505e-09,
+      "loss": 0.6869,
+      "step": 15336
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.432438820401458,
+      "learning_rate": 8.846338649005082e-09,
+      "loss": 0.7944,
+      "step": 15337
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.988691207300327,
+      "learning_rate": 8.784817115479183e-09,
+      "loss": 0.6627,
+      "step": 15338
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7639973533054554,
+      "learning_rate": 8.723510065585806e-09,
+      "loss": 0.7703,
+      "step": 15339
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.044199676779124,
+      "learning_rate": 8.662417501959507e-09,
+      "loss": 0.7922,
+      "step": 15340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.025269586421196,
+      "learning_rate": 8.601539427225414e-09,
+      "loss": 0.6521,
+      "step": 15341
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.760996597836465,
+      "learning_rate": 8.540875843999208e-09,
+      "loss": 0.7348,
+      "step": 15342
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2541278507681564,
+      "learning_rate": 8.480426754888805e-09,
+      "loss": 0.7588,
+      "step": 15343
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7742758924281565,
+      "learning_rate": 8.42019216249046e-09,
+      "loss": 0.8666,
+      "step": 15344
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.478686621986209,
+      "learning_rate": 8.360172069393768e-09,
+      "loss": 0.7801,
+      "step": 15345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2135904707001908,
+      "learning_rate": 8.300366478177779e-09,
+      "loss": 0.7755,
+      "step": 15346
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0906826706956596,
+      "learning_rate": 8.240775391411548e-09,
+      "loss": 0.717,
+      "step": 15347
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.326425752786376,
+      "learning_rate": 8.18139881165636e-09,
+      "loss": 0.8255,
+      "step": 15348
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2200056173290128,
+      "learning_rate": 8.122236741464618e-09,
+      "loss": 0.6287,
+      "step": 15349
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1278239179227625,
+      "learning_rate": 8.063289183377065e-09,
+      "loss": 0.7035,
+      "step": 15350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2652327045532754,
+      "learning_rate": 8.004556139927788e-09,
+      "loss": 0.7139,
+      "step": 15351
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1735606334322153,
+      "learning_rate": 7.946037613640878e-09,
+      "loss": 0.9113,
+      "step": 15352
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7074194236245406,
+      "learning_rate": 7.887733607030435e-09,
+      "loss": 0.8029,
+      "step": 15353
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.879668382437602,
+      "learning_rate": 7.82964412260223e-09,
+      "loss": 0.7438,
+      "step": 15354
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1187706882535124,
+      "learning_rate": 7.77176916285205e-09,
+      "loss": 0.6723,
+      "step": 15355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.549116626649985,
+      "learning_rate": 7.714108730267345e-09,
+      "loss": 0.8146,
+      "step": 15356
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3216435304326137,
+      "learning_rate": 7.656662827326134e-09,
+      "loss": 0.8098,
+      "step": 15357
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.901568449201383,
+      "learning_rate": 7.599431456495888e-09,
+      "loss": 0.728,
+      "step": 15358
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5415205698036916,
+      "learning_rate": 7.542414620237414e-09,
+      "loss": 0.8305,
+      "step": 15359
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8620585210893927,
+      "learning_rate": 7.485612321000424e-09,
+      "loss": 0.834,
+      "step": 15360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2529900132662313,
+      "learning_rate": 7.4290245612246294e-09,
+      "loss": 0.804,
+      "step": 15361
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.268722937505718,
+      "learning_rate": 7.372651343343639e-09,
+      "loss": 0.9011,
+      "step": 15362
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.15327311344073,
+      "learning_rate": 7.316492669778297e-09,
+      "loss": 0.7192,
+      "step": 15363
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.987666376478295,
+      "learning_rate": 7.260548542943335e-09,
+      "loss": 0.7675,
+      "step": 15364
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0104464589708555,
+      "learning_rate": 7.2048189652412784e-09,
+      "loss": 0.8251,
+      "step": 15365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2657492422277783,
+      "learning_rate": 7.149303939067986e-09,
+      "loss": 0.8201,
+      "step": 15366
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4410897059355996,
+      "learning_rate": 7.094003466808774e-09,
+      "loss": 0.8901,
+      "step": 15367
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.281139978848566,
+      "learning_rate": 7.038917550840074e-09,
+      "loss": 0.8163,
+      "step": 15368
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9900155523757946,
+      "learning_rate": 6.984046193528881e-09,
+      "loss": 0.7438,
+      "step": 15369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6128056191563434,
+      "learning_rate": 6.9293893972338654e-09,
+      "loss": 0.8173,
+      "step": 15370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5121698509708006,
+      "learning_rate": 6.874947164302592e-09,
+      "loss": 0.8942,
+      "step": 15371
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.108130342806249,
+      "learning_rate": 6.820719497074857e-09,
+      "loss": 0.7914,
+      "step": 15372
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9915855472708828,
+      "learning_rate": 6.7667063978815724e-09,
+      "loss": 0.8132,
+      "step": 15373
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.165020805913904,
+      "learning_rate": 6.712907869043661e-09,
+      "loss": 0.7046,
+      "step": 15374
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.15895887318214,
+      "learning_rate": 6.659323912872606e-09,
+      "loss": 0.8664,
+      "step": 15375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3933174482653152,
+      "learning_rate": 6.60595453167101e-09,
+      "loss": 0.8749,
+      "step": 15376
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2534188025165403,
+      "learning_rate": 6.552799727732595e-09,
+      "loss": 0.9267,
+      "step": 15377
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0198246104282234,
+      "learning_rate": 6.4998595033410886e-09,
+      "loss": 0.6535,
+      "step": 15378
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.011681789193445,
+      "learning_rate": 6.447133860771893e-09,
+      "loss": 0.8195,
+      "step": 15379
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.890334954443396,
+      "learning_rate": 6.394622802290973e-09,
+      "loss": 0.5172,
+      "step": 15380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1651903310279434,
+      "learning_rate": 6.342326330153747e-09,
+      "loss": 0.6198,
+      "step": 15381
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7538320377964856,
+      "learning_rate": 6.290244446608973e-09,
+      "loss": 0.6874,
+      "step": 15382
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0856510279704465,
+      "learning_rate": 6.238377153893749e-09,
+      "loss": 0.5332,
+      "step": 15383
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.467284877093055,
+      "learning_rate": 6.186724454236847e-09,
+      "loss": 0.7426,
+      "step": 15384
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.084580147435911,
+      "learning_rate": 6.13528634985816e-09,
+      "loss": 0.9855,
+      "step": 15385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.593324142267427,
+      "learning_rate": 6.084062842968696e-09,
+      "loss": 0.8394,
+      "step": 15386
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1481447633771493,
+      "learning_rate": 6.0330539357689175e-09,
+      "loss": 0.957,
+      "step": 15387
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6367329616394706,
+      "learning_rate": 5.98225963045096e-09,
+      "loss": 0.8802,
+      "step": 15388
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2102844449468533,
+      "learning_rate": 5.9316799291969654e-09,
+      "loss": 0.6135,
+      "step": 15389
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2735787058346988,
+      "learning_rate": 5.881314834181862e-09,
+      "loss": 0.7332,
+      "step": 15390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0489687565907184,
+      "learning_rate": 5.831164347568918e-09,
+      "loss": 0.7991,
+      "step": 15391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.765818667583224,
+      "learning_rate": 5.781228471513633e-09,
+      "loss": 0.6945,
+      "step": 15392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0933064970147544,
+      "learning_rate": 5.731507208160958e-09,
+      "loss": 0.8392,
+      "step": 15393
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.702336669918469,
+      "learning_rate": 5.682000559649181e-09,
+      "loss": 0.8536,
+      "step": 15394
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1953205805520102,
+      "learning_rate": 5.63270852810438e-09,
+      "loss": 0.6091,
+      "step": 15395
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0152040987414215,
+      "learning_rate": 5.583631115644861e-09,
+      "loss": 0.7632,
+      "step": 15396
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2496870682186136,
+      "learning_rate": 5.534768324380046e-09,
+      "loss": 0.8147,
+      "step": 15397
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3598751773111952,
+      "learning_rate": 5.486120156409924e-09,
+      "loss": 0.7699,
+      "step": 15398
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5072472851329812,
+      "learning_rate": 5.437686613823934e-09,
+      "loss": 0.9634,
+      "step": 15399
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9016027240207125,
+      "learning_rate": 5.389467698704298e-09,
+      "loss": 0.6725,
+      "step": 15400
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9202811151369104,
+      "learning_rate": 5.341463413123249e-09,
+      "loss": 0.8067,
+      "step": 15401
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5148989404590267,
+      "learning_rate": 5.29367375914247e-09,
+      "loss": 0.7355,
+      "step": 15402
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.499520473206462,
+      "learning_rate": 5.246098738816985e-09,
+      "loss": 0.8535,
+      "step": 15403
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.116548384178757,
+      "learning_rate": 5.198738354190158e-09,
+      "loss": 0.8163,
+      "step": 15404
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0148329817623396,
+      "learning_rate": 5.151592607298139e-09,
+      "loss": 0.5693,
+      "step": 15405
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1815353654929615,
+      "learning_rate": 5.104661500165975e-09,
+      "loss": 0.7228,
+      "step": 15406
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1208795110636904,
+      "learning_rate": 5.057945034810385e-09,
+      "loss": 0.5762,
+      "step": 15407
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1867126134518473,
+      "learning_rate": 5.0114432132397636e-09,
+      "loss": 0.695,
+      "step": 15408
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2204916225110076,
+      "learning_rate": 4.9651560374514015e-09,
+      "loss": 0.6853,
+      "step": 15409
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.307841060597286,
+      "learning_rate": 4.919083509434819e-09,
+      "loss": 0.9884,
+      "step": 15410
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.464101104375962,
+      "learning_rate": 4.873225631170098e-09,
+      "loss": 0.7914,
+      "step": 15411
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.256422098174574,
+      "learning_rate": 4.827582404627884e-09,
+      "loss": 0.8219,
+      "step": 15412
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9778041527058505,
+      "learning_rate": 4.782153831768832e-09,
+      "loss": 0.7786,
+      "step": 15413
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8561164829623717,
+      "learning_rate": 4.736939914545824e-09,
+      "loss": 0.8071,
+      "step": 15414
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.601467911361137,
+      "learning_rate": 4.691940654901195e-09,
+      "loss": 0.9823,
+      "step": 15415
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9351614739722094,
+      "learning_rate": 4.647156054769508e-09,
+      "loss": 0.7705,
+      "step": 15416
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.062095491638619,
+      "learning_rate": 4.602586116074781e-09,
+      "loss": 0.6476,
+      "step": 15417
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.280126660818931,
+      "learning_rate": 4.558230840731592e-09,
+      "loss": 0.7101,
+      "step": 15418
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3234321395506012,
+      "learning_rate": 4.514090230647305e-09,
+      "loss": 0.6558,
+      "step": 15419
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3154443290444577,
+      "learning_rate": 4.470164287717627e-09,
+      "loss": 0.7359,
+      "step": 15420
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.038091302523622,
+      "learning_rate": 4.4264530138310445e-09,
+      "loss": 0.6003,
+      "step": 15421
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2333731538349597,
+      "learning_rate": 4.382956410865502e-09,
+      "loss": 0.6292,
+      "step": 15422
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.705772079002665,
+      "learning_rate": 4.339674480690059e-09,
+      "loss": 0.8855,
+      "step": 15423
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.735496968508907,
+      "learning_rate": 4.296607225164895e-09,
+      "loss": 0.8376,
+      "step": 15424
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.061885703936921,
+      "learning_rate": 4.253754646140196e-09,
+      "loss": 0.9741,
+      "step": 15425
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.16834602536583,
+      "learning_rate": 4.211116745458377e-09,
+      "loss": 0.6696,
+      "step": 15426
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.349033940973905,
+      "learning_rate": 4.168693524950196e-09,
+      "loss": 0.9143,
+      "step": 15427
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.015151744428591,
+      "learning_rate": 4.1264849864403044e-09,
+      "loss": 0.8486,
+      "step": 15428
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.251244092132396,
+      "learning_rate": 4.084491131741697e-09,
+      "loss": 0.8518,
+      "step": 15429
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.828814121106205,
+      "learning_rate": 4.042711962658486e-09,
+      "loss": 0.7054,
+      "step": 15430
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.980703342113554,
+      "learning_rate": 4.001147480987011e-09,
+      "loss": 0.848,
+      "step": 15431
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0067934699642587,
+      "learning_rate": 3.959797688513067e-09,
+      "loss": 0.8069,
+      "step": 15432
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2143068220096898,
+      "learning_rate": 3.9186625870130115e-09,
+      "loss": 0.6497,
+      "step": 15433
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.178221190321435,
+      "learning_rate": 3.877742178254873e-09,
+      "loss": 0.7597,
+      "step": 15434
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.107681129673171,
+      "learning_rate": 3.837036463997246e-09,
+      "loss": 0.9375,
+      "step": 15435
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9663432831941081,
+      "learning_rate": 3.796545445989286e-09,
+      "loss": 0.5693,
+      "step": 15436
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.32378040287855,
+      "learning_rate": 3.756269125970713e-09,
+      "loss": 0.5458,
+      "step": 15437
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1507063325134537,
+      "learning_rate": 3.7162075056729196e-09,
+      "loss": 0.7908,
+      "step": 15438
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6548308247985473,
+      "learning_rate": 3.6763605868167516e-09,
+      "loss": 0.6819,
+      "step": 15439
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.407553379420524,
+      "learning_rate": 3.636728371114173e-09,
+      "loss": 0.7681,
+      "step": 15440
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7239476419051756,
+      "learning_rate": 3.5973108602699313e-09,
+      "loss": 0.8257,
+      "step": 15441
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.038518926136439,
+      "learning_rate": 3.558108055976006e-09,
+      "loss": 0.7834,
+      "step": 15442
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4561033592523107,
+      "learning_rate": 3.5191199599182712e-09,
+      "loss": 0.7201,
+      "step": 15443
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.293078958811557,
+      "learning_rate": 3.4803465737714983e-09,
+      "loss": 0.5401,
+      "step": 15444
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.399304039844068,
+      "learning_rate": 3.4417878992021315e-09,
+      "loss": 0.8231,
+      "step": 15445
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0698171594240025,
+      "learning_rate": 3.403443937867179e-09,
+      "loss": 0.7837,
+      "step": 15446
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0736841856849255,
+      "learning_rate": 3.365314691414212e-09,
+      "loss": 0.7053,
+      "step": 15447
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0220541347739482,
+      "learning_rate": 3.3274001614819195e-09,
+      "loss": 0.7179,
+      "step": 15448
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.12975159600881,
+      "learning_rate": 3.289700349698999e-09,
+      "loss": 0.7561,
+      "step": 15449
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.005372770369834,
+      "learning_rate": 3.252215257686375e-09,
+      "loss": 0.704,
+      "step": 15450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.122314455772646,
+      "learning_rate": 3.2149448870544273e-09,
+      "loss": 0.6835,
+      "step": 15451
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9918235106349602,
+      "learning_rate": 3.1778892394046525e-09,
+      "loss": 0.8671,
+      "step": 15452
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2513520586015678,
+      "learning_rate": 3.1410483163291093e-09,
+      "loss": 0.6991,
+      "step": 15453
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3023481997844404,
+      "learning_rate": 3.104422119411532e-09,
+      "loss": 0.8909,
+      "step": 15454
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.224555309997502,
+      "learning_rate": 3.0680106502256613e-09,
+      "loss": 0.7075,
+      "step": 15455
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.549570970321342,
+      "learning_rate": 3.0318139103363564e-09,
+      "loss": 0.8975,
+      "step": 15456
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.45575629436806,
+      "learning_rate": 2.9958319012984847e-09,
+      "loss": 0.6682,
+      "step": 15457
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4157383620390216,
+      "learning_rate": 2.9600646246591425e-09,
+      "loss": 0.9707,
+      "step": 15458
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.014543137795539,
+      "learning_rate": 2.9245120819543226e-09,
+      "loss": 0.7065,
+      "step": 15459
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3178016000955208,
+      "learning_rate": 2.8891742747128025e-09,
+      "loss": 0.6494,
+      "step": 15460
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2625489273790307,
+      "learning_rate": 2.854051204452257e-09,
+      "loss": 0.7509,
+      "step": 15461
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.642179445536426,
+      "learning_rate": 2.8191428726820347e-09,
+      "loss": 0.7569,
+      "step": 15462
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0700996278239874,
+      "learning_rate": 2.7844492809031567e-09,
+      "loss": 0.8776,
+      "step": 15463
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.081654121835923,
+      "learning_rate": 2.749970430605542e-09,
+      "loss": 0.9366,
+      "step": 15464
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0342928467995005,
+      "learning_rate": 2.715706323271339e-09,
+      "loss": 0.778,
+      "step": 15465
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9033548227991575,
+      "learning_rate": 2.681656960372703e-09,
+      "loss": 0.7397,
+      "step": 15466
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2426781568940495,
+      "learning_rate": 2.6478223433734627e-09,
+      "loss": 0.8485,
+      "step": 15467
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.019995730411732,
+      "learning_rate": 2.614202473726346e-09,
+      "loss": 0.6628,
+      "step": 15468
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.118126083816835,
+      "learning_rate": 2.5807973528768626e-09,
+      "loss": 0.7276,
+      "step": 15469
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2970618591275414,
+      "learning_rate": 2.547606982260531e-09,
+      "loss": 0.6217,
+      "step": 15470
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.381269449416328,
+      "learning_rate": 2.5146313633028775e-09,
+      "loss": 0.8572,
+      "step": 15471
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.844677488567724,
+      "learning_rate": 2.481870497422212e-09,
+      "loss": 0.8652,
+      "step": 15472
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.337598785895897,
+      "learning_rate": 2.449324386025187e-09,
+      "loss": 0.691,
+      "step": 15473
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.930379244189388,
+      "learning_rate": 2.416993030511239e-09,
+      "loss": 0.7086,
+      "step": 15474
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7754960926051386,
+      "learning_rate": 2.3848764322687014e-09,
+      "loss": 0.967,
+      "step": 15475
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7035195768007347,
+      "learning_rate": 2.3529745926786916e-09,
+      "loss": 0.7348,
+      "step": 15476
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.057995716356013,
+      "learning_rate": 2.3212875131117805e-09,
+      "loss": 0.8643,
+      "step": 15477
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.41217961170916,
+      "learning_rate": 2.2898151949296566e-09,
+      "loss": 0.842,
+      "step": 15478
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0561997714201796,
+      "learning_rate": 2.258557639484571e-09,
+      "loss": 0.8241,
+      "step": 15479
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.498831335087788,
+      "learning_rate": 2.227514848119894e-09,
+      "loss": 0.7024,
+      "step": 15480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.106814118758657,
+      "learning_rate": 2.196686822169003e-09,
+      "loss": 0.7732,
+      "step": 15481
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.324988076782785,
+      "learning_rate": 2.1660735629580598e-09,
+      "loss": 0.9204,
+      "step": 15482
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1120749738111992,
+      "learning_rate": 2.1356750718010133e-09,
+      "loss": 0.6984,
+      "step": 15483
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5738854704086678,
+      "learning_rate": 2.1054913500051512e-09,
+      "loss": 0.8669,
+      "step": 15484
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.158651768202401,
+      "learning_rate": 2.0755223988672136e-09,
+      "loss": 0.6827,
+      "step": 15485
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.748899820846655,
+      "learning_rate": 2.0457682196750595e-09,
+      "loss": 0.9234,
+      "step": 15486
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.105724732953678,
+      "learning_rate": 2.0162288137071105e-09,
+      "loss": 0.8169,
+      "step": 15487
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.338203415659204,
+      "learning_rate": 1.9869041822329073e-09,
+      "loss": 0.8849,
+      "step": 15488
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5810597678462663,
+      "learning_rate": 1.957794326513107e-09,
+      "loss": 0.8209,
+      "step": 15489
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6768023440159108,
+      "learning_rate": 1.928899247797822e-09,
+      "loss": 0.8841,
+      "step": 15490
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2219964963106844,
+      "learning_rate": 1.9002189473288356e-09,
+      "loss": 0.8035,
+      "step": 15491
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.294842870095805,
+      "learning_rate": 1.8717534263390513e-09,
+      "loss": 0.7656,
+      "step": 15492
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3348309417336475,
+      "learning_rate": 1.8435026860513794e-09,
+      "loss": 0.7824,
+      "step": 15493
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2423103969344855,
+      "learning_rate": 1.8154667276798488e-09,
+      "loss": 0.8579,
+      "step": 15494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.502599867699983,
+      "learning_rate": 1.7876455524290514e-09,
+      "loss": 1.0131,
+      "step": 15495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3612284083747714,
+      "learning_rate": 1.7600391614952529e-09,
+      "loss": 0.8274,
+      "step": 15496
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2845700313651953,
+      "learning_rate": 1.7326475560636157e-09,
+      "loss": 0.7267,
+      "step": 15497
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0043461764931467,
+      "learning_rate": 1.7054707373126423e-09,
+      "loss": 0.8592,
+      "step": 15498
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7584727985818422,
+      "learning_rate": 1.6785087064086213e-09,
+      "loss": 0.8578,
+      "step": 15499
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9676081933926617,
+      "learning_rate": 1.6517614645111812e-09,
+      "loss": 0.7361,
+      "step": 15500
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1699789874507664,
+      "learning_rate": 1.6252290127699577e-09,
+      "loss": 0.7173,
+      "step": 15501
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.126449900828759,
+      "learning_rate": 1.5989113523240396e-09,
+      "loss": 0.811,
+      "step": 15502
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9826422002789221,
+      "learning_rate": 1.5728084843052992e-09,
+      "loss": 1.051,
+      "step": 15503
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1283478739645494,
+      "learning_rate": 1.546920409834507e-09,
+      "loss": 0.9474,
+      "step": 15504
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.385916321824235,
+      "learning_rate": 1.5212471300252163e-09,
+      "loss": 0.6873,
+      "step": 15505
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0273545775318786,
+      "learning_rate": 1.4957886459798787e-09,
+      "loss": 0.8335,
+      "step": 15506
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2085442742554795,
+      "learning_rate": 1.4705449587931742e-09,
+      "loss": 0.7661,
+      "step": 15507
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3180997237116725,
+      "learning_rate": 1.4455160695492354e-09,
+      "loss": 0.6397,
+      "step": 15508
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.11062495180028,
+      "learning_rate": 1.4207019793238686e-09,
+      "loss": 0.9713,
+      "step": 15509
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1630352068560645,
+      "learning_rate": 1.3961026891828878e-09,
+      "loss": 0.9328,
+      "step": 15510
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.8049264863187378,
+      "learning_rate": 1.3717182001843355e-09,
+      "loss": 0.7519,
+      "step": 15511
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5322588297025628,
+      "learning_rate": 1.347548513375707e-09,
+      "loss": 0.7301,
+      "step": 15512
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.243562350131891,
+      "learning_rate": 1.3235936297956165e-09,
+      "loss": 0.6804,
+      "step": 15513
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.505612385290232,
+      "learning_rate": 1.299853550472685e-09,
+      "loss": 0.863,
+      "step": 15514
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9386108981848764,
+      "learning_rate": 1.276328276428318e-09,
+      "loss": 0.7448,
+      "step": 15515
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.39456050003107,
+      "learning_rate": 1.2530178086728183e-09,
+      "loss": 0.7944,
+      "step": 15516
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0737423201667604,
+      "learning_rate": 1.2299221482076074e-09,
+      "loss": 0.708,
+      "step": 15517
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4921274425714297,
+      "learning_rate": 1.2070412960257793e-09,
+      "loss": 0.8425,
+      "step": 15518
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3102713853433445,
+      "learning_rate": 1.1843752531104368e-09,
+      "loss": 0.7653,
+      "step": 15519
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 36.999456911027174,
+      "learning_rate": 1.1619240204352455e-09,
+      "loss": 0.9731,
+      "step": 15520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.136725789809754,
+      "learning_rate": 1.1396875989655443e-09,
+      "loss": 0.7605,
+      "step": 15521
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3006332927964919,
+      "learning_rate": 1.1176659896561247e-09,
+      "loss": 0.6102,
+      "step": 15522
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7824023890269776,
+      "learning_rate": 1.0958591934534525e-09,
+      "loss": 0.6298,
+      "step": 15523
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9615438180255829,
+      "learning_rate": 1.0742672112951103e-09,
+      "loss": 0.6997,
+      "step": 15524
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3254535006388632,
+      "learning_rate": 1.052890044109245e-09,
+      "loss": 0.8255,
+      "step": 15525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8029566457557324,
+      "learning_rate": 1.0317276928134557e-09,
+      "loss": 0.7271,
+      "step": 15526
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.475255134721409,
+      "learning_rate": 1.0107801583175702e-09,
+      "loss": 0.7861,
+      "step": 15527
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.025281496430098,
+      "learning_rate": 9.900474415219795e-10,
+      "loss": 0.7411,
+      "step": 15528
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1407345159609346,
+      "learning_rate": 9.695295433170826e-10,
+      "loss": 0.6721,
+      "step": 15529
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9400908955215566,
+      "learning_rate": 9.492264645855065e-10,
+      "loss": 0.9594,
+      "step": 15530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0189760979776707,
+      "learning_rate": 9.291382061987763e-10,
+      "loss": 0.6067,
+      "step": 15531
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2597269042402965,
+      "learning_rate": 9.092647690206457e-10,
+      "loss": 0.8464,
+      "step": 15532
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8004062702655503,
+      "learning_rate": 8.896061539048762e-10,
+      "loss": 0.6593,
+      "step": 15533
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.251521316364478,
+      "learning_rate": 8.701623616963472e-10,
+      "loss": 0.7204,
+      "step": 15534
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.23119046643004,
+      "learning_rate": 8.509333932310571e-10,
+      "loss": 0.622,
+      "step": 15535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4238388993302853,
+      "learning_rate": 8.319192493344563e-10,
+      "loss": 0.7818,
+      "step": 15536
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.193323177384786,
+      "learning_rate": 8.131199308236692e-10,
+      "loss": 1.0688,
+      "step": 15537
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3951893307416685,
+      "learning_rate": 7.945354385074933e-10,
+      "loss": 0.6444,
+      "step": 15538
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8364805212943214,
+      "learning_rate": 7.761657731836236e-10,
+      "loss": 0.7569,
+      "step": 15539
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.332174760501655,
+      "learning_rate": 7.580109356419841e-10,
+      "loss": 0.7358,
+      "step": 15540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1106971152677465,
+      "learning_rate": 7.400709266625062e-10,
+      "loss": 0.7998,
+      "step": 15541
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.127467394917939,
+      "learning_rate": 7.223457470156847e-10,
+      "loss": 0.7626,
+      "step": 15542
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8753013789944846,
+      "learning_rate": 7.04835397464243e-10,
+      "loss": 0.6063,
+      "step": 15543
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0193015159308287,
+      "learning_rate": 6.87539878759802e-10,
+      "loss": 0.9154,
+      "step": 15544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3353193296082235,
+      "learning_rate": 6.704591916462111e-10,
+      "loss": 0.8445,
+      "step": 15545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.646236877008405,
+      "learning_rate": 6.535933368567726e-10,
+      "loss": 0.6475,
+      "step": 15546
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.067906601155733,
+      "learning_rate": 6.369423151164622e-10,
+      "loss": 0.6794,
+      "step": 15547
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.91654036504449,
+      "learning_rate": 6.205061271408186e-10,
+      "loss": 0.7852,
+      "step": 15548
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1309820228274805,
+      "learning_rate": 6.042847736364987e-10,
+      "loss": 0.5881,
+      "step": 15549
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8712439381981714,
+      "learning_rate": 5.882782553001676e-10,
+      "loss": 0.6163,
+      "step": 15550
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.989693538586779,
+      "learning_rate": 5.724865728201634e-10,
+      "loss": 0.7632,
+      "step": 15551
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9533558626436909,
+      "learning_rate": 5.569097268742773e-10,
+      "loss": 0.8945,
+      "step": 15552
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9986918309610187,
+      "learning_rate": 5.415477181325291e-10,
+      "loss": 0.8042,
+      "step": 15553
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2644255652654723,
+      "learning_rate": 5.264005472549461e-10,
+      "loss": 0.7125,
+      "step": 15554
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8802945325562734,
+      "learning_rate": 5.114682148921191e-10,
+      "loss": 0.8753,
+      "step": 15555
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6652035726788736,
+      "learning_rate": 4.967507216863121e-10,
+      "loss": 0.7337,
+      "step": 15556
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9582639452786959,
+      "learning_rate": 4.822480682692421e-10,
+      "loss": 0.7001,
+      "step": 15557
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2019603662476697,
+      "learning_rate": 4.679602552648543e-10,
+      "loss": 0.8058,
+      "step": 15558
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6777746874442934,
+      "learning_rate": 4.538872832865471e-10,
+      "loss": 0.8909,
+      "step": 15559
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.214228218300208,
+      "learning_rate": 4.4002915293939186e-10,
+      "loss": 0.8118,
+      "step": 15560
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.076218241518644,
+      "learning_rate": 4.2638586481846823e-10,
+      "loss": 0.6039,
+      "step": 15561
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0394462670215967,
+      "learning_rate": 4.1295741951108417e-10,
+      "loss": 0.5938,
+      "step": 15562
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4962996731934997,
+      "learning_rate": 3.9974381759289027e-10,
+      "loss": 0.8134,
+      "step": 15563
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.124287120712738,
+      "learning_rate": 3.867450596328759e-10,
+      "loss": 0.7384,
+      "step": 15564
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3492924228358225,
+      "learning_rate": 3.739611461889281e-10,
+      "loss": 0.7572,
+      "step": 15565
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9410032431316178,
+      "learning_rate": 3.6139207781060723e-10,
+      "loss": 0.8021,
+      "step": 15566
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3035484851511256,
+      "learning_rate": 3.490378550380369e-10,
+      "loss": 0.593,
+      "step": 15567
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.548338550045549,
+      "learning_rate": 3.368984784024587e-10,
+      "loss": 0.582,
+      "step": 15568
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.36991270518799,
+      "learning_rate": 3.2497394842512244e-10,
+      "loss": 0.8937,
+      "step": 15569
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.900147475481181,
+      "learning_rate": 3.1326426561839597e-10,
+      "loss": 0.7007,
+      "step": 15570
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9040523456291356,
+      "learning_rate": 3.017694304852103e-10,
+      "loss": 0.8269,
+      "step": 15571
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0858603499654487,
+      "learning_rate": 2.9048944352072505e-10,
+      "loss": 0.5958,
+      "step": 15572
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.718538752007495,
+      "learning_rate": 2.794243052084422e-10,
+      "loss": 0.8466,
+      "step": 15573
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1804963002854723,
+      "learning_rate": 2.685740160240924e-10,
+      "loss": 0.7751,
+      "step": 15574
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.12829107153095,
+      "learning_rate": 2.5793857643396924e-10,
+      "loss": 0.7291,
+      "step": 15575
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5101139091599758,
+      "learning_rate": 2.475179868960398e-10,
+      "loss": 0.6319,
+      "step": 15576
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1493204587175738,
+      "learning_rate": 2.3731224785661365e-10,
+      "loss": 0.6448,
+      "step": 15577
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.825954426915879,
+      "learning_rate": 2.273213597553392e-10,
+      "loss": 0.7057,
+      "step": 15578
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9960053908993611,
+      "learning_rate": 2.1754532302076247e-10,
+      "loss": 0.7344,
+      "step": 15579
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6856270069846095,
+      "learning_rate": 2.0798413807365802e-10,
+      "loss": 0.7546,
+      "step": 15580
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6550880506018673,
+      "learning_rate": 1.9863780532480836e-10,
+      "loss": 0.9049,
+      "step": 15581
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2378278034842216,
+      "learning_rate": 1.89506325175004e-10,
+      "loss": 0.7705,
+      "step": 15582
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.216088018086585,
+      "learning_rate": 1.80589698017819e-10,
+      "loss": 0.7908,
+      "step": 15583
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.990040686515409,
+      "learning_rate": 1.718879242357252e-10,
+      "loss": 0.6762,
+      "step": 15584
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6690880255863427,
+      "learning_rate": 1.6340100420286776e-10,
+      "loss": 0.8103,
+      "step": 15585
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7897106157559013,
+      "learning_rate": 1.5512893828395493e-10,
+      "loss": 0.7815,
+      "step": 15586
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2578829094089632,
+      "learning_rate": 1.470717268342581e-10,
+      "loss": 0.5934,
+      "step": 15587
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9126296982187934,
+      "learning_rate": 1.392293702001668e-10,
+      "loss": 0.9283,
+      "step": 15588
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0709449155599544,
+      "learning_rate": 1.316018687191889e-10,
+      "loss": 0.975,
+      "step": 15589
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9699405003846122,
+      "learning_rate": 1.2418922271772993e-10,
+      "loss": 0.8621,
+      "step": 15590
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1116071886631507,
+      "learning_rate": 1.1699143251553414e-10,
+      "loss": 0.6675,
+      "step": 15591
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9931789961633527,
+      "learning_rate": 1.1000849842179862e-10,
+      "loss": 0.7126,
+      "step": 15592
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8217283146487242,
+      "learning_rate": 1.032404207362836e-10,
+      "loss": 0.7359,
+      "step": 15593
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8804373604620839,
+      "learning_rate": 9.668719974986751e-11,
+      "loss": 0.644,
+      "step": 15594
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.014549403377854,
+      "learning_rate": 9.03488357439919e-11,
+      "loss": 0.6946,
+      "step": 15595
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8874123513379448,
+      "learning_rate": 8.422532899121649e-11,
+      "loss": 0.7971,
+      "step": 15596
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3577717897981003,
+      "learning_rate": 7.831667975466417e-11,
+      "loss": 0.8431,
+      "step": 15597
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9817667481177588,
+      "learning_rate": 7.2622888288576e-11,
+      "loss": 0.7807,
+      "step": 15598
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2197332823696825,
+      "learning_rate": 6.714395483720105e-11,
+      "loss": 0.9096,
+      "step": 15599
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0417634223214436,
+      "learning_rate": 6.187987963590658e-11,
+      "loss": 0.6312,
+      "step": 15600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8159943664484786,
+      "learning_rate": 5.6830662911178115e-11,
+      "loss": 0.9274,
+      "step": 15601
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.729539771811265,
+      "learning_rate": 5.199630488006424e-11,
+      "loss": 0.8819,
+      "step": 15602
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0325897010986744,
+      "learning_rate": 4.737680575017667e-11,
+      "loss": 0.7767,
+      "step": 15603
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8911373121433566,
+      "learning_rate": 4.297216571969021e-11,
+      "loss": 0.7043,
+      "step": 15604
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3820441032042203,
+      "learning_rate": 3.878238497845299e-11,
+      "loss": 0.8178,
+      "step": 15605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.927647849491125,
+      "learning_rate": 3.4807463705766044e-11,
+      "loss": 0.6963,
+      "step": 15606
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0034265175936183,
+      "learning_rate": 3.104740207371393e-11,
+      "loss": 0.8534,
+      "step": 15607
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1763329251098933,
+      "learning_rate": 2.7502200242723874e-11,
+      "loss": 0.7496,
+      "step": 15608
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.142325298597799,
+      "learning_rate": 2.417185836545155e-11,
+      "loss": 0.8263,
+      "step": 15609
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9850556730533981,
+      "learning_rate": 2.1056376585115723e-11,
+      "loss": 0.8579,
+      "step": 15610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4674293682249986,
+      "learning_rate": 1.815575503549827e-11,
+      "loss": 0.9103,
+      "step": 15611
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.529739152812592,
+      "learning_rate": 1.5469993840944163e-11,
+      "loss": 0.9258,
+      "step": 15612
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9836067874262675,
+      "learning_rate": 1.2999093117471717e-11,
+      "loss": 0.7915,
+      "step": 15613
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3199308374762286,
+      "learning_rate": 1.0743052971107225e-11,
+      "loss": 0.7877,
+      "step": 15614
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.002935357592006,
+      "learning_rate": 8.701873498995205e-12,
+      "loss": 0.5867,
+      "step": 15615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9608790279995774,
+      "learning_rate": 6.875554787733052e-12,
+      "loss": 0.7573,
+      "step": 15616
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.47162882908754,
+      "learning_rate": 5.2640969172568225e-12,
+      "loss": 0.7692,
+      "step": 15617
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.222922779278431,
+      "learning_rate": 3.867499955845233e-12,
+      "loss": 0.8733,
+      "step": 15618
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.315783904048678,
+      "learning_rate": 2.6857639640054387e-12,
+      "loss": 0.8559,
+      "step": 15619
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2847166425136156,
+      "learning_rate": 1.7188889922525876e-12,
+      "loss": 0.6566,
+      "step": 15620
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0696625065934007,
+      "learning_rate": 9.668750822200423e-13,
+      "loss": 0.6332,
+      "step": 15621
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5920122554118294,
+      "learning_rate": 4.2972226665938255e-13,
+      "loss": 0.8304,
+      "step": 15622
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2234522461213686,
+      "learning_rate": 1.0743056777506866e-13,
+      "loss": 0.7455,
+      "step": 15623
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.380849603295449,
+      "learning_rate": 0.0,
+      "loss": 0.7652,
+      "step": 15624
+    },
+    {
+      "epoch": 1.0,
+      "step": 15624,
+      "total_flos": 2901524184932352.0,
+      "train_loss": 0.824478470777098,
+      "train_runtime": 61973.8267,
+      "train_samples_per_second": 16.136,
+      "train_steps_per_second": 0.252
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 15624,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 2901524184932352.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}